Title Royal Commission into the failure of HIH Insurance Report by the Royal Commissioner the Honourable Justice Owen, April 2003 Volume II-Reasons, circumstances and responsibilities
Source Both Chambers
Date 13-05-2003
Parliament No. 40
Tabled in House of Reps 13-05-2003
Tabled in Senate 13-05-2003
Parliamentary Paper Year 2003
Parliamentary Paper No. 99
House of Reps Misc. Paper No.
Senate Misc. Paper No.
Paper Type Parliamentary Papers
Deemed Paper Type Amendments
Disallowable No
Journals Page No. 1775
Votes Page No. 863
House of Reps DPL No. 162
House of Reps DPL Date 13-05-2003
Number of Deemed Papers 0
Linked Address
Author Body URL
Federal Register of Legislative Instruments No.
URL Description
System Id publications/tabledpapers/19870


Royal Commission into the failure of HIH Insurance Report by the Royal Commissioner the Honourable Justice Owen, April 2003 Volume II-Reasons, circumstances and responsibilities

THE FAILURE OF HIH INSURANCE

Volume II

Reasons, circumstances and responsibilities

April 2003

© Commonwealth of Australia 2003

This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be

reproduced by any process without prior written permission from the Commonwealth available from

the Department of Communications, Information Technology and the Arts. Requests and inquiries

concerning · reproduction and rights should be addressed to the Commonwealth Copyright

Administration, Intellectual Property

Technology and the Arts, GPO

http://www.dcita.gov.au/cca.

Branch, Department of Communications, Information

Box 2154, Canberra ACT 260 I or posted at

National Library of Australia Cataloguing-in-Publication data:

Australia. HIH Royal Commission.

The failure ofHIH Insurance.

ISBN 0 9750678 1 8 (volume II)

ISBN 0 9750678 50 (set)

I. HIH Insurance. 2. Insurance companies-Australia.

3. Insurance- Australia. 4. Business failures-Australia.

I. Title.

368.006594

Printed by National Capital Printing, Canberra

Canberra Publishing and Printing, Canberra

..

Contents

12 The Winterthur years 1

12.1 HIH before Winterthur 12.2 The CIC Insurance group 2

12.3 Early discussions between HIH and Winterthur 3

12.4 Due diligence investigations 3

12.5 Williams's letter of 31 May 1995 8

12.6 Williams as director of CE Heath 9

12.7 Williams's private share sale to Winterthur 10

12.8 Approvals for merger 13

12.9 Winterthur's concerns following the merger 14

12.10 The response to due diligence concerns 14

12.11 The business in the United Kingdom 16

12.12 The US business 19

12.13 The Credit Suisse - Winterthur merger 21

12.14 Winterthur's proposed disposal of its shares 22

12 .15 The sale ofWinterthur's interest 24

13 Unprofitable international operations 35

13.1 Operations in the United Kingdom 35

13.2 Operations in the United States 64

13.3 Operations in Hong Kong and elsewhere in Asia 80

13.4 Conclusion 88

14 The impact of the FAI acquisition 105

14.1 F AI provisioning 105

14.2 The negotiation of the GCRA contract 139

14.3 The negotiation of the NI contract 163

14.4 The preparation ofthe APRA returns and FAI's accounts for the year ending 30 June 1998 176

14.5 The takeover ofF AI 200

14.6 The 'total return swap' transaction 230

14.7 The financial consequences ofHIH's takeover ofFAl 237

14.8 Possible contraventions and referrals 248

The failure of HIH Insurance iii

15 Under-provisioning for claims 289

15.1 Overview ofHIH's business and provisions 289

15.2 Introduction to Slee and his retainer by HIH 296

15.3 Other actuarial assistance 302

15.4 Matters affecting HIH's provisions generally 303

15.5 Liability 334

15.6 Professional indemnity 358

15.7 Workers compensation 365

15.8 Run-off Treaty Account and RMR division 366

15.9 UK operations 375

15 .10 US operations 386

15.11 Slee 419

15.12 Andersen 424

15.13 The directors 434

15.14 Possible contraventions and referrals 443

16 Use of whole-of-account reinsurance 467

16.1 The significance of reinsurance in the failure of HIH 467

16.2 Arrangements with Hannover 468

16.3 Swiss Re 500

16.4 Another whole-of-account reinsurance contract 512

16.5 HIH and the NI and GCRA AXOL contracts 513

16.6 Information acquired prior to the takeover 513

16.7 Discovering the side letter 514

16.8 Accounting treatment for period ending 30 June 1999 516

16.9 The appropriate accounting treatment for period ending 30 June 1999 518

16.10 Misleading conduct and audit approach for period ending 30 June 1999 520

16.11 Amount of recoveries available under the NI contract 520

16.12 Accounting treatment for year ending 30 June 2000 522

16.13 The appropriate accounting treatment for year ending 30 June 2000 525

16.14 Misleading conduct and audit treatment during 2000 526

16.15 Andersen's audit approach 527

16.16 Possible contraventions and referrals 529

iv Contents

17 The demise of HIH 544

17.1 The Allianz transaction 544

17 .2 The concerns of W estpac and the retention of Ernst & · Young 560

17.3 Cash flow difficulties and the involvement of APRA and ASIC 570

17.4 Events subsequent to 4 December 2000 572

17.5 The final cash crisis 576

17.6 Asset sales and joint ventures 580

17.7 The retention ofKPMG and the final days 587

17 .8 Last minute payments to creditors 593

17.9 Provisional liquidation and the assessment of the ultimate deficiency of assets 597

17.10 Possible contraventions and referrals 598

The failure ofHIH Insurance v

VOLUME I

Explanatory notes

The failure of HIH: a critical assessment

Policy recommendations: a summary

Understanding this report

2 How the inquiry was conducted

3 A brief corporate history

4 The industry and regulatory context

5 Provisioning and reinsurance: general principles

6 Corporate governance

7 Financial reporting and assurance

8 Regulation of general insurance

9 State and territory regulation

10 Taxation and general insurance

11 A policyholder support scheme

Appendix A

Appendix B

Appendix C

Appendix D

Appendix E

Appendix F

Appendix G

Appendix H

Glossary

Abbreviations

vi

Terms of reference

Parties

Witnesses

Policy submitters

Operational statistics

The Commission team

Offence provisions: an outline

Correspondence from the public

Contents

VOLUME III

18 The inadequacy of management information

19 Regulatory solvency

20 The effect of incorrect accounting

21 The audit function

22 Home Security International: a case study

23 Other aspects of the governance of HIH

24 The regulators

'

Th e failure of HIH Insurance vii

..

12 The Winterthur years

During the three years from 1995 to 1998 the HIH share register was dominated by Winterthur, an insurance company based in Switzerland. Although it owned 51 per cent of the shares in HIH, Winterthur chose not to exercise that control by seeking a majority on the HIH board. Ray Williams and his fellow executive directors retained effective control.

There was an inherent tension in this arrangement, and it was evident throughout 1995 to 1998, a period referred to here as 'the Winterthur years' . The tension was fuelled by some fundamental differences in approach between the two groups.

Winterthur considered it had an understanding with HIH that the focus of operations would be largely, if not exclusively, the Australasian market. In contrast, HIH­ mainly at the urging of Williams-saw expansion into the United Kingdom and the United States as a priority. This conflict over the company's basic strategy was never satisfactorily resolved.

Another deep-seated difference concerned operating cultures and procedures. In this, Winterthur was conservative compared with HIH, which was entrepreneurial. As a result, the two companies often had different approaches to questions such as risk management and other aspects of corporate governance.

It was during these years that the early warning signs of what were later to become major problems within HIH emerged; in particular, these included HIH's under­ reserving and the deterioration in the company's UK and US operations. Throughout the period Winterthur regularly communicated its concerns in this regard to the HIH board and, ultimately, the losses in those areas contributed significantly to the collapse of HIH. These were the years when the problems that

led to the failure ofHIH began to emerge.

12.1 \ HIH before Winterthur

In the early 1990s HIH saw a need for fresh capital in order to strengthen its position in the market. On 6 April 1992 it issued a prospectus and was listed on the stock exchange. New shares were offered for subscription to provide capital to fund significant long-term opportunities for growth 1, and HIH raised the capital necessary to enable it to pursue a growth strategy.

As a result of the flotation, 45 per cent of the issued capital was owned by the public, 44 per cent by CE Heath pic and 11 per cent by the Australian management. 2 But the float was not entirely successful: the listing on the Australian Stock Exchange was achieved only after Terence Cassidy, George Sturesteps and Williams agreed to take up additional shares. Williams had to borrow $10 million

The failure of HIH Insurance

from Hambros Corporate Advisory to meet his commitments as a sub-underwriter. It was at this time that Colin Richardson first met Williams. Richardson was a director of Hambros and was the investment banker retained by HIH in connection with its acquisition and disposal of businesses during the next decade. HIH was a subsidiary of CE Heath, in which Hambros held a substantial stake and to whom it was financial adviser. 3

Williams attributed the limited success of the float to several factors , among them competitors besmirching HIH's reputation; an analyst circulating a negative report on the company; and the lead underwriter, Ord Minnett, failing to support the stock, contrary to advice that it would do so. 4 In any event, the negative perception of HIH persisted after the float and the share price remained flat. 5

As part of the arrangements for the float, Williams, Cassidy and Sturesteps entered a private agreement with CE Heath, whereby the latter would pay them $6.22 million as compensation for them forgoing their special remuneration service agreements. Williams received by far the largest portion of that sum. 6

In the prospectus published in April 1992 CE Heath announced that it was a long­ term investor in HIH and, subject to various qualifications, made a commitment to refrain from selling any of its shares for at least two years from the date of the prospectus. 7 In spite of this, in 1993 CE Heath reduced its shareholding in HIH from 70 754 133 8 to 41 649 695 9 shares. It appears that CE Heath was perceived as a potential seller of its HIH stock from this time. 10 In response, HIH sought to achieve two concurrent objectives- the replacement of its majority shareholder and a dramatic increase in the size of its business.

When Hambros, on behalf of HIH, began the search for a shareholder to replace CE Heath, it identified Winterthur and its subsidiary CIC as suitable candidates.

12.2 The CIC Insurance group

At the time CIC was the tenth-largest private insurer in Australia, with a market share of 5 per cent. 11 In June 1994 Hambros identified CIC as an under-performing insurer with an uncertain future. 12 Ultimately, a proposal involving a merger of the operations of HIH and CIC materialised. This was motivated by a range of considerations- improved balance sheet strength, improved capital-raising prospects, share register stability, strategic expertise, and so on.

The CIC Insurance group had been formed in January 1987, the holding company being CIC Holdings Limited and the authorised insurer being CIC Insurance Ltd. By late 1991 Winterthur held 67 per cent of the issued capital in CIC Holdings. The directors of CIC Holdings were Alexander Gon·ie, Charles Abbott, Willi Schurpf and Randolph Wein. 13

2 Th e Winterthur years

12.3 Early discussions between HIH and Winterthur

At the start of 1993 Richardson contacted Winterthur on behalf of HIH. At first Winterthur expressed no interest in a merger. 14 Hambros' initial analysis of CIC, in June 1994, identified an increase in market capitalisation upon merger that would benefit the share price ofHIH. 15

In mid-1994 Hambros made a presentation to Winterthur in respect of its interest in CIC Holdings 16, and the HIH board held discussions in relation to the negotiations with Winterthur.17 Senior management of Winterthur became more favourably disposed towards the merger proposal. 18

12.4 Due diligence investigations

HIH and Winterthur signed confidentiality undertakings in February 1995, to facilitate discussions and the exchange of information between Winterthur, CIC and HIH. 19 From the outset it was critical to Winterthur that it would have a majority shareholding and the ability to move to board control. 20

12.4.1 Williams's role in negotiations

At that time Williams was a director ofCE Heath, having been appointed in 1980.21 The potential for a conflict of interest arose as a result of Williams's directorship of CE Heath, the potential vendor, and his position as a director, chief executive officer and significant shareholder of HIH, the potential buyer. The interests of the major

shareholder (CE Heath) in seeking to quit its shareholding on the most favourable terms were not identical to the interests of HIH. No action was taken at any time to confront or resolve this conflict.

In early March 1995 HIH announced its results for the year ending 31 December 1994, describing it as 'easily the best year on record for the Company' .22

Williams negotiated the major elements of a merger transaction with Wein, who was chief regional executive ofWinterthur in the Asia-Pacific area. 23 The purchase price was to be based on a formula that took account of both net tangible assets as at 31 December 1994 and earnings. Winterthur was to acquire the remaining shares in CIC plus CE Heath's shareholding in HIH and would be responsible for obtaining the resignation of the CIC chief executive. Winterthur would also be invited to

appoint three representatives to the HIH board, which would be enlarged to 10 directors. In addition, it was to warrant the reserves of CIC, although HIH initially indicated it was in no position to give a similar warranty. 24

The failure ofHIH Insurance 3

12.4.2 Heads of agreement: April1995

In late March 1995 Williams travelled to Switzerland for further negotiations with Winterthur. 25 Heads of agreement were signed on 11 April 1995 ? 6 A joint media release from HIH and Winterthur27 announced entry into heads of agreement on th e basis that HIH would acquire CIC from CIC Holdings. The consideration for the acquisition was to be the issue of approximately 95 million HIH shares. In addition, CIC Holdings agreed to acquire CE Heath's shares in HIH for $1.50 a share.

The media release stated that this would result in Winterthur becoming the largest shareholder in HIH, with a holding of approximately 48.2 per cent. The transaction remained subject to entry into formal agreements, the completion of due diligence, HIH shareholders' approval of the transaction, and compliance with regulatory

. 28

reqmrements.

On the same day Richardson wrote to Hambros in the United Kingdom, enclosing the press release. He stated that, as a result of the transaction, HIH's gross written premium would rise from $350 million in 1992 to $1.2 billion and that ' HIH will be the largest listed writer of insurance in Australia . . . unfortunately Ray is already talking about acquiring one of his competitors' ?9 This was a flippant reference to the possible acquisition ofFAr.J0

12.4.3 Due diligence inquiries

CIC Holdings established a board due diligence committee relating to the merger with HIH; the committee consisted of Gorrie, Abbott and W ein. CIC commissioned Ernst & Young to prepare a due diligence report and Blake Dawson Waldron to undertake legal due diligence.

By early May 1995 Winterthur had received a number of reports, from both its actuary and Ernst & Young, expressing unease about HIH's reserving and actuarial practices. 3 1

Ernst & Young sent to Wein a summary that noted that HIH's reserving policies were not as conservative as CIC's and raised other concerns, including the fact that HIH did not maintain a prudential margin. The summary also questioned the independence of David Slee, HIH' s consulting actuary, and commented that the due diligence report by the Winterthur actuarial team revealed very unfavourable underwriting. 32

In May 1995 33 , at a CIC board due diligence committee meeting, Ernst & Young tabled a preliminary report on the matters that had arisen in connection with the HIH due diligence and expressed further concerns. 34

The minutes also record a comment

about a need for more legal documentation and the absence of a clear 'paper trail' leading to important decisions. 35

Slee presented to the HIH due diligence committee an interim report on CIC expressing anxiety about the value of the CIC business. 36 Richardson wrote to Winterthur's financial advisers, Potter Warburg, reporting that the due diligence investigations into CIC led to significantly lower values for that company, since

4 The Winterthur years

both the historic results and the forecasts were worse than originally thought. As a result, Potter W arburg suggested that a lower number of shares be issued to Winterthur, which would give it 44.12 per cent of the enlarged capital rather than 48 per cent originally contemplated. 37 Similarly, on 2 May 1995 Hambros sent to Winterthur a facsimile listing a number of matters affecting the value of CIC?8 The

suggestion of a 44 per cent shareholding was rejected by Winterthur: its objective was to have a majority shareholding in HIH. Ultimately, CIC and HIH offered each other an indemnity that provided that, in relation to certain of their portfolios of business, the total reserves booked for each of their insurance portfolios, calculated as at 30 June 1995, would be compared in two years' time and any deficiency in reserves would require one party to indemnify the other. Liability would arise under the indemnity only if the shortfall exceeded 5 per cent of the warranted reserves. Each party's liability was capped at $20 million. 39

Among other material issues arising from HIH's due diligence on CIC was an assertion that CIC's historical results had been distorted by the creation and subsequent release of prudential margins, particularly in 1994, when $34.4 million of prudential margins was released to the profit-and-loss account. Adjustments to eliminate this distortion and an increase in the net discount rate in 1992 and 1994

resulted in CIC incurring underwriting losses of over $40 million. Prospectively, the restated forecasts led to an underwriting result of a $31 million loss. Most lines of CIC's business were reported to be worse than HIH had originally been told. 40

HIH provided a written response to the Ernst & Young due diligence queries, rejecting all the concerns that had been raised.41 Slee also rejected the concerns expressed by Ernst & Young's actuary. 42

At a CIC board due diligence committee meeting held on 16 May 1995 Ernst & Young reported on outstanding matters, particularly claims reserving for workers compensation, public liability, professional indemnity and marine, and noted that there were still significant differences.43 In fact, Ernst & Young's actuarial report stated that HIH's central estimate was under-reserved by $60 million and that future claims-handling costs were understated by $9.9 million. An HIH note reported that after further discussions the variance in the central estimate was reduced.

44

12.4.4 The Blake Dawson Waldron due diligence report

On 18 May 1995 Blake Dawson Waldron issued its legal due diligence report on HIH for CIC; the report noted a number of major concerns 45 :

• HIH had not made a complete transition from an entrepreneurial company strongly influenced by senior management-and from which senior management benefited considerably-to a company listed on the Australian Stock Exchange and run primarily in the interests of shareholders.

• Price adjustments remained under the agreement for HIH's sale of Heath Cal, which was completed on 19 August 1994.

The failure of HIH Insurance 5

• A lack of coordination, stemming in part from HIH's decentralised management structure, gave rise to 'potential exposure at corporate governance level'.

• The terms of service agreements with directors and senior executives, whereby payments to Williams, Cassidy and Sturesteps amounted to $6.22 million, were noted.

• Retirement deeds had been entered into with non-executive directors, whereby Neville Head, Geoffrey Cohen and Robert Stitt were to be paid up to three times their annual average emoluments upon retirement.

• Employee loan schemes, employee share-ownership plans, executive option plans and incentive bonus schemes for executive directors and semor management were noted.

• Property and joint-venture projects-the most important of which was the Forum site at St Leonards-were controlled by two associates of HIH, Gary Rothwell and Matthew Braden through Rimcorp Capital. HIH had provided funds without the protection of any joint-venture documentation.

• Rothwell and Braden had a close relationship with HIH and Williams and Cassidy. Williams, in particular, had been a director of Rimcorp Capital, which paid substantial amounts ($228 000) in aggregate directors fees.

12.4.5 The Ernst & Young due diligence report

Ernst & Young' s financial due diligence report was finalised on 18 May 1995. The financial adjustments in relation to central estimate differences, prudential margin differences, the marine portfolio estimate, property valuation differences, and taxation issues amounted to $87 790 000.

46

In the report's executive summary

various questions were raised in connection with financial due diligence-including the fact that HIH' s auditors, Andersen, had assessed the audit of HIH as one of 'maximum audit risk'. This was because of factors that suggested management had the ability and incentive to 'manage' the reported result and to adopt potentially

invalid positions and policies. 47

Ernst & Young also noted that additional reserves amounting to $64.19 million would be needed if the HIH reserves were to have a 75 per cent probability of sufficiency.48 In its opinion, the reserving methodology adopted by HIH was 'very aggressive and appears to have already proven inadequate' .

49 It was also of the

opinion that the estimation of liabilities for the UK branch was subject to significant uncertainty as a result of the immaturity of the portfolio; it recommended that a prudential margin be adopted.

6 Th e Winterthur years

Among other concerns, Ernst & Young reported:

HIH has traditionally been under strong influence by senior management and may be described as having a strong entrepreneurial flavour. The management structure of HIH is also very decentralised. The impression formed by a number of individuals involved in the due diligence review is that the management style and approach of HIH is very different to the management style and approach ofCIC.50

Ernst & Young had encountered problems in receiving on a timely basis the information it sought from HIH. In fact, some requested information had not been provided and on occasion statements made by HIH officers were subsequently found to be incorrect.

51 Ernst & Young noted the possibility of a 'high incidence of potentially controversial transactions and accounting treatments'. 52

The board of HIH met in May 1995 53 and received from Hambros a briefing paper that noted certain of the financial issues identified in the Ernst & Young report54, albeit in a summary form that was inadequate to put the board on notice of the seriousness of the matters raised. Richardson stated that he did not see the final due diligence report on HIH.55

At a CIC board meeting on 21 May 1995 Abbott briefed the directors on his recent discussions with Williams concerning HIH's commitment to high standards of corporate governance and legal and accounting compliance. Williams's assertions were at odds with the due diligence reports CIC had received from Ernst & Young and Blake Dawson Waldron. The directors approved the share sale agreement, and Blake Dawson Waldron and Ernst & Young gave each other a copy of their due diligence reports. 56

12.4.6 The outcome

Richardson considered that both HIH and CIC had an incentive to find as many valuation points as possible during the due diligence process- in order to gain a commercial advantage in the merger. 57 Richardson's evidence is an accurate reflection of HIH's approach to the due diligence exercise. Legitimate criticism and concerns expressed by CIC were not assessed on their merits but were dismissed by HIH for short-term commercial advantage.

It is clear that during the due diligence process both HIH and CIC identified the principal risks in each other's business. CIC was not as profitable as its financial statements suggested: its profitability had been inflated by the release of prudential margins that-along with questions about its reserving practices- had the potential to lead to further significant losses. HIH was replete with present and potential problems, including under-reserving, lack of internal controls, conflicts of interest, and a culture of entrepreneurial expediency. The mixture of the identified problems and risks for both entities did not augur well for the future of a combined group.

Many of the observations both Ernst & Young and Blake Dawson Waldron made about HIH's reserves and accounting practices, and its management culture, came to

The failure of HIH Insurance 7

have a haunting poignancy. They were as true in 2001, when the HIH group collapsed, as they had been in 199 5. In the intervening six years the HIH board failed to confront and resolve the major problems that had become so evident during the due diligence process.

12.5 Williams's letter of 31 May 1995

Throughout the Winterthur years, international expansion by HIH was a significant source of tension and misunderstanding between Winterthur and HIH. Ultimately, HIH's international operations- particularly those in the United Kingdom­ contributed to Winterthur's decision to dissociate itself from HIH.

On 25 May 1995 David Williamson of Blake Dawson Waldron sent to Wein and Hans-Peter Naef, Winterthur's vice-president Asia-Pacific, a facsimile entitled 'Gentlemen's Agreement with Ray Williams'; attached to it was a draft letter from Williams to Schurpf, described as a draft 'side letter'. 58 The letter, which was subsequently signed by Williams on 31 May 1995, was addressed to Peter Spalti, chairman and chief executive of Winterthur. In it Williams welcomed Winterthur's investment and stated that the ' current members of the HIH Board look forward to cooperating with Winterthur in every way possible to advance the joint interests of both Companies'. Williams also noted a number of areas in which HIH could take advantage of the expertise and economies of scale Winterthur offered and highlighted the focus on HIH's expansion in Australia (as opposed to internationally), confirming the 'geographical focus as to becoming a major player in close cooperation with [Winterthur] and its network in Austral-Asia'. Further, Williams acknowledged that various shortcomings that had been identified in CIC's due diligence of HIH did require HIH's attention and that HIH would work closely with Winterthur to remedy the situation. 59

In closing submissions to the Commission, Williams asserted that in the letter he signed on 31 May 1995 he did not give an assurance that HIH would confine its geographic focus to Australasia.60 But it is clear from the terms of the letter that Winterthur was seeking comfort in relation to HIH's international ambitions and wished to confine the focus of those ambitions to Australasia. Schurpf gave evidence that, before the acquisition, Winterthur considered it had reached agreement with HIH that HIH would not expand in any significant way on an international basis. 61

Richardson made a file note recording details of a June 1995 discussion with Leigh Brown of Minter Ellison concerning Williams's letter of comfort of 31 May 1995. The note disclosed that Hambros had not seen the letter and that Brown had advised that Hambros had no responsibility to insist on seeing the letter. In the note, Richardson stated that he had confirmed with Williams that nothing in the letter served to make the information memorandum misleading and that the HIH board would approve the letter, the contents of which Williams had disclosed to the

8 Th e Winterthur years

board. 62 Richardson understood the 31 May letter to be a 'feel good' private letter between Williams and Spalti.

Williams testified to the Commission that the chairman and members of the HIH board discussed the 31 May letter and confirmed that part of the arrangement with Winterthur was that development should focus on Australasia. 63 But the chairman, Cohen, denied that Williams had shown him the letter and claimed that he would not have dealt with Winterthur as he did in 1997 and 1998 had he known of the letter's existence. 64

Cohen's evidence on this topic is preferable to that of Williams. It is consistent with the objective evidence of his actions and is supported by the evidence of Head, who said that he had no recollection of the letter being tabled at a board meeting or of having seen it prior to giving evidence to the Commission. 65 Before the acquisition he was not aware of any discussions or advice that Winterthur was concerned about HIH expanding its operations overseas. Sturesteps could not recall seeing the letter

before he gave evidence, although his interpretation of it from the witness box is not one with which I can agree. 66

Williams may have thought the board members knew about the letter and the geographical limitations it expressed. 67 They did not. It is strange that the chief executive thought he had authority to execute a letter of such potential significance without the formal approval of the board. If the approval were to be obtained through informal discussions it should have been done with clarity and precision.

12.6 Williams as director of CE Heath

Because he was a director of CE Heath pic, the vendor of HIH shares to Winterthur, Williams withdrew from the HIH board meeting held on 19 May 1995 , when the transaction was discussed and approved. 68 He testified that the chairman or Richardson must have told him it was not appropriate for him to be present and he took the advice. He acknowledged that he was the prime mover in bringing this transaction to fruition and that he negotiated with CE Heath. But he said he gave no consideration to whether he had a conflict of interest arising from his directorship of CE Heath. In his view, it was his task to negotiate with the chief executive of CE Heath, Peter Presland, who was also a director ofHIH at the time.69

The CE Heath - HIH negotiations should not have taken place between Williams and Presland. They both had obligations to CE Heath that had the potential to conflict with the interests of HIH, to whom they both also had fiduciary obligations. Williams submitted that he was acting under the express authority and with the approval of all directors of CE Heath, HIH and Winterthur. 70 Nonetheless, the central role he played in negotiating this transaction was inconsistent with his action

in withdrawing from the board meeting at which HIH resolved that the share sale agreement be approved for execution.

Th e failure of HIH Insurance 9

12.7 Williams's private share sale to Winterthur

Williams negotiated the sale to Winterthur of approximately 6 million of his personal shareholding in HIH at a time when he had negotiated on behalf of eE Heath for the sale of its HIH shares at a significantly lower price. Further, Williams failed to make appropriate disclosure to HIH shareholders and the HIH board of his private share deal with Winterthur. Both Winterthur and Williams had always understood that this transaction was part of the overall arrangements.

Williams had advised the HIH board meeting in May 1995 that he might enter into an option agreement with Winterthur for Winterthur's purchase of a portion of his shareholding in the company. The chairman remarked that such an agreement would require disclosure to the shareholders in the information memorandum.71

Williams testified that Winterthur was anxious to increase its shareholding to 51 per cent by 31 December 1995, to allow for consolidation of the HIH and Winterthur results, but that it was not possible for Winterthur to acquire the additional 2 per cent on the stock market. 72 That proposition is difficult to accommodate. Winterthur could have purchased the necessary HIH shares on the market or HIH could have issued them to Winterthur and obtained shareholder approval for that transaction in a legitimate and transparent way.

Williams said that the only 'readily available' source of the additional 2 per cent was the shares he held- in particular, the 6.6 million shares for which he had borrowed $10 million from Ham bros in order that the 1992 float might proceed. Williams explained that he agreed to sell those shares to Winterthur, thus enabling them to increase their shareholding to the required 51 per cent; the shares were sold to Winterthur in December 1995. Williams confirmed that after repayment of the $10 million plus outstanding interest to Hambros, the transaction resulted in a profit to him of approximately $3 million. 73

12.7.1 The synopsis of agreement: May 1995

On 25 May 1995 Hambros provided Wein with a synopsis of the Winterthur­ Williams agreement in relation to the shares. 74 In essence, the arrangement would

have given ere Holdings a right of first refusal over 6 million ordinary shares until 1 December 1995. Between 1 December 1995 and 30 June 1996 ere was to be granted a call option from Williams at an exercise price of the market price or $2.05, whichever was the greater. A put option was to be granted to Williams on the same terms. 75

Williams claimed he also became aware that it was important for Winterthur not to lose control of ere. 76 Richardson gave evidence that one reason for Winterthur not wanting to lose control of ere was the shareholders agreement covering AAMI. Under that arrangement, Winterthur could lose its interest in AAMI if it lost ultimate control of ere, and AAMI was thought to be an attractive ere asset and thus something the restructured group needed to retain. 77

10 Th e Winterthur years

Richardson said that, to achieve the objective, Williams granted a call over part of his personal holding in HIH, representing 3 per cent to Winterthur, together with the voting rights associated with those shares. 78 Williams, however, denied that there was such an agreement between him and Winterthur. 79 In fact, he gave evidence that an agreement pursuant to the synopsis of 25 May 1995 80 was not entered into and that a sale at $2.05 was under negotiation but was never realised. 81 Wein ' s evidence was to the effect that 'a gentlemen's agreement' operated during 1995. 82

In closing submissions Williams asserted that there was no evidence that the option agreement was entered into. 83 Even accepting that the formal agreement was never executed, there was an arrangement or understanding to a similar effect: the suggestion that the arrangement struck in December 1995 was not foreshadowed during the first half of 1995 cannot be accepted.

While Williams was negotiating with Winterthur in the first part of 1995 , HIH ' s share price fluctuated from a low of $1.24 to a high of $1.40. 84 Had Williams been required to repay the $1 0 million to Ham bros at that time, the proceeds of a sale at that price would have resulted in a significant shortfall for Williams, which would have had to have been met from his private resources. In closing submissions Williams rejected the assertion that he did not sell his shares at the same time as CE Heath because of the fact that that would have resulted in a loss on the shares he

had bought at $1.50. He claimed that in June 1995, when the agreement was finalised, the HIH share price was approximately $1.7 5 and that, accordingly, he could have achieved a substantial profit if he had wanted to sell his shares to Winterthur then. 85 It must, however, be remembered that at the very time Williams was negotiating the sale of his own shares at a price in excess of $2 .00, he was

negotiating the sale of the parcel held by CE Heath at a price of $1.52.

Williams confirmed that Richardson put a price of $2.05 on the shares after talking to him. But he could not recall discussing the figure of $2.05 with Presland at CE Heath or whether it might have been possible to negotiate a transaction at that price for CE Heath's shareholding in HIH.86 Williams saw no incongruity between the conclusion he had arrived at in his role as a director of CE Heath-that $1 .52 was a fair price for HIH shares-and the idea that Winterthur was to pay him personally

$2.05 .87

The information memorandum tabled at the HIH board meeting on 13 July 1995 disclosed the following arrangement between Williams and CIC Holdings:

As stated previously throughout this document, on Completion of the Transaction, CIC Holdings will hold a 48.4 per cent shareholding in Heath Insurance.

Mr R.R. Williams (who is the Chief Executive Officer of Heath Insurance) has executed an 'Appointment Deed'. When effective this deed will give CIC Holdings the right to exercise the votes attaching to 6.5 million Ordinary Shares held by him (representing approximately 2.3% of Heath Insurance's enlarged issued share capital) on shareholders' resolutions relating to the appointment or removal of Directors. The Appointment

The failure ofHIH Insurance 11

12.7.2

Deed (and the rights granted thereunder to CIC Holdings) will terminate upon the earlier of:

(i) CIC Holdings acquiring (or acquiring voting control of) 50% of the total issued capital of Heath Insurance;

(ii) CIC Holdings and Williams each agreeing to terminate the Appointment Deed; and

(iii) Completion not occurring on or before 15 July 1995 .

The principal purpose of this arrangement is to ensure that after the acquisition of CIC by Heath Insurance, certain assets of the Group including its shareholding in the AAMI Insurance Group are retained.88

The agreement of 22 December 1995

Williams did subsequently sell a number of shares to Winterthur, as evidenced by a letter dated 22 December 1995. He lodged a substantial shareholder's notice annexing the letter pursuant to which he agreed to sell the parcel of 6.5 million shares to Winterthur for $2.03 a share. The arrangement included a term stipulating that, if HIH's pre-tax profit performance exceeded specified amounts that he and Wein had agreed for the years 1995 , 1996 and 1997, Williams was to receive an amount equal to 10 per cent of the excess. The relevant amounts were not specified

in the notice, and this aspect of the transaction was not disclosed to the board of HIH. There was also a term entitling Williams to an amount equal to the dividend declared on the shares in March 1996. Wein signed the letter on behalf of Winterthur in confirmation of agreement with its terms. 89

12.7.3 Directors' knowledge of the arrangements

The directors of HIH had differing levels of knowledge about these arrangements. Wein confirmed that it was always Winterthur's idea to move to 51 per cent90 and that there were discussions with Williams as to how he would sell enough shares to bring Winterthur up to that level. 91

With the exception ofWein, in view of their participation at the board meetings, the directors of HIH either knew of the transaction but not the detail or had grounds to suspect such a transaction. 92 The directors knew that Winterthur wanted to obtain a 51 per cent shareholding in HIH, and they had been put on notice at the May 1995 board meeting that Williams was considering entering an agreement with Winterthur to sell part of his shareholding. In those circumstances the directors of HIH should have confirmed whether Winterthur did intend to acquire additional shares to become a 51 per cent shareholder. They should also have inquired whether Williams had an arrangement to help Winterthur achieve that objective and, if such an arrangement had been entered into, whether that arrangement required disclosure in the information memorandum.

12 The Winterthur years

12.8 Approvals for merger

12.8.1 The June 1995 meetings

On 9 June 1995 an HIH board and an audit committee meeting were held consecutively. At the board meeting a draft information memorandum for submission to the Australian Stock Exchange and to shareholders, for approval of the entry into the share sale agreement, was tabled and discussed.93 Later that

morning Williams reported to the audit committee that the result of the due diligence carried out on HIH in relation to the acquisition of CIC was satisfactory. 94 It appears that nothing of substance was said at the meeting about the numerous due diligence concerns identified in the reports of Blake Dawson Waldron and Ernst &

Young. Williams testified that he could not recall seeing those reports.95 The meeting lasted only 20 minutes and no other director could recall whether any of the substantial concerns identified in the due diligence reports were raised at the meeting. It is not at all clear how Williams could have reported that the result of the due diligence was satisfactory or why this should have been reported to the audit committee rather than the board.

12.8.2 The information memorandum

The information memorandum for shareholders on the acquisition of CIC by HIH and the acquisition by CIC Holdings of a 48.4 per cent shareholding in CE Heath was issued on 16 June 1995.96 Clause 12 of the memorandum recorded that Williams and Presland had not voted on any directors' resolutions pertaining to the transaction because they were associated with one of the vendors, CE Heath. But both Williams and Presland recommended that shareholders vote in favour of the transaction.97

Because Winterthur was to acquire over 20 per cent of the shareholding ofHIH, but without making a full takeover bid, an extraordinmy general meeting of HIH was required. This meeting was held on 13 July 1995. The resolutions were all passed in terms of the information memorandum.98 At the board meeting held later that day,

Pres land resigned as director and Erwin Heri, Schurpf and Wein were appointed. 99

HIH then issued a media release announcing the completion of the transaction. It noted, 'Heath Insurance is now poised to capitalise on growth opportunities within the Australian insurance sector' .100

The first board meeting of the merged entity was held on 30 August 1995. Abbott, Schurpf and W ein attended and were welcomed. Gorrie was appointed an alternate director for Schurpf and Abbott as an alternate director for Heri. 101

The failure ofHIH Insurance 13

12.9 W interthur's concerns following the merger

After the merger Winterthur sought to redress some of the problems identified during the due diligence process. Accordingly, Winterthur representatives introduced to HIH the guidelines and internal control procedures that were maintained by Winterthur internationally. Greg Waters, Winterthur's regional internal audit manager Asia-Pacific, was appointed general manager of internal audit for HIH Insurance. 102

In a memorandum prepared in August 1996 Winterthur noted that HIH's corporate culture was 'a closely knit and related club where one is either part of it or "out'" Williams was described as having a 'charismatic and somewhat patriarchal leadership'. These two factors were said to have led to inadequate communications and information sharing with Winterthur. 103

The memorandum noted that Winterthur received only limited information other than the board papers. Important decisions were made at divisional board level, where there was limited, if any, access by Winterthur. It also noted that there were insufficient independent checks and ba lances to verify the accuracy and appropriateness of the information that was supplied. Winterthur considered that the internal audit function should be strengthened, particularly in view of the past close relationship between Andersen and HIH. 104

At a meeting with Williams in September 1996 Schurpf raised concerns about the information provided to directors, who he said were not adequately informed or consulted before HIH decided on major new projects.105

By early 1997 Winterthur considered it had been reduced to a fi nancial investor, with little influence on strategy and practically no say on operational matters. Spalti addressed the HIH board in February 1997, noting that Winterthur expected the board and management ofHIH to agree to allow Winterthur a reasonable exercise of its authority by giving it enough time and information about new projects, so as not to be confronted by a fait accompli. In a script prepared by or for Spalti it was noted that, in the absence of compliance, Winterthur would need to move to a majority position on the board or reconsider its position as HIH's majority shareholder. 106 It is not clear that he actually put the position as bluntly as that.

12 .10 T he response to due diligence concerns

Abbott, Gorrie, Heri, Schurpf and W ein all had full access to the due diligence reports on HIH prepared by Ernst & Young and Blake Dawson Waldron. Abbott and Wein gave evidence of discussions with Williams and other directors concerning the issues raised in the reports 107, but there is no evidence that the non­ ere directors called for or saw the reports.

14 The Winterthur years

Williams stated that there could be no melding of the corporate cultures of HIH and CIC; as a result, CIC staff needed to 'embrace our culture' or leave.108 Cassidy and Sturesteps claimed that the criticisms raised were based on misconceptions and that any initial concerns diminished in significance.109

Discussion of the matters raised in the Ernst & Young and Blake Dawson Waldron reports occurred only at a general level. 110 Abbott asserted that such discussions, although general, were held at board level. 111 He also testified that he had raised a number of due diligence questions informally with Williams but that Williams had dismissed them.

112 Abbott believed that Andersen would have raised a number of the matters identified in the Ernst & Young report if it had had concerns. 113 In any event, Andersen did not tell the directors that it had assessed HIH as a ' maximum

audit risk'. Davies confirmed this. 114

W ein said that the due diligence reports were not made available to the HIH board for reasons of confidentiality, although excerpts from the Ernst & Young report were given to Williams. No one from HIH ever asked for the reports.115

Cassidy testified that he was aware due diligence was carried out but could not recall seeing a copy of the reports. 116 Similarly, Cohen stated he could not recall the due diligence arrangements being discussed by the board and confirmed that the Ernst & Young report was not tabled or referred to at any meeting of the audit committee or the board. 117 Gorrie was also aware of the Ernst & Young report but

did not recall specific discussions about it; he did not think to inform the HIH board of the matters he had become aware of as a result of the due diligence process. 118

As a result of his discussion with Wein, Williams was, at the very least, aware of the general tenor of the due diligence reports. He advised the board that due diligence had been performed satisfactorily. But in many respects the results expressed in those reports were unsatisfactory. Williams failed to advise the board of the criticisms in clear and direct terms. Similarly, the directors of HIH ought to have obtained detailed information about the results of the due diligence undertaken by

Ernst & Young. They did not.

Winterthur had sufficient presence on the HIH board to give comfort to the non­ executive directors that the problems identified were being dealt with. Nevertheless, the series of complaints Winterthur made at various board meetings during 1996 and 1997 about the level of information provided to it should have alerted those

directors to the fact that such comfort was limited and diminishing.

The internal audit reports prepared by Winterthur revealed various weaknesses in corporate governance. Not enough was done by the directors of HIH to take action in response. When reports revealing continuing shortcomings were produced the directors should have involved themselves in formulating and implementing a plan to resolve the problem and then should have monitored the remedial action. Nowhere is the evidence of this failure more clear than in the rapid expansion of

HIH's operations in the United Kingdom.

The failure ofHIH Insurance 15

12.11 The business in the United Kingdom

There was continuing tension between Winterthur and HIH over the entire UK strategy of HIH and the way the UK operations were being managed. Ultimately, Winterthur saw only one solution to the problems it encountered with HIH's UK operations-- to completely withdraw from involvement in HIH.

The letter of comfort Williams wrote on 31 May 1995 led Winterthur to understand that HIH's future expansion would be confined to Australasia and that the geographic focus would not be international:

Whilst recognizing that HIH is an independent listed company with the majority of its shares for the time being held by a wide range of

shareholders, I and the other current members of the HIH Board look forward to co-operating with Winterthur in every way possible to advance the joint interests of both Companies. In particular, we consider it important that Winterthur be fully involved in:

• HIH's planning to ensure HIH takes fully [sic] advantage of the expertise and economies of scale which can be introduced by Winterthur in areas such as:

geographical focus as to becoming a major player in close cooperation with WSI [Winterthur] and its network in Austral­ Asia.119

In the final paragraph of the letter Williams stated, 'I feel that the proposed transaction will bring the two entities with complimentary [sic] strengths together which in the process will create a unique insurance carrier of formidable proportions in Australia' .120

During the course of the due diligence review, Winterthur briefly looked at HIH's London business. W ein explained, however, that Ernst & Young, who conducted the review, was not overly interested in the London business, since it was not significant in HIH's overall operations and in its existing form did not seem to present undue exposure.

121

Ernst & Young noted in its due diligence report that the UK operations were under the control of Michael Payne. The report contained a projection that by 1998 treaty business written by Payne would represent 75 per cent of the premium income in the United Kingdom. In fact, Payne was the only person authorised to write treaty business, and in 1995 it was already projected to account for half the business. At that time the question of a successor to Payne was already being raised, particularly considering that by 1998 he would be 71 years old. 122 It was evident that the UK operations were highly dependent (at least at that stage) on one line of business, treaty, and on one person, Payne.

16 The Winterthur years

12.11.1 Preparation of internal audit reports

During 1995 Winterthur became concerned about the expansion of the UK operations and asked for a detailed review of the Australian operations of HIH Marine and a general review of HIH's offshore operations. Waters prepared an internal audit report on the UK branch as at 31 October 1995, which reflected signs of improvement but highlighted weaknesses and deficiencies, both minor and major. He reported that there had been significant growth since the start of the UK operations in 1993, that there had been continuing improvement in technological support, and that plans had been implemented to improve accounting systems. 123

In May 1996 Winterthur began a strategic review of overseas operations and reinsurance. The result was a proposal for immediate action, primarily to bring some of the underwriting activities in London under control. 124 Senior management of Winterthur met in August 1996 and expressed concern about the UK operations and HIH's international activities in general. 125 The risk posed to Winterthur by its lack of control over HIH became increasingly pressing. 126 Concern continued to be expressed within Winterthur from 1997 onwards. 127

Winterthur began to give serious consideration to its options. It considered that moving to majority control of the HIH board would probably require the replacement of Williams and Payne, since it was thought that neither would accept a sale of the UK operations. Winterthur acknowledged, however, that replacing the two men would be very difficult and risky because of the company's perceived dependence on Williams and Payne. Winterthur concluded that selling its shareholding in HIH was the preferable option. 128

Winterthur's strategic review team prepared various internal audit reports; the reports caused disquiet in the UK branch. 129 This led to an agreement between Winterthur and HIH that the Winterthur strategic team would cease its work and Waters would undertake an internal audit review. 130

12.11.2 The River Thames and Sphere Drake proposals

Subsequent proposals by HIH to expand its UK operations further aggravated Winterthur's concerns. At an HIH board meeting in August 1996 Williams tabled a paper in relation to the proposed acquisition of the River Thames Insurance Company. 131 Williams testified before the Commission that he was aware Winterthur was opposed to the acquisition of River Thames. 132 Abbott stated that initially Williams threatened to resign and take his management team with him if the acquisition was not approved. 133 Wein acted as mediator, and it was agreed that the acquisition would be deferred while the future conduct of HIH operations in the London market was discussed with Winterthur. 134

This prompted Schurpf to schedule a meeting between Williams and various representatives of Winterthur-for 'an objective discussion on the international strategy of BIB-Winterthur'. The meeting was held in September 1996. 135 HIH's UK activities were discussed at length and Winterthur expressed its lack of

Th e failure of HIH Insurance 17

enthusiasm about the London market as a whole and its reservations about the River Thames proposal. 136 It was agreed that the River Thames proposal would not be pursued. 137 After the meeting, Schurpf sent a facsimile to Williams:

I am confident that the conclusions of that meeting on strategy, the early information of Winterthur on important strategic proposals and the measures to be taken in respect of your UK Underwriting policy form a clear and valid basis for successful cooperation. 138

At the board meeting in October 1996 Williams reported that 'rather than pursue the acquisition of River Thames (RTI), management could achieve the principal aim of the acquisition through the employment of RTI' s personnel'. He also reported that an agreement had been reached whereby the River Thames information and financial systems would be made available to HIH. 139 Williams gave evidence that the objective of the acquisition of senior River Thames management personnel was to strengthen the UK branch staff; this was an attempt to resolve some of the concerns expressed by Winterthur and identified in the internal audit reports. 140

Similar difficulties arose when HIH acquired certain lines from another UK operation called Sphere Drake. In January 1997 Heri became aware from a newspaper article of the acquisition of Sphere Drake's P&I (protection and indemnity) business. He wrote to Williams complaining that this was an inappropriate way for directors to learn of the acquisition and noting that this was not an isolated incident of inadequate communication. 141 Williams responded that the Sphere Drake matter was business arising in the normal course of the company's activities and was therefore not advised to HIH's non-executive directors, with the exception ofWein. 142

12.11.3 Expansion and financial deterioration in the UK business

In response to the Winterthur internal audit reports, Waters prepared an internal audit report that reviewed the inwards treaty reinsurance portfolio of the UK branch as at 30 September 1996. 143 The report highlighted the problems stemming from Payne's expansion into the marine excess-of-loss whole-of-account line ofbusiness (a market in which the UK branch had no experience) and, in particular, the UK branch's involvement with Charman.

144 Charman was a Lloyd's syndicate formed

by John Charman; Payne had become involved in it through the writing of the marine excess-of-loss whole-of-account lines of business.

Winterthur's concern in relation to further expansion of the UK operations grew more and more serious, particularly when Schurpf read a press report claiming that Payne expected a premium increase of 33 per cent for HIH (UK) in 1997. 145

Later

that month Williams wrote to Winterthur in an attempt to allay its fears about developments in the UK branch and the questions raised as a result of the review of the UK operation. 146

At an HIH audit committee meeting in August 1997 Schurpf expressed the view that the UK branch continued to outgrow the rest of the HIH group. He remained

18 The Winterthur years

concerned about the explosive growth, including that gained through the acquisition of the River Thames and Sphere Drake portfolios.147 Waters tabled 148 an internal audit activity report that noted 'a continuation of the problems previously identified in the HIH (UK) operations'. 149

An internal audit report prepared by Waters as at 30 September 1997 stated that conditions in the United Kingdom had further deteriorated. 150 HIH's accounting operations and financial integrity were said to be in an unacceptable state, and there had been a major deterioration in the reliability of financial results. 15 1 The report also noted there was an unusually high proportion of accrued premium to gross written premium (63 per cent) 152, which Waters said indicated a lack of control of the accounting function or of underwriting. 153 Payne considered that the report would be 'literal dynamite in the hands ofthe Swiss' .154

In December 1997 Winterthur's senior management reached the conclusion that Winterthur should sell its interest in HIH. 155 Later that month, Winterthur' s chief auditor, Eric Sieber, expressed the view that the main cause of concern to Winterthur in connection with HIH was its UK operations, although it would not be

sufficient for HIH only to sell its UK operations. 156

12.12 The US business

Williams also promoted expansion in the United States, even though he had assured Winterthur in his letter of 31 May 1995 that HIH would confine its geographic focus to Australasia. He pursued the US expansion strategy despite known problems in the Californian workers compensation industry, the absence of the usual warranties from the vendors, and Winterthur's opposition to HIH's re-entry into the US market.

HIH had sold its Californian business operations, Heath Cal (subsequently renamed), to CareAmerica in 1994. Two years after the sale, however, Sturesteps became aware that CareAmerica was considering the disposal of its workers compensation business. 157 The result was that in December 1996 Dominic Fodera circulated to the HIH directors a briefing paper prepared by Hambros. 158

The briefing paper 159 was tabled at the board meeting held in December 1996, and approval was given to Williams to enter into negotiations for the purpose of acquiring Heath Cal for US$67 million. 160 Schurpf and Heri were unable to attend the board meeting. Before the meeting Schurpf had asked Gorrie to require full due diligence into the business, including an assessment by an independent expert of likely developments in the Californian workers compensation market before HIH

made a final decision. 161

On 23 December 1996 Schurpf wrote to Williams, setting out a number of Winterthur's concerns in respect of the proposed acquisition. Winterthur was primarily concerned that HIH's confidence in the market upturn might not be justified, that CareAmerica's reserve levels appeared inadequate and that the

The failure of HIH Insurance 19

purchase of CareAmerica's workers compensation business would damage the credibility of HIH and Winterthur in the market. Winterthur was also concerned that the acquisition would be encroaching on the business of associated companies in the United States. 162

Schurpf sent a facsimile to Williams, asking that the board meeting scheduled for 27 December 1997 be postponed due to the short notice. Further, 'If you don't rearrange the date, I ask you not to vote on the topic. If you don't accede to this wish of Winterthur, I formally reject the proposal to reacquire Care America also on behalf of Erwin Heri and Randolph We in'. 163

12.12.1 The 27 December 1996 board meeting

The board meeting was duly held on 27 December 1996. The status report on the CareAmerica negotiations, Andersen's report on financial due diligence, and actuarial information were tabled. A vote to approve the acquisition was deferred until 3 January 1997. 164 According to the minutes, Williams advised that the Winterthur directors had expressed their opposition to the acquisition, but the board members present were unanimous in their support of the proposal. 165

It was resolved that Cohen would write to advise Schurpf of the outcome of the meeting. 166 He did so the same day, reiterating the support that the transaction received from other directors. He also pointed to the conflict between the interests of Winterthur and those of HIH, suggesting that Winterthur wished HIH to put the

interests of the majority shareholder before the interests of the substantial minority: he described it as 'an absolutely fundamental corporate governance issue' .167 In his letter Cohen made no mention of the 31 May 1995 letter from Williams to Spalti, which had confirmed that HIH's geographical focus would be Australasia. 168

Abbott and Gorrie sent a facsimile to Schurpf and Heri, confirming that a vote had been taken to allow more time to consider HIH' s proposed acquisition. They advised Schurpf and Heri of their strong support for the proposal and asked that Schurpf and Heri give further consideration to withdrawing their objection. They warned that there would be a conflict of interest at board level if the Winterthur directors were to vote against the proposal for reasons that were not related to the best interests of HIH. 169

12.12.2 Approval of the acquisition

At a board meeting on 3 January 1997 Schurpf reiterated Winterthur's opposition to the proposed acquisition. 170 Williams, Cassidy, Fodera and Sturesteps responded to the concerns he raised. A written record of the concerns raised was annexed to the minutes. 171

In any event, the motion in favour of the acquisition was approved, with Abbott, Schurpf and Wein abstaining from voting. Apart from Abbott, the Australian

20 The Winterthur years

directors were unanimous in their wish to reacquire Heath Cal. 172 The agreement­ entailing a purchase price of $79.3 million 173 -was signed on 31 December 1996. 174

HIH attributed Winterthur's opposition to the acquisition to a desire to prevent competition against it in the US market. Williams reiterated that view in evidence put to the Commission. It is undoubtedly the case, however, that Winterthur's representatives raised a number of matters that went directly to the financial merit of the transaction and were unrelated to any aspect of potential competition with Winterthur's American business. 175 Sturesteps testified that some of the objections raised by the Winterthur directors did in fact prove accurate.176

12.13 The Credit Suisse - Winterthur merger

In August 1997 Winterthur approved a proposal to merge with the Credit Suisse group. The proposal prompted Winterthur to re-examine its troubled relationship with HIH and resulted in Winterthur's exit from HIH.

12.13.1 HIH's reaction to the merger proposal

When he was advised of the merger proposal, Williams travelled to Switzerland and met Spa1ti. He was told that the merger would not interfere with the relationship between Winterthur and HIH. 177

Williams told the HIH board of the discussions.178 Then, on 15 August, he sent a facsimile to Schurpf, attaching to it a letter from Cohen to Brian Self of Winterthur Holdings Australia detailing the ramifications of the merger. The letter set out a number of points described as 'the negative implications for HIH Winterthur' and demanded a response by 20 August. 179 Cohen also asserted that, as a result of Credit

Suisse' s banking activities, HIH may be precluded from continuing its insurance business in the United States.

It is difficult to see any justification for the urgent and imperative tone of Cohen's letter. It made an unreasonable, inappropriate demand of the majority shareholder. HIH's attitude to the announcement of the Credit Suisse - Winterthur merger gave Winterthur another reason to seek to dispose of its shareholding in HIH.

12.13.2 Winterthur's response to HIH

Schurpf and Heri responded to Williams by facsimile. They objected to being pressured to provide an answer at extremely short notice. They also expressed astonishment at the 'alarmist attitude shown in your letter with a lot of requests to your major shareholder and few proactive and constructive proposals' .180

Williams sent Schurpf a facsimile to which was attached a letter from Cohen to Heri. In that letter Cohen confirmed that HIH saw the United States as an important strategic market and asked that Winterthur agree to compensate HIH for any loss

The failure of HIH Insurance 21

suffered through an inability to expand in the United States or an obligation to divest itself of its operations there as a result of the merger. 181

At a board meeting in late August 1997 Schurpf advised that Winterthur would not propose that HIH be forced to sell its US operations at unattractive prices and confirmed that the Winterthur - Credit Suisse merger would be very beneficial for all parties concemed. 182

12.13.3 The merger

By late 1997, 98 per cent of the Winterthur shares had been tendered pursuant to the exchange offer, the Credit Suisse merger had been approved by shareholders, and all the necessary regulatory approvals for Credit Suisse to acquire a controlling interest in HIH had been received. 183

In closing submissions to the Commission, Heri and Schurpf asserted that-without the support of the majority of the HIH board (and, in particular, Williams and Cohen) and despite their best efforts- they were not in a position to influence the board's decision making, especially in respect of major acquisitions and important strategic matters.184 This is an accurate assessment of Winterthur's predicament in circumstances in which it was not prepared to bring about changes to the composition of the HIH board.

By the end of 1997 Winterthur had reached the conclusion that it should sell its interest in HIH. 185

12.14 Winterthur's proposed disposal of its shares

Winterthur's decision to sell its interest in HIH posed a dilemma both for HIH and for Williams. On one hand, a majority shareholder with the strength and standing of Winterthur had obvious benefits for HIH, among them access to capital and expertise and the resulting reassurance it gave to lenders, investors and policyholders. On the other hand, the dilemma- for Williams in particular- was that, unlike Winterthur, a new majority shareholder would almost certainly insist on board control. Replacement of Winterthur with another majority shareholder that would tolerate HIH management's effective control of the board was most unlikely. It seems that Williams was anxious to retain his control over HIH without the

interference of a majority shareholder. Accordingly, he was opposed to the trade­ sale option favoured by Winterthur and pursued an alternative approach consistent with his desire to retain control. This ultimately led to the abandonment of the trade­ sale option and, instead, the pursuit of a book-build to accommodate Winterthur' s departure from HIH.

22 The Winterthur years

12.14.1 Evaluation of the options

Between November 1997 and January 1998 Winterthur instructed Credit Suisse First Boston to make detailed assessments of the options available to Winterthur were it to sell its shareholding in HIH.186 CSFB considered a variety of means, including selling the shareholding to institutions in blocks of 19.9 per cent or less, a trade sale to a strategic buyer, and a global offering. 187 It came to the view that the

most beneficial outcome for Winterthur, and for HIH's minority shareholders, was for Winterthur to be replaced by a strategic shareholder that already had a presence in the insurance sector. Among the benefits this option afforded were balance sheet strength, improved capital-raising prospects, share register stability, and strategic expertise supporting a positive credit-rating impact. 188 Essentially, these were the same benefits the HIH board had identified when recommending the CIC merger in

1995.

12.14.2 HIH's response to Winterthur's exit

In early 1998 the board of HIH was advised that Credit Suisse First Boston was considering ways of disposing of Winterthur's interest in HIH. 189 Williams and W ein travelled to Switzerland and met Thomas W ellauer, chief executive of Winterthur. 190 On returning to Australia Williams sent a handwritten note to Wein 191 , saying that he (Williams) was convinced that the disposal of Winterthur's

51 per cent interest in HIH should be 'Williams driven' and that ideally 'management [would take] 5 per cent (of the 51 per cent) on favourable terms provided by Winterthur'.

The note from Williams is important in several respects. Its terms are consistent with other evidence that it was Williams's view that he should be in control of the sale of the Winterthur share parcel, rather than Winterthur or its advisers. It also revealed Williams's opposition to the involvement of CSFB and indicated his general support of the sale of the Winterthur shares to a diversified group of

shareholders, rather than to a single purchaser. The terms of the note suggest that Williams was trying to take advantage of Winterthur's dependence on the cooperation of senior management during the sell-down process, to secure for himself and other senior management a benefit of the kind that had been secured at the time of the float and the CIC merger.

In closing submissions to the Commission Williams stated that the reference to 'favourable terms' in his note was a reference to a loan facility on favourable terms (so that management could buy 5 per cent of company shares), not a reference to a reduction in the price of those shares. 192 Even so, the precise nature of the benefit sought does not materially affect the tenor of Williams's attempt to secure it. Williams's request to Winterthur-even if it were for a loan facility on terms that would enable management to acquire those shares in HIH-was inappropriate. He should have been seeking to advance the interests of HIH shareholders; he should

not have been trying to secure from Winterthur a collateral financial benefit for HIH management.

The failure ofHJH Insurance 23

In a memo Williams prepared on 27 January 1998 he stated that he regarded the prospect of Winterthur divesting its shares in HIH as 'disappointing' and noted that HIH ' s rating would probably be adversely affected and that new capital in the order of $100 million would have to be raised. Williams proposed that he should manage the transaction, saying it was necessary for him to find replacement shareholders not only for the Winterthur shares but also for the capital raising. 193 He noted alternative strategies and stressed the need for the HIH board's cooperation. He further noted that it was unlikely that a single trade buyer for Winterthur's parcel of HIH shares could be located and that a buyer would probably want due diligence-which raised

'a number of serious issues'. He said he would seek to ascertain management' s intentions. 194

On 30 January 1998 Alan Highet, managing director of CSFB, sent to Wellauer a letter that stated,

The di scus sion with Hambros (Richardson called me this morning) corroborated the assessment we discussed last night, in terms of RW ' s motives and the fact that it appears that no genuine consideration has been given to strategic (as opposed to institutional placement) altematives. 195

Williams also wrote to Wein, claiming that the approach of seeking a strategic purchaser was contrary to an agreement reached with Winterthur. He expressed his concern that the decision to seek a buyer for the 51 per cent, with a view to making a full takeover offer for the company, was totally wrong and 'treacherous' to the interests of the company, its shareholders, its customers and staff. 196

On any objective assessment, a move towards a takeover offer was not likely to be adverse to the interests of HIH shareholders. Williams ' s evidence of an agreement in this regard is not accepted. There is no other evidence of such an agreement, and it seems improbable that Winterthur would have entered into an agreement so obviously contrary to its interests and the advice it had received. The contemporaneous conduct of Winterthur in seeking out the trade sale alternative is inconsistent with the idea that such an agreement had been reached.

12.15 The sale of Winterthur's interest

In response to Winterthur' s decision to evaluate strategic options for its shareholding, the HIH board appointed a steering committee consisting of Abbott, Cohen, Williams and Wein. 197

At the board meeting in February 1998 Leigh Brown of Minter Ellison suggested that a book-building exercise, on the basis of market experience, would tend to enhance the share price. It would also identify parties who were genuinely keen to take an interest in the company. The board agreed on a number of things. First, a

book-building exercise was the preferred route for the disposal of Winterthur's shareholding. Second, Hambros should be retained to act as financial adviser to the company. Third, there was no need to consider the confidentiality agreements

24 Th e Winterthur years

proposed by Winterthur since only public information would be made available under the book-building process. Fourth, action needed to be taken swiftly-to coincide with the company's profit announcement, scheduled for 4 March 1998-to resolve the matter. Finally, any exchange of information between the internal audit departments of Winterthur and HIH was to be suspended. 198

The events of that board meeting are more fully described in a letter of report from Wein to Wellauer. Wein noted that the meeting started on a 'somewhat unfriendly and emotional note' as Williams gave an account of recent events, including his trip to Switzerland. Williams also 'expressed his unhappiness about the handling of the

case and the tum of things after his return from Switzerland where he had gotten the impression to have a deal, ie to do it his way in the form of an off market placement' .199 Williams said that he and all management would refuse to cooperate with the trade-sale alternative and would not prepare information or road shows or provide any due diligence. W ein informed the Commission of his view:

The attitude of the directors was interesting in that directors generally did not support RW's approach, on the other hand clearly RW 's judgment is highly regarded and the uncompromising attitude with its veiled threats scared the Board.200

Cohen wrote to W ellauer, telling him the board was very concerned about a strategy that relied primarily on the sale of the shares in a single parcel. He said the board unanimously supported the use of a parallel approach, with a public institutional book-building that might encourage a single purchaser to make itself known. He also noted that one impediment to granting due diligence was that the company

'cannot provide information to any party which is not in the public domain' ? 01 It

appears that Williams remained fundamentally opposed to a trade sale and had little intention of encouraging a single-trade purchaser for Winterthur's stake. Williams 's response to the prospect of the sale of Winterthur's shareholding reveals his desire to ensure that he did not lose control of HIH. He was using his position as chief executive for that purpose, rather than using his power and position as director for the benefit of HIH.

The board as a whole (Cohen, Head, Payne, Stitt, Sturesteps and Williams) failed to attach adequate weight to the importance of securing a major shareholder to replace Winterthur-one with international insurance expertise and the financial strength to support any capital raising that might be necessary in the future.

12.15.1 Discussions between HIH and Winterthur

Wellauer responded to Cohen's letter after discussions with him. He confirmed the adoption of a 'multi pronged secondary offering strategy' in which a prospectus would be prepared while alternative sale options were also pursued. 202

On the same

day Williams told Winterthur's chief auditor, Sieber, that he would not be permitted to attend the audit committee meeting scheduled for 2 March 1998? 03 Wellauer advised Cohen that he was becoming 'increasingly worried about Ray Williams's

attitude towards his majority shareholder' ? 04

The failure ofHIH Insurance 25

A steering committee meeting was held on 20 March 1998 ? 05 The main point of discussion was Winterthur' s wish to carry out comprehensive due diligence. Representatives of Winterthur confirmed that a comprehensive review was necessary, to instil confidence in the institutional investors and thereby obtain a higher price for its shares. It would also reassure Winterthur that the information memorandum was complete- with no material omission.206 Minter Ellison noted that Winterthur's request went beyond what HIH was willing to accommodate and might disclose competitive information.

During March 1998 CSFB made discreet approaches to a number of large insurance companies that had been identified as most likely to be interested in acquiring Winterthur' s shareholding in HIH. A Belgian insurer, Fortis, and a US-based insurer, Liberty Mutual, were among the parties approached. 207

The HIH board subsequently agreed to work in cooperation with Winterthur in the assessment, evaluation and implementation of third parties' proposals for acquiring Winterthur' s shareholding. It also agreed that HIH would disclose information not already in the public domain to genuine third party strategic buyers, subject to certain qualifications. 208 The most important of those qualifications was that Williams first be satisfied as to the 'chemistry' of an alliance with any particular prospective party, which effectively gave Williams the power to veto any prospective bid.

12.15.2 Proposals by Fortis and Liberty Mutual

Fortis and Liberty Mutual submitted written preliminary non-binding proposals, each of which sought due diligence. 209 Wellauer subsequently spoke with Williams about the two proposals. During this discussion, Williams reiterated his opposition to Winterthur pursuing the trade-sale alternative. 210 In discussions between CSFB and Abbott, Highet reported on the principal elements of the two proposals. He testified to the Commission that he emphasised that the proposals were serious, very attractive in terms of price, supported by strong balance sheet and credit ratings, well structured, and warranted appropriate access to management and due diligence. 2 11 The share prices Fortis and Liberty Mutual had submitted were between $3.57 and $3.60.212

In late April 1998 Williams met separately with representatives of Fortis 2 13 and Liberty Mutual. 214 Subsequently, representatives of Fortis arrived in Australia to undertake due diligence but withdrew from discussions after HIH denied access to due diligence materials. 215

Abbott informed the Commission that Highet did not tell him the precise details of the price.216 Similarly, Williams denied being aware of the price and terms being offered by Fortis and Liberty Mutual; in fact, he stated that, on seeing this information when preparing to present evidence to the Commission, he felt sick and almost wept. 217 Wein confirmed, however, that he (Wein) was aware of the terms of

26 The Winterthur years

the offers and that the other Winterthur directors were also aware of the precise terms and conditions attached to the Fortis and Liberty Mutual offers.

Abbott wrote to W ellauer, saying that the provision of confidential information would be delayed until the board could be assured that a deal had been struck that had a good chance of success and was subject only to final due diligence.218

In May 1998 Williams advised the steering committee that Liberty Mutual remained interested in acquiring Winterthur's shareholding in HIH but was not interested in making a full bid. 219

But in its proposal dated 17 April 1998 Liberty Mutual advised that it would be prepared to tender for 100 per cent.220 Richardson explained that Liberty Mutual wished for a degree of control equivalent to that provided by a I 00 per cent shareholding, as opposed to a 51 per cent shareholding in a publicly

listed company. This resulted in HIH suggesting that Liberty Mutual make a takeover bid, which it declined to do. 221

Highet gave evidence that he disagreed with Richardson's assertion that Liberty Mutual wanted a degree of control equivalent to a 100 per cent shareholding. Highet's understanding was that if Liberty Mutual acquired Winterthur's 51 per cent shareholding it expected to have a level of influence commensurate with such a shareholding and that-unlike the position adopted by Winterthur-Liberty Mutual proposed to move to exercise control of the HIH board. Highet said he understood that Williams and perhaps other members of the HIH board regarded such an objective as unacceptable.222 In fact, Williams and Fodera summarily discontinued all discussions with Liberty Mutual after Liberty indicated it would like to move to

board control. In this light, it is not surprising that Liberty Mutual's proposal failed.

CSFB subsequently confirmed to Winterthur that Fortis was still interested in proceeding with the acquisition of the Winterthur shareholding if the problem of due diligence could be overcome.223 Williams informed the Commission that Fortis was simply interested in acquiring the 51 per cent shareholding of Winterthur and lacked commitment to HIH's international operations.

224 It is clear that Williams was

opposed to Fortis acquiring a majority stake in HIH.

12.15.3 The offering memorandum

HIH was anxious that only minimal information be included in the offering memorandum, whereas Winterthur sought the inclusion of as much information as possible in order to obtain a higher price. Ultimately, extensive information was included in the memorandum. In particular, the document contained actuarial assessments and a detailed claims-development table for HIH for the period 1987 to

1997. This claims-development table was the first public revelation of HIH's historical under-reserving. It showed the extent of under-reserving in the accounts of HIH over a number of years. In the offering memorandum the explanation put forward for the decline in the adequacy of claims provisions was the longer-than­ expected adverse development of certain long-tail portfolios.

225 Even though various

of the HIH directors claimed they did not see the due diligence reports of Ernst &

The failure of HIH Insurance 27

Young and Blake Dawson Waldron, they should have been aware of HIH's reserving problems by virtue of the inclusion of the claims-development table in the information memorandum.

On 13 July 1998 Wein signed the offering memorandum on behalf of Winterthur Australia Holdings. 226 The memorandum provided for instalment receipts whereby investors would pay an initial instalment of $1.60 per instalment receipt, with the price of the second instalment to be determined by an institutional book-building process.Z27

Under the agreed sale structure-which was announced to the public when the offer opened on 14 July 1998- the 169 179 709 HIH shares held by Winterthur were to be sold as instalment receipts in two tranches. The first tranche comprised 145 million instalment receipts, representing 86 per cent under the primary global offering; the second comprised 24 179 709 instalment receipts, being the remainder of Winterthur's shareholding. 228 At the HIH board meeting on 31 July 1998, CSFB advised that the response from the investment community had been favourable and the book-build would finish over-subscribed.229

12.15.4 The outcome of the public offering

As a result of its own participation in the public offering, Winterthur maintained a significant minority interest in HIH- 22 885 142 HIH instalment receipts and 13 069 356 of the HIH 1996 convertible notes. 230 Even though the sale of Winterthur' s remaining shareholding in HIH was discussed towards the end of

1999, at the time ofHIH's liquidation Winterthur still held a number ofHIH shares and convertible notes. 231 The spread of shareholdings following the global offering was diverse: the 20 largest shareholders owned less than 50 per cent of the issued capital of HIH, and no single shareholder held over I 0 per cent of the capital. 232

Williams's control over HIH was tighter than it ever had been.

4

28

CIV.OO 1.999 057.

CIV.OOI.999 160at999 195.

WITS.0155 .001 at 006.

WITS.0156.001A at 029A. CIV.001.999 057 at 999 072. - -CIV.001.999 057 at 120.

CIV.OOJ.999 057 at 999 072.

CIV.001.432 at437.

CIV.001.552 at 557.

Th e Winterthur years

10

Tll542 to Tll543; WITS.Ol56.001A at 030A. II

12

13

14

15

16

SGHA.0017.036 at 044 to 045.

SGHA.0017.036 at 038 .

See WITS.Ol53.015 at 017 (annexure B). WITS.Ol53.015 (annexure B) at 017.

SGHA.0017.036 at 086.

SGHA.0017.018. 17 BRD.001.028 at 029. 18

19

20

21

22

23

24

25

WITS .0153.015 (annexure B) at 018 to 019.

WITS .OI53 .015 (annexure B) at 019 par. 20.

WITS .OI54.001 at003 par. 8.

CIV.001.552 at 559.

ABB.0010.123 at 123.

WITS .0153.015 (annexure B) at 019 par. 23.

SGHA.0018.131.

SGHA.0017.017 . 26 ERNY.0004.126 at 128 . 27

ABB.0010.117. 28

29

ABB.0010.117; APRA.0364.0118.

SGHA.0018.036. 30 WITS .0155.001 at 014 . 31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

WITS.0153 .015 (annexure B) at 023 to 028 .

WITS.0153.163 at 165; WITS .0153 .015 at 026 (annexure B) .

ROY.0077.0017.

SBB.005.290 001.

ROY .0077.0017 at 0018.

SGHA.0018.028 .

SGHA.0018.085 .

WITS .0153 .015 (annexure B) at 031 par. 62.

WITS.0153.015 (annexure B) at 031 to 032 pars 64 and 65 .

SGHA.OO 18 .053.

ERNY.0004.147 at 148.

SGHA.0018 .061.

ROY.0077.0009 at 0011.

SBB.005 .196 001.

ROY.0077.0107. 46 ROY.0077.0026 at 0036. 47

ROY.0077.0026 at 0028.

Th e failure of HIH Insurance 29

48

ROY .0077.0026 at 0029 and 0035 . 49 ROY .0077.0026 at 0029. 50

ROY.0077.0026 at 0035 . 51 ROY.0077.0026 at 0036. 52

ROY.0077.0026 at 0062. 53 BRD.005.00l. 54

BRD.005 .015 . 55 WITS .Ol55 .001 at 012 par. 45. 56

ABB0.0002.026. 57 WITS .Ol55 .001 at 013 par. 46(c). 58

BDWC.0003.001. 59 ROY .0077.000l. 60

SUBP.0053 .001 at 028 par. 70. 61 WITS.Ol54.001 at 005 par. 21. 62

SGHA.OO 18 .00 I. 63 Tll974/24 to Tll974/49 . 64

Tl4658/6 to Tl4658/55 . 65 Tl5521 to Tl5522. 66

Tl5149/48 to Tl5150/38. 67 TII974/24 to Tll975/3. 68

BRD.005.001 at 002. 69 Tll965/3 to Tll965/50. 70

SUBP.0053 .001 at 023 to 024 pars 56 to 58 . 71 BRD.005 .001 at 002. 72

Tl1966/ 12 to Tl1966/27. 73 WITS .0156.001A at 051A. 74

SGHA.OO 18.005. 75 SGHA.0018.004; SGHA.0018.005 . 76

Tl1966/ II to Tll966/21. 77 WITS.Ol55.001 at 013 par. 46(d). 78

WITS.Ol55.001 at 013 to 014 par. 46(d). 79 Tll970/41 to Tll970/46. 80

SGHA.OO 18.004; SGHA.OO 18 .005. 81 Tll969/ 10 to Tll969/28. 82

TI4725/41 to Tl4725/50 . 83 SUBP.0053.001 at 024 par. 60. 84

CIV.OOI.999 160at999 189 . - -85 SUBP.0053.001 at 024 par. 59.

30

·'

Th e Winterthur years

86

Tl197112 to Tl1971150. 87 Tl1972/ 12 to T11972/ 16. 88

BRD.009.006 at 027. 89 CJV.001.012 at 019. 90

Tl4719/2 to Tl4719/4. 91

92

93

T14725114 to T14725/22.

See, for example, T15209/13 to Tl5210/ 10 ; WITS.0168 .00 1 at 055 par. 15 1. BRD.008.000 at 000 001. 94 AUDC.002A.OOl. 95

Tl1951/20 to Tll957 /20, Tll962/6 to Tll962/l 0. 96 BRD.009.006. 97

BRD.009.006 at 031. 98 BRD.009.001 . 99

BRD.009.001 at 002 to 003 . 100 ABB.0010.114 at 114. 101

BRD.010.000 at 000 and 000 001. 102 WITS.0097.001 at 002 par. 3. 103

WITS.0153 .015 (annexure B) at 058 to 059 par. 170. 104 WITS.0153.015 (annexure B) at 059 to 060 par. 170. 105

WITS.0153 .596 at 604. 106 WITS.0153.431 at 435. 107

WITS.O l 79.00 1 at 037 to 039 pars 181 to 184; T14703/13 to Tl4703/19. 108 WITS.0156.001A at 049A to 050A.

Tl4726toT14727;Tl510.

110

WITS.O 179.001 at 044 par. 201; Tl5212/ 17 to Tl5212/4l. 111 WITS .Ol98 .001 at 005; Tl5212/39 to T15212/4l. 112

Tl5212/28 to Tl5212/34. 113 Tl5213/7 to Tl5213/12. 114

Tl7360/3 to T17360/7. 115 Tl4706/38 to T14707/14. 116

T14364/26 to T14364/34. 117 WITS.0168.001 at 052 par. 143 and 055 to 056 pars !53 to 155 . 11 8

Tl5464/7 to T15464/32. 11 9 ROY.0077 .0001 at 0001 to 0002. 120

ROY.0077 .0001 at 0002. 12 1 WITS.Ol21.001 at par. 104. 122

ROY.0077.0026 at 0034 and 0052.

Th e failure of HIH Ins urance 31

123

WITS.0153.517 at 521; SBB.019.196_001 at 196_005; see also WITS.0153.001 at par. 129. 124 WITS.Ol53.543; WITS.0153.0 15 (annexure B) at pars 130 to 132. 125

WITS.0153 .001 at par. 26 . 126 WITS.0153.001 at par. 29 . 12 7

WITS.0153.001 at pars 26 to 29. 128 WITS .0153 .001 at pars 33 and 35. 129

WITS .0153.547; WITS.0153.015 (annexure B) at par. 134; SBB.020.586_001; ROY.0076.0192; WITS.OI65.001A at par. 346. 130 WITS .0153.589; see also WITS.0153.015 (annexure B) at par. 135. 131

BRD.017.000 at 000 003 . 132 Tl1999/8. 133

WITS.0179.001 at par. 303 . 134 WITS.0179.001 at pars 303 to 304. 135

WITS.Ol53.683. 136 WITS.0153.596 at 598 to 599; see also WITS.0153.015 (annexure B) at pars 136 to 140. 137

WITS .OI53.596 at 604; see also WITS.0153 .015 (annexure B) at pars 136 to 140. 138 WITS.0153.606 at 607; see also WITS.0153.015 (annexure B) at par. 141. 139

BRD.018.001 at 003; see also WITS.0153.015 (annexure B) at par. 153. 140 Tl1999/15toT11999/26. 141

WITS.0153 .691; WITS.0153.015 (annexure B) at par. 156. 142 WITS.0153.692; WITS.0153.015 (annexure B) at par. 157. 143

SBB.020.573 001. 144 SBB.020.573 001 at 573 010 to 573 012. - - -145

WITS.0 153.643; see also WITS.0153.015 (annexure B) at par. 144. 146 SBB.Ol9.812 001. 147

AUDC.009.001 at 004; BRD.029.122 at 125; WITS.0153.650 at 653 . See also WITS.Ol53.015 (annexure B) at par. 147 ; T9725/55 to T9726/ 14. 148 AUDC.009.001 at 003. 149

BRD.026.150 at 160; AUDC.009.091 at 101 ; WITS.Ol68.001 at pars 187 to 190. 150 SBB.021.152 001. 151

SBB.021.153_001 at 153_003; PA YN.0004.015 at 015; Tl4233/7 to Tl4234/ll. 152 SBB.021.153 001 at 153 012. - -153 T9729/9 to T9729113. 154 PA YN.0004.007; WITS.Ol65.001A at par. 354. 155 WITS.Ol53.00l at par. 38. 156 WITS.O I53.00l at par. 39. 157 WITS.Ol76.00l at par. 79. 32 The Winterthur years

158

WITS.0153.695 . 159 BRD .020.040. 160

BRD.020 .000 at 000 003. 161 WITS.0153 .643 ; WITS.0153 .001 at par. 160. 162

WITS.OI53 .730; WITS.0153 .015 (annexure B) at par. 163. 163 WITS .O 153 .732. 164

BRD.021.000 at 000 001. 165 BRD.021.000 at 000 001. 166

BRD.021.000 at 000 001. 167 WITS .0153.733. 168

T14658/57 to T14659/45 . 169 WITS .0153.736 at 737 . 170

BRD.021.000 at 000 002 . 171 BRD.021.000 at 000 005 . 172

BRD.021.000 at 000 004. 173 CIV. 001.999 160 at 999 213 . - -174

SBB .019.131 001 at 131 015 . - -175 WITS.0156 .001A at par. 10; T12140/32 to T12140/35 . 176

Tl5154/53 to T15154/56 . 177 BRD.025.000 at 000 001. 178

BRD.025.000 at 000 001. 179 WITS .0153.801. 180

WITS.0153.805 at 808. 181 WITS.0153.810 at 812. 182

BRD.026.000. 183 BRD.027.000. 184

SUBP.0088.001 at par. 144. 185 WITS.0153.001 at par. 38. 186

WITS.OI80.001 at 004 and 005. 187 WITS.0180.001 at 005 . 188

WITS.0180.001 at 005. 189 BRD.028.000 at 000 001. 190

WITS.0153.015 (annexure B) at par. 193 . 191 TEND.0033.001. 192

SUBP.0053 .001 at par. 75; WITS.0305 .001 at par. 1.7 . 193 WJTS .0153.825 at 827. 194

WITS .0153.825 at 828 to 830. 195 WINT.0001.163.

Th e failure ofHIH Insurance 33

196

WITS.0153 .832. 197 BRD.029.000 at 000 001 . 198

BRD.028.000. 199 WITS.0153 .837; WINT.0001.174. 200

WITS.Ol53 .837 at 838; WINT.0001.174 at 175. 20 1 HIH.0127.0057 at 0059. 202

WINT.0001.183 at 184. 203 WINT.0001.182. 204

WINT.0001.182. 205 DUDC.0008 .031. 206

DUDC.0008.031 at 032 . 207 WITS.0180.001 at 015 to 017 pars 65 to 69. 208

WITS.O 153 .878 at 880; BRD.031.000 at 000 _ 002. 209 WITS.0153.885; WITS.0153.885 at 896. 2 10

WITS.0180.001 at 022 par. 82. 21 1 WITS.Ol80.001 at023 par. 87 . 212

WITS.Ol80.001 at029par.114. 2 13 WITS.0156.001A at 060A. 2 14

WITS.Ol56.001A at 062A. 215 ABNA.0005 .001; WITS.0153.918; WITS.Ol80.001 at 027 . 216

WITS.Ol83 .001 at 004. 2 17 T12576/39 to T12577/40. 218

SBB.027.590 001 to 590 002 . - -2 19 WITS.0153 .921. 220

WINT.OOOI.Ol2. 22 1 WITS.0155.001 at 015 to 016 par. 52 . 222

WITS.Ol80.001 at031 par. 127. 223 WITS.Ol53.928. 224

WITS.Ol56.001Aat060A. 225 CIV.OOI.999 160 at 219. 226

CIV.OOI.999 160 at 169. 227 CIV.001.999 160 at 164. 228

CIV.OOI.999 160 at 163. 229 BRD.035.000. 230 WITS.0153 .001 at 010 pars 53 and 54; WITS.Ol53 .015 (annexure B) at par. 247. 23 1

WITS.0153 .001 at pars 55 and 56. 232 CIV.001.838 at 870.

34 Th e Winterthur years

13 Unprofitable international operations

In 1986 HIH began to expand its operations into the United Kingdom, the United States, and Hong Kong and elsewhere in Asia. The company's international operations warrant particular attention because, combined, they form the largest single source of loss in the collapse of HIH. The provisional liquidators have assessed outstanding liabilities from these operations to amount to approximately A$2.5 billion. 1

Despite the magnitude of these losses, the evidence relating to them did not reveal many instances of possible contraventions of the law by officers and employees of HIH. Rather, it was indicative of mismanagement. HIH's expansion outside Australia was the product of bad decision making and a lack of business judgment in circumstances of adverse insurance market conditions. HIH failed to respond appropriately to the losses that started to flow from its international operations and continued to write business despite these losses.

13.1 Operations in the United Kingdom

The greatest losses HIH suffered internationally are attributable to its operations in the United Kingdom. The UK business was initially managed by Michael Payne. Ray Williams had recruited Payne in 1992 to head HIH's expansion into the UK market. This expansion included a foray into areas of business in which the company had no experience. The failure of HIH to respond adequately to the problems that consequently surfaced led ultimately to its collapse.

HIH's operations in the United Kingdom encompassed the activities of its UK branch and, following the acquisition of the Cotesworth syndicates in 1998, its activities within Lloyd's.

13.1.1 Losses stemming from UK operations

As at December 2002 the estimated deficiency of HIH's UK operations on a net, undiscounted central-estimate basis was A$660 million for the UK branch of HIH Casualty and General Insurance Limited, A$1 077 million for Cotesworth Capital Limited, and A$48.5 million for FAI Underwriting Limited. 2 This amounts to more than $1.7 billion. In dollar terms, the UK operations were the largest single contributor to the collapse of HIH.

The failure of HIH Insurance 35

The nature of long-tail insurance of the type written by HIH in the United Kingdom is such that the losses were not revealed until some years after the contracts had been entered into. As a result, once those losses were apparent, the measures that could be taken to confine the financial consequences were limited. The losses occurred in two main areas : the acquisition of Cotesworth, which occurred after extensive due diligence; and the writing of one-off business, notably with the Charman syndicate and in film finance.

13.1.2 The relationship between Williams and Payne

Williams and Payne first met in 1964 during one of Payne's visits to Australia.3 Thereafter a close relationship developed between them, which led to the establishment of MW Payne Liability Agencies Pty Limited, or PLA, in Melbourne in 1968. PLA was registered at Lloyd ' s as a broker4 ; it operated on behalf of two Lloyd' s syndicates, Sir William Garthwaite & Others and Wynn Davies & Associates. 5 Williams was chief executive officer of PLA, and Payne and Williams were directors with equal shareholding.6

In 1969 the Garthwaite Managing Agency withdrew support for PLA. Payne approached the CE Heath syndicate about taking up opportunities in Australia. This culminated in the CE Heath group's acquisition of PLA in 1971. PLA became CE Heath Underwriting Agencies Pty Limited and took over the control of the underwriting. 7

13.1.3 Initial operations

In 1992 Payne ceased as an active underwriter in his Lloyd's syndicate. 8 Williams subsequently approached him to establish a branch office of HIH in the United Kingdom. Payne was appointed chief executive of HIH (UK) Limited in July 1993 9, and the UK operations started underwriting in September 1993. 10

HIH Casualty and General was one of three Australian-licensed insurers of HIH. It was also registered as a foreign company in the United Kingdom and was authorised to write insurance and reinsurance business there as a branch operation of the Australian company. It is referred to here as the 'UK branch'. HIH C&G held its

interests in the United Kingdom through a wholly owned subsidiary, HIH European Holding Company Limited, which also owned a service company, HIH (UK) Limited. 11

Payne held the position of chief executive of the UK branch until November 1997, when he became ill. Peter Thompson then took over in an acting role. 12 On 1 July 1998 Harvey Simons was appointed chief executive of the UK branch. 13

At first, the UK operations of HIH were successful; in fact, the 1994 operations were described as an 'outstanding success' .14 The initial business written consisted mainly of public liability and professional indemnity, with some inwards treaty business. 15 These lines of business were later expanded. During 1995 Payne started

36 Unprofitable international operations

writing some marine inwards treaty reinsurance business. Among it was a proportional treaty written for the Charman syndicate; this business subsequently deteriorated significantly. 16

The UK branch's involvement in the Charman syndicate is discussed in Section 13.1.5.

13.1.4 Internal audit of UK operations

The purpose of an internal audit department is to provide an objective assessment of a company's performance and the effectiveness of its systems of internal control. Internal audit review can be an effective means of management-but only if recommendations for improvement are acted on. In this way, the internal audit function can help a company avoid unjustifiable risks.

Internal audit reports: 1995 and 1996 As a result of concerns Winterthur Swiss Insurance Company expressed in 1995 about the expansion of the UK operations, Greg Waters, general manger of HIH's internal audit, prepared an internal audit report on the UK branch as at 31 October

1995. 17 The general impression of that report was 'good' with a 'few minor' and a 'few major' weaknesses and deficiencies, matters that either were isolated or were intended to assist the move of the UK branch into the future . The report failed to alleviate Winterthur's concerns and during 1996 the Winterthur internal audit team prepared various reports. 18 The criticisms expressed in those reports-and the negative reaction ofPayne19-led to Waters being asked to carry out further internal audit reviews of the UK operations?0 The first of these reports, as at 30 June 1996, revealed further deterioration in the eight months that had elapsed since the preparation of the previous report.2 1 Waters next reviewed the inwards treaty reinsurance portfolio of the UK branch as at 30 September 1996.22 The resultant report highlighted the problems stemming from Payne's expansion into the marine

excess-of-loss whole-account line of business (a market in which the UK branch had no experience) and, in particular, the UK branch's involvement with Charman. 23 Soon after that report's publication, Payne stopped writing whole-of-account reinsurance. 24

Another important criticism made by Waters concerned Payne's personal handling of the inwards treaty business. There was no reporting structure that allowed others in the organisation to know what Payne was writing and the risks being taken. 25

Payne 's response to criticisms The results of Waters' reports and the criticisms directed at Payne (in particular, his management of the UK branch) prompted Payne to write to Williams and Dominic Fodera on 24 October 1996. The letter used emotional language and contained an offer to resign. 26 Waters' findings were disclosed to the audit committee of HIH in Australia in February 1997.27

Payne responded to criticism of the lack of controls in the UK branch by acknowledging his inability to manage the business as well as carry out his

The failure of HJH insurance 37

underwriting tasks. Accordingly, on Williams's instruction, Payne recruited Simon Bird as managing director, underwriting. Bird began work on 1 March 1997.28

Waters produced an internal audit report in relation to the UK operations in May 1997; it included a catalogue of outstanding matters. 29 The review revealed some improvements but there remained a number of major concerns-including unprotected aggregation of the exposure in the marine whole-account portfolio30, significant deterioration in the Charman accoune I' adverse developments in all portfolios except property32 , and the need for 'restriction of underwriting activities

into areas with appropriate expertise'. 33

At the audit committee meeting of HIH in Australia in August 1997, during the course of which Willi Schurpf expressed his concern at the explosive growth experienced in the UK branch, Waters tabled an internal audit activity report that summarised the findings ofthe December 1996 and May 1997 internal audit reports. In it Waters noted 'a continuation of the problems previously identified in the HIH (UK) operations' .34

Concerns: 1997 A further internal audit report was prepared by Waters as at 30 September 1997.35 The general impression it conveyed was 'bad'.36 The report emphasised that accounting operations and financial integrity were in an unacceptable state and that there had been a deterioration in the reliability of financial results. 37 It was clear that management controls and systems in the United Kingdom had further deteriorated. Payne, Chris Lusty (finance director) and Robert Piper (managing director, operations) met Waters and discussed the report in mid-November 1997.38 Later on the same day Payne sent a facsimile to Thompson, in which he stated that the report would be 'literal dynamite in the hands of the Swiss'.39 This was a reference to Winterthur's repeatedly expressed concerns about the management and expansion of the UK branch. Payne described the day to Williams as a 'desperate' one. 40

During this period Andersen, HIH' s auditor, was having 'grave concerns'41 about the UK branch's ultimate loss ratios on the marine excess-of-loss account, the Charman business. 42 Payne and Williams were aware of Andersen's views in this regard. 43

Williams and Fodera instructed Thompson to ensure that the necessary procedures were implemented so that the UK branch could redress the problems identified by Waters.44 In response, Thompson wrote to Piper, Payne and Fodera on 19 November 1997, proposing an action programme to 'correct' the problems identified in the internal audit report. It had been agreed that Andersen should be appointed in that regard.45 Six days later Payne suffered a series of strokes that were to prevent him from working for three months. 46 He never returned to the full-time 'hands on' role he previously had.

The UK branch continued to expand despite Winterthur' s expressions of concern. As at 31 December 1997 gross earned premium totalled $228.2 million, an increase

38 Unprofitable international operations

of 125 per cent on the previous year. Revenue had increased from $90.4 million in 1995 to $113.5 million in 1996 and to $266.2 million in 1997. 47

At the audit committee meeting in Australia on 27 February 1998 Waters tabled an internal audit activity report. 48 It stated that the UK operations had been subject to further operational problems and detailed the findings of the September 1997

internal audit report.49

The Andersen management letter In February 1998 Andersen in the United Kingdom prepared a management letter50 in relation to the year ending 31 December 1997. This was in response to Thompson's request for a review of HIH's UK operations. The letter updated a previous draft that had been presented to the November 1997 audit committee

meeting of the UK branch. 51 The letter, dated 17 February and addressed to the board of HIH European, was forwarded by Thompson to Fodera in early March 1998. Fodera's handwritten notation on the covering memorandum recorded that it was to be taken to London for the May board meeting52 ; the letter was discussed at the audit committee meeting held in the United Kingdom during that month. 53

Andersen's management letter was critical of the UK operations and contained suggestions for improved accounting procedures, internal controls, and other aspects of the business. 54 In particular, in relation to underwriting the letter stated that the UK branch had grown rapidly since 1993 and now wrote in a wide range of classes.

It added, however, 'to the outside observer it is difficult to understand what the strategy of the branch has been and whether the appropriate level of due diligence has been performed prior to the business being underwritten'. 55

The letter also commented, in relation to organisational structure, that Payne had been away ill and noted Thompson's recent extensive direct contact with the UK branch. Andersen said it understood that there were direct reporting responsibilities to Austral ian management but that the responsibilities of each of the positions of the chair, chief executive, managing director of operations and managing director of

underwriting did not appear to be clearly defined. As a consequence, Andersen believed that the lines of responsibility had become 'blurred', making it difficult to implement change, delaying the restructure of the back office and thus effective management of the problems facing the UK branch generally.

56 The letter also

stated that the quality of management information provided by the UK branch made it difficult for the Australian management to monitor and control the UK 0

57

operatwns.

Thompson forwarded the Andersen management letter to Payne in March. Payne responded by facsimile, stating that the report was 'an absolute indictment of [his] management' and further, 'maybe I got myself too obsessed with Charman but I quite clearly lost the plot'. He foreshadowed his withdrawal from the UK branch.

58

As it happened, Thompson wanted to supplant Payne as chief executive and considered that Payne's ill-health offered an opportunity to do so. 5 9

The failure ofHJH Insurance 39

Results of the UK branch: March to July 1998 The results for the three months to 31 March 1998 disclosed that the UK branch made a core earnings loss of £1.235 million; this compares with a profit of £2.313 million in the equivalent period of 1997. The loss was attributed to a 31 per cent decrease in gross written premium. The core earnings profit forecast was also revised.60 In addition, HIH was experiencing security problems: a number of cedants of HIH were no longer prepared to accept reinsurance policies written by HIH as adequate $ecurity.61

In June 1998 Waters produced a further internal audit report on the UK branch as at 31 March 1998, primarily to establish the adequacy of any action taken on the findings detailed in his report dealing with the period to 30 September 1997. The June report recorded that significant progress had been made in all areas and that some cultural change had been effected in the UK branch office. Implementation of a strong financial reporting system remained of critical importance.62 Waters tabled the findings in an internal audit activity report at the audit committee meeting of HIH held in Australia in August 1998.63 The report's findings were also expressly mentioned at the UK audit committee meeting, and Thompson reported that the UK branch was now much more into 'self management mode'. 64

On 1 June 1998 Williams asked Waters to prepare an internal audit report on the marine division of the UK branch operations. His request had been prompted by the failure of the division's management to explain satisfactorily their business at a board meeting held in May 1998. Accordingly, Waters conducted a review and issued his report in early August 1998. 65 He concluded that the UK marine operations were at a vulnerable stage; an analysis of the balance sheet revealed the net asset position had deteriorated between 1996 and June 1998. 66 Waters commented that since 31 December 1997 the UK marine operation, as a single entity, had been technically insolvent. 67 The results of the report were also tabled by Waters at the August 1998 audit committee meeting ofHIH in Australia. 68

As a result of Waters' findings, the managing director of the marine division was instructed to report to Thompson. Early in 1999 the division's activities were integrated into the respective geographic offices and placed under the control of local executive management. The division thus ceased to be a separate entity for internal audit purposes.69

Further deterioration in the financial results of the UK operations remained a matter of concern. The results for the month of July 1998 disclosed gross written premiums of £2.93 million and a core earnings loss of £52 000, bringing cumulative gross written premium for the year to £59.29 million and core earnings to £45 000. 70

Continuing deterioration: September 1'998 to Apri/1999 To follow up the results of both the 1997 audit result and the interim review in June 1998, Waters prepared an internal audit report as at 30 September 1998. The general condition of the operations was recorded as 'satisfactory'.7 1 The report stated that the UK operations continued to undergo significant change and the deteriorating

40 Unprofitable international operations

results for 1998 continued to reflect the poor state of the market. 72 The small profit of £2.468 million made in 1997 had turned into a loss of £9.335 million by September 1998. 73

Waters tabled the results ofthat report at the February 1999 audit committee meeting of HIH held in Australia. The findings were to the effect that there had been ·ongoing and considerable efforts to rectify previously reported problems and improve the level of confidence in the activities in the United K. d 74 mg om.

But the results for the three months to 30 September 1998 disclosed a loss of £7.3 5 million, bringing the core earnings loss for the nine months of 1998 to £9.17 million. The board members acknowledged that this was disappointing. They were informed that the UK branch was actively working to 'tum the situation around'. 75

On 1 December 1998 representatives of Cotesworth and the UK branch made presentations to the HIH board in Sydney. For 1998, gross written premium for all the Cotesworth syndicates was forecast at A$326.8 million and core earnings at A$15 million. The 1998 results for the UK branch were forecast at A$170.2 million

in written premium but a core earnings loss of A$15.8 million. 76

HIH had also acquired Syndicate 1236 as a result of its acquisition of FAI, which was completed in January 1999. 77 After it acquired the Cotesworth group in December 1998, Cotesworth became responsible for managing Syndicate 1236. The syndicate was placed into run-off in September 1999, a loss having been projected for the first nine months of 1999.78

By the middle of April 1999 it was clear that there would be a loss of about £10 million. The loss in the previous 15 months had offset the underwriting profits made since the UK branch had been founded in 1993. 79 It was reported to the May 1999 board meeting in the United Kingdom that the UK results for the three months to 31 March 1999 were poor, with a draft core earnings loss of £10.8 million on gross written premiums of £76.8 million. 8° Fodera said it was clear to the board that the steps taken six months earlier had not resulted in a turnaround of the position.

81

At the same meeting it was reported that the Cotesworth group had made a core earnings loss of £1.9 million for the first quarter of 1999, even though the Cotesworth syndicates had outperformed the market by 4 per cent for the reported 1996 results. It was expected that the situation for 1997, 1998 and 1999 would be more difficult. 82

Results to June 1999 The report detailing the results for the second quarter of 1999 disclosed further deterioration and a core earnings loss of £16.9 million for the quarter. There was a year-to-date core earnings loss of £27.7 million.

83

The chief executive's report on the UK branch for June 1999 disclosed two particularly noteworthy instances of poor underwriting control in the UK branch: the Taiwanese military procurement facility and the Israeli cover for physical damage to motor vehicles.

The failure of HIH Insurance 41

The report stated that reserves on the Taiwanese military procurement facility had been increased. 84 This was cover written on the personal accident account; it gave personal accident cover to members of the Taiwanese army, whether on military duties or otherwise. A specific exclusion in respect of military personnel was in place in the underwriting guidelines in 1997 but for some inexplicable reason it had been deleted in 1998. Waters, in his internal audit report as at 30 September 1999, commented that the writing of this cover had been in breach of the underwriting guidelines. 85

The report also stated that by July 1999 the accident excess-of-loss account was revealing a substantial deterioration during 1996 and 1997. This was mainly attributable to a series of claims on a book of stop-loss insurance written in those years for physical damage to or loss of Israeli motor vehicles. 86 When the Arab­ Israeli problems escalated the number of cars stolen and destroyed increased. That cover was subsequently withdrawn. 87

The only positive aspects of the report were the continuing profits booked for the professional indemnity and liability excess-of-loss accounts for non-marine and the positive results for the hull, liability, and protection and indemnity books for marine.88 The professional indemnity and liability accounts were the original lines of business entered into by the UK branch.

The writing of the Taiwanese military cover and the Israeli motor vehicle cover is glaring evidence of the UK branch's failure to adopt adequate underwriting guidelines and controls that would limit underwriting activities to less risky lines of business. The branch's involvement in the Charman and film finance lines of business (see Sections 13.1.5 and 13 .1.9) further demonstrates this shortcoming. All four instances show the UK branch's willingness to underwrite lines of business in which it had no, or insufficient, experience. The combination of a lack of underwriting controls and of relevant experience in those particular lines of business proved to be a formula for financial disaster.

The bottom-line loss for the 18 months to 30 June 1999 was £42.344 million before Charman losses and reinsurance recoveries. 89 Consideration was given to closing the UK branch, but the market was rising so a decision was deferred. 90

The UK branch run-off The results for the three months to 30 September 1999 were also poor, disclosing a core earnings loss of £8.5 million. At the board meeting held on 23 November 1999 Gary Ankcorn, the UK branch's internal actuary, expressed the view that in subsequent years there would probably be further deterioration in the accounts that had formed the key components of that loss . Williams was recorded as expressing the opinion that the results for the UK branch were extremely disappointing. 9 1

Thereafter the UK branch was placed into run-off; it had proved to be a ' financial basket case' .92 Waters produced two further internal audit reports in December 1999, one of which was a report in relation to the UK branch as at 30 September 1999.93 That report highlighted the 'extraordinary turnaround'

94

in the net asset

42 Unprofitable international operations

position-from £290 644 at 31 December 1998 to a negative £40.514 million only six months later. 95 The report was tabled at the February 2000 audit committee meeting of HIH in Australia. It highlighted the deteriorating financial performance, the acceptance of cover outside authority and the underwriting guidelines, and the

inadequate management of a binding authority given to an underwriting agency. 96

The other internal audit report prepared by Waters in December 1999 concerned Cotesworth and was critical of the performance of its compliance and internal audit group. 97

By the end of February 2000 Cotesworth had recorded a loss of £17.6 million, representing an increase of £6.6 million on the loss reported at December 1999.98

HIH's presentation to APRA In early March 2000 Williams and Fodera made a presentation to the Australian Prudential Regulation Authority, expressing optimism about the UK operations. In the presentation it was asserted the UK market was at the bottom of the cycle and some recovery was evident. 99 Fodera admitted to the Commission that the expression of optimism was not based on historical results.100 In fact, at the November 1999 board meeting held in the UK, Piper had disclosed that the results

for the three months to 30 September 1999 were poor, revealing a core earnings loss of £8.5 million . It was at that meeting that Williams expressed his extreme disappointment. 101

From a historical perspective, the reports emanating from the United Kingdom provided no basis for the optimism Williams and Fodera expressed to APRA in March 2000. They should not have conveyed to APRA an expectation of positive results for the UK operations.

And the poor results continued. The results to the end of February 2000, revealing a loss of £17.6 million for Cotesworth, were communicated to Williams about two weeks after the presentation to APRA. 102 By 31 March 2000 the core earnings loss for the first quarter was assessed as likely to be approximately £18.3 million; the core earnings loss for the nine months to 31 March 2000 was assessed to be £35.8 million. Cash flow shortages were being experienced and cash flow analyses and projections were being reported monthly to head office. At that time it was expected the shortfall would be £3.3 million. 103

Given the continuing losses and the deterioration in the position of the UK operations, Williams and Fodera should have taken steps to inform APRA of the true position and to withdraw the expressions of optimism that had been made at the March 2000 presentation.

The 30 June 2000 results The second-quarter results for the UK branch disclosed a core earnings loss for the three months of £8.6 million and a year-to-date core earnings loss of £45 million. 104

The failure of HIH Insurance 43

As a result of the deteriorating situation, Piper and his team conducted a detailed review of the treaty business underwritten by Bird. That review showed that the underwriting of the treaty account had been unsatisfactory in many respects. 105

The underwriting results of the UK branch and Cotesworth for the 12 months to 30 June 2000 revealed losses of $119 million and $44 million respectively. That result was communicated to the September 2000 audit committee meeting of HIH in Australia. 106 The board's resultant concern led to arrangements being made for Norman Britten, managing director of Cotesworth & Co. Limited, to travel to Australia and address the board. 107

Developments in late 2000 On 19 October 2000 HIH gave another presentation to APRA, showing a core earnings loss of $39.9 million for Cotesworth as at June 2000; this compares with the previous year's core earnings profit of $1.7 million. HIH forecast a core earnings profit of $13.9 million as at June 2001. It was also noted in the presentation that the Lloyd's market had experienced the worst trading conditions in 10 years and that HIH expected it to recover. Melodramatically, the presentation stated,

'There is blood on the floor, blood on the walls, blood on the ceiling'. 108

A matter of weeks later, on 2 November 2000, Fodera advised the board that on present projections Cotesworth would have to be sold in two years' time; otherwise, it would require substantial amounts of cash, preventing that cash being used elsewhere in the group. 109 Britten attended the board meeting of 29 November 2000, at which the position of the UK operations was again discussed. It was agreed that the introduction of a joint-venture partner would be welcome, that the result of the 2000 accounts would be known by January 2001 and that, subject to extension of

relevant letters of credit, Cotesworth's operations should be supported into 2001. 110

After the board meeting Britten proceeded to canvass the London market for new capital. 111 Randolph Wein suggested that they needed a person with expertise in Lloyd's to handle the capital raising or sale and, at the suggestion of Charles Abbott, Patrick Moore was appointed.

11 2 Moore had only provided preliminary

advice by the time HIH was placed into 1iquidation.11 3

On Wein's instruction, Williams visited the UK operations in early 2001, in particular to discuss the results for the year ending 31 December 2000. 114 In March 2001 Robert Stitt met with Piper at HIH's offices in London and obtained detailed information of the position of the UK operations. Piper reiterated to Stitt the firm belief in the London market that insurance rates would harden dramatically and that many classes of business would return to profitability in the short to medium term.115

Conclusions The chronological narrative just detailed reveals a series of milestones in the period 1997 to 200 I. From time to time the directors of HIH, particularly those present at the audit committee meetings, were told of the deterioration in HIH's UK

44 Unprofitable international operations

operations. Twice a year Waters tabled an internal audit activity report at the audit committee meetings held in Australia. The reports conveyed the findings of the internal audit during the preceding period. Communication to the audit committee of the continuing deterioration in the results of the UK operations should have prompted the directors to ensure that the systems established to resolve the problems

leading to the deterioration were in fact effective. Each account of deterioration constituted additional notice that the systems were inadequate. The directors' failure to respond appropriately and effectively to the recurrent advice of increasing losses contributed to the continuing financial deterioration of the UK operations.

Counsel assisting the Commission submitted that the directors of HIH did not react appropriately to the reports of deterioration in the UK operations. The directors responded, describing the initiatives taken to allay those concerns-the appointment of Bird and later Simons; the obtaining of personnel from River Thames Insurance Company, Sphere Drake Insurance plc and through the Cotesworth acquisition; and the introduction of Thompson to the UK operations. The directors argued that these

initiatives were appropriate to ensure that there were adequate controls over the UK operations.

In my view, the fact that management of the UK branch was ineffective and inappropriate and that underwriting decisions were made in breach of guidelines does not necessarily lead to the conclusion that there was a failing on the part of the non-executive directors of HIH. From the directors' perspective, I agree that each of the initiatives outlined above could reasonably have been expected to have increased the level of control over the business written in the UK branch.

However, in comparison with the non-executive directors who were aware of the results stemming from the UK operations, there was a greater onus on Williams, Payne, Fodera and George Sturesteps to respond appropriately to the continuing reports of deterioration in the financial results. They had additional knowledge of

matters pertaining to the UK operations. Unlike the non-executive directors, they received the internal audit reports and other correspondence on matters relating to the UK operations. They attended various of the board and audit committee meetings in the United Kingdom and were more generally aware of the status of the

UK operations than the other directors. In these circumstances, their failure to respond adequately was a serious departure from what is appropriate conduct on the part of executive directors.

The failure ofHIH insurance 45

13.1.5 Charman

Charman, a line of business written by the UK branch, proved particularly problematic. 11 6 Payne considered that the UK business expanded 'slowly but

successfully' until 1996, at which time it branched out into other forms of insurance business. 117 During 1995

118 and 1996 119 Payne wrote marine excess-of-loss contracts for John Charman, an active underwriter for Syndicate 488 at Lloyd' s. He knew Charman as an eminent and highly respected Lloyd ' s underwriter, who produced consistent profits for his names (an underwriting member of Lloyd's), of which Payne was one until 2001. 120 Significant losses resulted from the lines written by Payne, however, and litigation was initiated.12 1 Thompson was gi ven authority to manage the commutation or settlement discussions and the litigation.

The shortcomings raised in Waters ' internal audit report of October 1996 related almost exclusively to business conducted with the Charman syndicate. In the light of that criticism, Payne acknowledged that the involvement with the Charman syndicate was out of proportion to the whole account and that he had been aware of the problem from the beginning and took full responsibility. 122 Steps had been taken to reduce HIH ' s involvement in the Charman account by up to 50 per cent. Nevertheless, Payne remained confident that the Charman account would make a

substantial overall profit. 123

Six months later, however, in his internal audit report of 23 May 1997 Waters reported that the Charman account had deteriorated significantly. It was clear that management's response, after the October 1996 report, had been 'overly optimistic and unsustainable' . 124

Coopers & Lybrand was appointed to prepare an actuarial review of Charman covers by mid-1997. It assessed the ultimate claims in respect of the 1996 underwriting year as at 31 March 1997 and suggested a required reserve of $20.05 million net of reinsurance. 125 That figure was updated six months later to a required reserve net of reinsurance and before whole-of-account protections of $14.9 million as at 30 September 1997.

126

During the fourth quarter of 1997 the loss position deteriorated dramatically. 127 At the audit committee meeting held in February 1998 it was reported that, as at 31 December 1997, $20 million had been booked in relation to Charman for the 1996 underwriting year, in line with the likely estimate, with the best and worst estimates put at $ 15 million and $30 million respectively. 128 Six months later, at 30 June 1998, the exposure to Charman (1996) had increased by a further $5 million for the likely, best and worst estimates. This meant a booked amount of $25 million. 129

At that time the UK branch was looking· to commute as much of the 1996 Charman layers as possible and to avoid litigation.130 As at 30 June 1999 gross losses of £13.022 million were booked in the UK in relation to Charman. 13 1

The first aspect of the problems arising from the UK branch's reinsurance of the Charman syndicates was direct reinsurance; the dispute in this connection was

46 Unprofitable international operations

settled through mediation in Sydney in 1999 on confidential terms. The second aspect involved reinsurance taken out by the Charman syndicates with AXA Reinsurance UK pic, which in tum retro-ceded to HIH; disputes about the 1996 cover were settled on a confidential basis. The third aspect of HIH's exposure, which involved litigation in relation to the 1997 treaty layers, resulted in HIH being

liable for a combined amount ofUS$11 million on the 1997 year. 132

In December 2001 the liquidators estimated HIH's potential liability to be in the order of A$20 to 25 million. 133 The total loss that HIH C&G would incur as a result of providing reinsurance to the Charman syndicate, both directly and indirectly, was £17.371 million. 134

13.1.6 River Thames

HIH's recruitment of personnel from River Thames was a further attempt to resolve the problems Waters identified in his internal audit reports-in particular, the concern that there was a need for additional management in order to relieve Payne of his various underwriting duties.

HIH twice attempted to acquire River Thames, in late 1996 and again in mid-2000, but was unsuccessful on both occasions. 135

At the 30 August 1996 board meeting, Williams tabled a paper136 discussing the proposed acquisition of River Thames. The paper, which supported the continuing expansion of HIH UK's operations, had been provided to some of the directors a few days previously. 137 At the board meeting Schurpf argued that more time was needed to give adequate consideration to the proposal. Even so, it was resolved that the acquisition of River Thames would proceed, subject to due diligence, and that,

in light of the views expressed by Schurpf and if he requested it, the proposal would be further considered at a future board meeting. 138

Winterthur remained concerned about further expansion in the United Kingdom. Accordingly, Schurpf sent a facsimile to Williams proposing a meeting to discuss HIH's international strategy. 139 That meeting, attended by Peter Spalti, Erwin Heri, Schurpf, Williams and Wein, was held on 6 September 1996 in Switzerland, and

Schurpf again expressed reservations about the acquisition of River Thames. It was agreed to 'shelve' the proposed acquisition, although some management staff could be recruited to resolve the then current succession problems posed by the focus on Payne in the UK branch. 140

Williams subsequently reported to the 25 October 1996 board meeting that, rather than pursue the acquisition of River Thames, the principal aim of the acquisition could be achieved through the employment of River Thames personnel. 141

River

Thames's information and financial systems were also to be made available to HIH, with implementation scheduled for July 1997. 142

Th e failure af H!H Insurance 47

In fact, HIH also acquired the property lines of business from River Thames at no cost; the people who transferred from River Thames to HIH simply began underwriting the business they had previously written at River Thames. 143 In November 1996 it was already forecast that the book acquired from River Thames would be the growth area in 1997. 144

At first the property lines of business performed well, accounting for a large proportion of the £39.5 million increase in gross written premium for the UK branch in 1997. 145

But they then began to deteriorate, and by 1998 and 1999 significant losses were being incurred. 146 The property direct and facultative account was in run-off by the first quarter of 1999, for which the UK branch booked a preliminary loss of £3.1 million as at 31 March 1999.147 By July 1999 it was clear that 1998 had been one of the worst years on record for the property catastrophe account: a further deterioration of £800 000 was booked in the second quarter of 1999 and reinsurance purchasing had proved ineffective. 148 The property risk account had also been very

unprofitable during part of 1997 and in 1998 and 1999. 149

13.1.7 Sphere Drake

Various personnel from the marine team at Sphere Drake were also recruited in an attempt to redress the shortcomings in marine underwriting at HIH.

By late January 1997 HIH had concluded a deal in principle whereby it took over Sphere Drake's marine team of eight to ten people. 150 The following month Williams sent a briefing paper to Schurpf in relation to Sphere Drake. 151 Soon after, Heri sent a facsimile to Williams, complaining of the lack of communication about the acquisition; he had learnt of it through newspaper reports. 15 2

Williams denied there had been an acquisition, on the basis that HIH had not actually bought Sphere Drake. 153 HIH had, however, acquired Sphere Drake's marine line of business together with its staff. HIH was paid to administer Sphere Drake's run-off but did not accept financial responsibility for its liabilities until that business was renewed in HIH's name.154

13.1.8 Payne's resignation

Payne told the Commission he profoundly regretted the extent of his involvement in the Charman syndicate and believed the stress arising from that was a major contributor to his stroke in November 1997. 155

When Payne became ill Thompson

took over the running of the UK branch until Simons became chief executive on 1 July 1998. By the time Payne returned to work, in March 1998, the Andersen management letter of 17 February 1998, which was highly critical of the UK operations, had been issued. It appeared to Payne that he could not remain chief executive of the UK branch; he believed his age and state of health would not allow him to continue at the same pace. 156 Payne and Thompson discussed the matter and it was decided that Payne should stand down, with immediate effect.

157 Thompson

advised Williams, and by March Williams had agreed on that course and a search

48 Unprofitable international operations

had begun for Payne's successor. 158 Payne took on the role of non-executive chairman from July 1998.159

Payne was due to retire from HIH at the November 1999 annual general meeting, (at which time he was 73 years old) but a resolution was passed at the meeting to extend his term as director. 160 Payne said it had been agreed with Williams that he would remain for one more year. Accordingly, his resignation was accepted at the board meeting due to be held on 8 September 2000 but adjourned to

12 September 16 1 ; his notice of resignation was dated 12 September 2000. 162

13.1.9 Film finance

Film finance is a peculiar species of insurance. Financiers lend money to film producers on the condition that insurance is obtained to cover the risk that revenues from the completed films might not be sufficient to cover the costs of production and to enable the financiers to be repaid. 163 It is pecuniary loss indemnity insurance

similar to mortgage indemnity cover.

HIH usually wrote this business by fronting for reinsurers who were themselves unable or unwilling to write the business directly. Often the reinsurer would accept the entirety of the risk, a practice known as ' fronting '.

The UK branch Film finance was another line of business written by the UK branch that proved particularly problematic. 164 It was very unsuccessful from the point of view of the insurers; most insurers and reinsurers were seeking to avoid liability on the basis of misrepresentation and other grounds. 165 The UK branch wrote film finance business for a comparatively short period, from 1997 to July 1998.

Before the UK branch stopped paying claims and started litigation, HIH paid out US$57 million in claims. 166 At the time of appointment of the provisional liquidators, HIH had started litigation against reinsurers to recover

US$37.1 million. 167 As at December 2002, however, all the reinsurers had been successful in having the reinsurance claims struck out. 168

The remaining film finance claims were all subject to litigation. At December 2002 the total remaining claims against HIH in the United Kingdom, as estimated by the KPMG actuaries, amounted to approximately A$179 million gross, or A$40 million after reinsurance recoveries. 169

Steve Mitchell, the contingency underwriter for the UK branch, wrote film finance business 170, at first without the knowledge ofthe UK branch's management. 171 By at least November 1997 Payne had become aware that film finance business had been written 172, although in closing submissions he informed the Commission that he was not aware of the fronting element of the contracts until the claims started to come

. 173

111.

Th e failure ofHIH In surance 49

Bird conducted a review of the contingency account, which included film finance, and reported to the November 1997 board meeting in the United Kingdom that the core contingency business was profitable, albeit in need of better financial management. The account continued to be written. 174 It subsequently became clear that Bird had not properly assessed the risk. 1 75

By late May 1998 the directors were expressing to Bird their concern about the contingency account. 176 It had become apparent that Mitchell had agreed to front certain risks on behalf of AXA, which thereby became the reinsurer of HIH. Fronting was prohibited without the express authority of HIH group executive

management. 177 In July 1998, after Simons became chief executive of the UK branch, the contingency account was reviewed and the decision was made to close it.1 78

Between May and July 1999 claims began to flow from the film finance business that had been written. 179 At a board meeting held in the United Kingdom on 23 November 1999, it was reported that total exposure to the film finance policies was in the order ofUS$124 million gross. 180 By March 2000 Thompson was heavily involved in recovering money from reinsurers to cover film finance losses, litigation against AXA was under way, and proposals were in progress for the commutation of contracts. 181

What happened in relation to film finance is indicative of serious systemic deficiencies. There were inadequate controls in the UK branch to guard against or detect poor underwriting. To the extent that controls did exist, they were ineffective, and film finance is a prime example of that inadequacy. At least until November

1997 it was Payne's responsibility to ensure that such controls-in the form of detailed underwriting guidelines and an adequate reporting structure-were in place in the UK branch. If they were, they were inadequate. Payne should have monitored the writing of this business to ensure that he was aware of the level of risk being assumed.

Adelaide Although it is a departure from the theme of international operations, it is convenient here to consider film finance business written out of Adelaide.

From late 1996 or early 1997 film finance insurance was written by the contingency division of HIH, which was based in South Australia. 182 At this time Sturesteps was chairman of the contingency division and Ian Small was managing director. The divisional company was HIH (Contingencies) Pty Ltd.

The contingency division was a specialised underwriting unit that focused on agricultural-type insurance risks such as bloodstock, other livestock and aquaculture. 183 It also wrote staff household and motor insurances. In late 1996 the division became involved in writing film finance insurance. Small explained to the Commission the process by which this risk was written:

50 Unprofitable international operations

In late 1996 or early 1997 I became aware that a decision had been taken by HIH senior management that C.E. Heath Casualty and General Insurance Ltd had agreed to a fronting arrangement with REAC [Reinsurance Australia Corporation Limited] for film finance cover and the Contingencies Division had been chosen to facilitate this by attending to appropriate documentation. For this, Contingencies Division would receive the exchange commission earned by C.E. Heath Casualty and General Insurance Ltd.

I believe that the decision to enter into this arrangement was taken by both George Sturesteps and Peter Thompson. I was not involved in this decision .

I understood the arrangement was that ... REAC would provide 100% reinsurance cover in respect of lines of cover that HIH would 'notionally' underwrite. I say 'notionally' because HIH would have no exposure to this film finance cover because it was fronting REAC and the cover was 1 00% reinsured.

HIH would not receive any of the premium which was paid directly to REAC by the broker, but would receive a fronting commission on the amount of premium paid. This was the benefit to HIH of the arrangement ... 184

According to Small, an insurance broker in London would send to HIH reinsurance slips that had already been signed by Reinsurance Australia Corporation Limited (REAC). This indicated that reinsurance cover was already operative. Small would then sign the slip, which also indicated the level of coverage by HIH, and forward this to the broker. 185 Small explained that his main concern was to ensure that, before he committed HIH to this film finance cover, the relevant documentation had

first been signed by the reinsurer, REAC. 186 The first time Small was provided with a reinsurance slip signed by REAC, in late 1996, he approached Sturesteps. The latter advised him that he had spoken to Thompson and that it was in order for the contingency division to write this business. 187

Sturesteps had originally been approached by Roger Basset, whom he had known since the early 1990s and who asked whether HIH would be prepared to front for REAC. 188 Sturesteps agreed because he considered that REAC was an acceptable risk and because he had a high opinion of the REAC chief executive. 189

Sturesteps testified that if Small had refused to write the business on the basis that it was a fronting arrangement he would have accepted the refusal. He did, however, concede that he had 'encouraged' Small to agree to the fronting arrangement. 190

The benefit to the contingency division from the fronting arrangement was that it would receive an exchange commission. Documents available to the Commission suggested that the exchange commission was in the order of 5 per cent of the premium paid to REAC by the broker. 19 1

All up, HIH fronted for REAC and its subsidiary Monde Re on seven slates 192 , for a

total of 29 films. On a further three films, HIH was replaced as underwriter by Monde Re midway through the slate. 193

The failure ofHIH Insurance 51

In relation to Monde Re, on 6 October 1 997 Sturesteps and Small received from Rupert Dixon of Heath North America and Special Risks Limited a facsimile that read in part 'Monde Re is the wholly owned subsidiary of REAC, who you already allow fronting for on our film business and we would be interested to learn if this would extend to Monde Re who have offered us extra capacity'.194 Dixon also enclosed the latest financial report of Monde Re.

In oral evidence Small recalled having received this facsimile from Dixon and asking Sturesteps whether Monde Re was an acceptable reinsurer. 195 Sturesteps confirmed that Monde Re was acceptable. 196

Non-compliance with the HIH underwriting guidelines Paragraph 4.4 of the 1996 underwriting guidelines was in the following terms:

Risk Fronting without warranties from the ultimate insurer-this practice is not permitted because of the Group's exposure to losses in the event of reinsurer disputes or insolvency. Proposals where the ultimate insurer has agreed to a clause are to be referred to the Group Deputy

Managing Directors. 1 7 [emphasis added]

Paragraph 3 of the guidelines limited the territorial scope to Australia, New Zealand, South East Asia and the Pacific Islands. 198 In relation to international business, the guidelines permitted underwriting only on the specific authority of the group managing director (international). 199 In 1996 this position was held by Sturesteps.

Small testified that he was aware of the underwriting guidelines that applied to the contingency division in 1996 and 1997.200 He said that, because Sture steps was at a higher level than the group deputy managing directors, he had assumed that approval had been given to permit fronting? 01 He also said he did not know whether any warranty had been given by the ultimate insurer202 and conceded that, if it had not been provided, the fronting arrangement would have been in breach of the guidelines. 203

Small said that, because senior management of HIH had indicated to him that they wanted the contingency division to front for REAC in respect of this line of business, he did not think it was necessary even to consider whether fronting was in breach of the guidelines or whether the division was acting in breach of the territorial limits. 204 He also expressed the view that had HIH been writing the business itself, without any fronting arrangement, he would not have accepted the risk. 205 Moreover, he admitted, in both oral evidence206 and his written submission207, that in giving effect to the fronting arrangement he had regularly signed documents of which he had no real understanding.

Sturesteps agreed that the underwriting guidelines included a strict prohibition on fronting. 208 He considered, however, that there were exceptions to the general rule209 and that this was a matter where he could exercise discretion. 210 This was the case even though the arrangement was in breach of the underwriting guidelines and HIH had undertaken no independent risk analysis. 211

52 Unprofitable international operations

In closing submissions Small argued that he did not act in contravention of the underwriting guidelines, since Sturesteps had 'exercised the authority given to him to extend those guidelines' .212

Lack of independent risk assessment by HIH A central feature of the film finance insurance written in Australia was the complete absence of any independent risk analysis, despite the fact that the line of business was relatively novel and HIH had no experience with it. 213 In addition, the fronting arrangement meant that HIH was required to rely on the expertise of REAC for a proper assessment of the risks of the film finance cover.

In closing submissions Small conceded that he did not have the 'authority, facilities, capacity, knowledge, or requirement ... ' to carry out an independent risk analysis? 14 He argued, however, that, since film finance insurance was underwritten by HIH in the United Kingdom and the risk had been assessed by both the HIH (UK) underwriter (along with other insurers who were also carrying some risk) and REAC, it was not necessary for him to do any independent risk analysis. 215

HIH received reports that had been prepared by risk managers ICE Media Limited. The reports were provided to the underwriters and reinsurers and contained an analysis of the risks associated with a particular film or slate of films. Normally, the broker would retain the risk managers to prepare the report.

On some occasions the borrower was a company named ICE Insured Finance Limited? 16 When he was presenting evidence Small's attention was drawn to the possibility that a conflict might have existed in that the risk manager and the borrower might have been related entities: Small agreed that ' ... from time to time I did wonder about some of the documents, but they seemed to be acceptable to REAC and everybody else. All the other reinsurers had signed, so I just accepted it must have been okay' .217 Similarly, he submitted that he 'did give consideration to the possibility [that the risk manager and borrower might be related entities] but in

all the circumstances' concluded that ' ... it was reasonable for me to consider that action on my part was not warranted' .218 When asked about the same issue, Sturesteps said he could not recall whether he had considered the possibility of a conflict. 219

In oral evidence Thompson explained that assessing risk for film finance business was very difficult:

It's very, very difficult to--if you're going to write film finance business, you've got to understand not just about the film industry, but banking and how financing for films takes place. So it is a complex class . .. If you don't understand the class of insurance you're writing, you are vulnerable to making a loss. How can you assess the risk? 220

The other obvious risk that arises in respect of a fronting arrangement is that the insurer is protected only to the extent that the reinsurer is able to meet any claims that arise. Small stated that he ' ... was comforted by the fact that the reinsurance was placed with a reinsurer who I understood to be approved. By approved I mean

The failure ofHIH Insurance 53

licensed by the Regulatory Authority and approved by the HIH remsurance committee'.221

As at 26 November 1998 REAC was included on the HIH list of acceptable reinsurers. 222 This is the earliest such list available to the Commission. Monde Re was not included on this or any subsequent lists. 223

Quantification of potential exposure On 15 March 2001 Rick Saywell (run-off manager, risk management and reinsuranc·e) sent a memorandum to Thompson following a review of the film finance files. 224 He also prepared a spreadsheet that showed the potential liability to HIH as at 15 March 2001, assuming nil reinsurance recoveries, to be approximately US$23 .17 million. 225

A more recent spreadsheet prepared by Saywell-with assistance from Antonino Isgro (property and special risk)--calculated the potential liability to HIH to be US$29.84 million. 226 HIH's share of known claims was calculated to be about US$17.3 million. 227 A further schedule prepared as at 4 February 2002 estimated that, in relation to the Australian business, the potential exposure to HIH from the film finance business fronted for REAC and Monde Re had increased to approximately US$40 million. 228

Conclusions I find that Sturesteps permitted film finance business to be written by HIH C&G by way of a fronting arrangement with REAC and Monde Re. This was in breach of the underwriting guidelines. In doing this, Sturesteps did not seek or obtain the approval of any other member of senior management. His failure to do so is more significant when one considers that this was an area in which he had inadequate experience. The failure is compounded . by the absence of any-or any adequate-risk assessment in relation to the arrangements. In my view, the failures of Sturesteps in these respects constitute an undesirable corporate governance practice.

13.1.10 Lureco and the UK branch

Luxembourg European Reinsurance SA (Lureco) was a financial reinsurer that issued reinsurance policies to the UK branch over a number ofyears.229 The Lureco policies were back-to-hack policies with a fellow subsidiary of the HIH group, CIC Insurance Limited, on exactly the same terms, in a 'mirror image deal'.

230 The UK

branch of HIH C&G bought reinsurance from Lureco and Lureco bought exactly the same reinsurance from CIC. 231

Lureco wrote two policies to cover HIH's UK branch. The first policy, written in 1996 and amended in 1997, covered the 1993, 1994 and 1995 underwriting years; the second, written in 1997, covered the l997 underwriting year. 232 The amendment to the 1996 policy was made to protect part of the UK branch's exposure to Charman. At that time it had become clear that certain of the excess-of-loss business written in the United Kingdom in respect of the Charman syndicates would inflict

54 Unprofitable international operations

serious damage on the 1996 underwriting results for the UK branch's treaty results. 233

In relation to the 1993, 1994 and 1995 underwriting years, the arrangement had the effect of allowing HIH C&G to report lower taxable income in the United Kingdom and CIC to report higher income in Australia. In later years the arrangement had the effect of allowing HIH C&G to improve its balance sheet solvency for regulatory purposes in the United Kingdom to the extent of £30 million. The overall effect was to decrease the UK branch's net liabilities by £30 million and to increase CIC's net

liabilities by £30 million. 234

The reason for involving Lureco as an independent reinsurer was that related-party reinsurance would not qualify as an asset for the UK branch for the purposes of its solvency margin requirements with the Financial Services Authority (FSA). 235

By late 2000 the UK branch was corresponding with HIH in Australia in relation to the future of the Lureco policies and the possibility of commutation.Z36 Claims under both the 1996 and 1997 policies were submitted on 13 August 2002, for £9.6 million and £20 million respectively.237

The collapse of the HIH group means that CIC will not pay all claims in full and Lureco will be required to pay approximately £30 million to the UK branch. Thomas Riddell, one of the UK liquidators, reported, however, that by December 2002 Lureco had disputed liability, and arbitration is currently under way. If Lureco

is successful, the net deficiency assets of the UK branch will have to be increased by a further £30 million. 238

In summary, the substance of the transaction was that risk was transferred from the UK branch to CIC. Had that been done openly, without an intermediary, the Lureco reinsurance policy could not have been counted as an eligible asset in the UK regulator's assessment of the solvency of the UK branch. The risks associated with such an arrangement were borne out when Lureco disputed liability in the manner just described. This arrangement with Lureco also led to a reluctance on the part of

the staff of the UK branch to make a claim on Lureco in case Lureco's identical claim on CIC could not be paid by that company.239

13.1.11 The acquisition of Cotesworth

In late 1997 Williams wanted to gain the capacity to underwrite in Lloyd's, which meant that he needed to acquire or establish a Lloyd's managing agency. This culminated in the acquisition of the Cotesworth group during 1998. The Cotesworth group was made up of an established managing agency, Cotesworth & Co., which

managed four Lloyd's syndicates-228, 535, 536 and 1069. Cotesworth Capital was a corporate vehicle writing at Lloyd's exclusively through Cotesworth & Co. syndicates. The acquisition would provide direct access to the Lloyd's market and other international markets where Lloyd' s was licensed to underwrite. It would also provide scope for HIH to consolidate its two other businesses in the United Kingdom, Sphere Drake and the UK branch.Z4° From January 1999 all new business

The failure ofHIH Insurance 55

previously written by the UK branch was transferred to Syndicate 1688, which was subsequently formed under the Cotesworth umbrella. 241

Williams foreshadowed the establishment of a Lloyd's syndicate at the November 1997 board meeting. 242 Fodera and Thompson were given responsibility for the acquisition

243 and, after discussions with various brokers, DJL Phoenix Securities Limited recommended that HIH consider buying Cotesworth & Co. 244 In early March 1998 Thompson held discussions with Phoenix and the Cotesworth group. 245 HIH subsequently retained Phoenix to act as financial adviser in relation to the possible acquisition of Cotesworth246; in this capacity Phoenix carried out an independent valuation ofCotesworth before due diligence was commenced.247

At the board meeting held on 13 March 1998 it was reported that four Lloyd's corporate syndicates had been identified as likely targets. HIH was, however, still considering the feasibility of establishing its own corporate syndicate. 248

Negotiations and due diligence Soon after, Phoenix provided HIH with draft heads of agreement, and on 26 March 1998 it provided Fodera with a valuation paper on the proposed acquisition of Cotesworth, recommending that there be a due diligence review of the group. 249 A few days later, representatives of Phoenix, Cotesworth, Cotesworth's financial advisers and HIH met and discussed the conditional memorandum of understanding prepared by Phoenix on 24 March 1998. The total consideration for the purchase of Cotesworth was agreed at £14.4 million. 250 In addition, Lloyd's required that HIH provide funds at Lloyd's or standby letters of credit issued by suitable banks to guarantee Cotesworth Capital's obligations.

At the April 1998 board meeting Williams tabled a paper on the proposed acquisition of the Cotesworth group. The acquisition was subject to due diligence, which was scheduled to begin on 20 April 1998. The due diligence team consisted of Andersen (financial), Coopers & Lybrand (actuarial) and Clyde & Co. (legal). 251

By 22 April 1998252 the conditional memorandum of understanding had been duly signed by the respective parties253 and HIH announced its agreement to acquire the entire share capital of Cotesworth Group Limited. 254 In May 1998 Britten, the managing director of Cotesworth, and Graham Davies, the senior underwriter of Syndicate 535, held discussions with HIH in Australia. Subsequent to those discussions, it was decided to proceed with the purchase of Cotesworth. 255

Between 3 and 5 June 1998 Andersen issued four volumes of a due diligence report to HIH256 and on 4 June Williams informed the board that due diligence had been carried out satisfactorily. The purchase was expected to be complete by 30 June. 257 As at 8 June, however, a list of due diligence queries remained outstanding

258 and

the following day Phoenix, Andersen and HIH (UK) representatives met to discuss Andersen's draft due diligence report.Z 59 Because of the lack of documentation from third parties, it was not possible for the Commission to investigate the due diligence exercise fully.

56 Unprofitable international operations

On 24 and 27 July 1998 the boards of HIH C&G and HIH European respectively approved various agreements, including a deed of option, that were necessary for the acquisition to proceed.260

The structure of the acquisition In total, Cotesworth was acquired for approximately £14.4 million. 261 The final structure of the transaction involved two parts-the acquisition of shares in Cotesworth Group Limited by HIH European, and also the acquisition by it of the

minority interests in Cotesworth Capital. The first part of the transaction (the purchase of the managing agency) was acquired for tangible assets plus £4.3 million for goodwill and a further £2.7 million for goodwill if Cotesworth management could increase HIH's corporate capacity on the Cotesworth syndicates to over

50 per cent over five years. A further £1.4 million was also to be paid to specific vendors. 262

Before the acquisition, the Cotesworth Group Limited owned 42.5 per cent of Cotesworth Capital; the remainder was owned by Walsham Brothers Co. Limited and Lureco.Z 63 Walter Copping, a director of Lureco, was made a director of Cotesworth Capital on 13 January 1998. The shares held by Walsham Brothers and Lureco in Cotesworth Capital were purchased pursuant to a share-purchase deed that was approved at a meeting of the directors of HIH European held on 3 December

1998?64 This formed the second part of the Cotesworth acquisition. The purchase price was an amount equal to Walsham Brothers and Lureco's share of the net tangible assets of Cotesworth Capital plus a goodwill payment of £555 556?65

Two documents-a declaration and an undertaking266 -were approved at the HIH board meeting held on 31 July 1998 after it had been resolved that the acquisition was in the best interests and to the benefit of the company. 267 The directors of HIH European met268 and the final documents relating to the acquisition were signed on

12 October 1998?69 By 2 December 1998 letters of credit270 totalling £84.16 million had been issued by Westpac and Societe Generale of Australia Limited (SocGen) for and on behalf of HIH. 271

Following the acquisition, Williams appointed Thompson as chairman of Cotesworth, an arrangement that lasted until January 2000. At that time Britten informed Williams that all directors of Cotesworth would resign if Thompson continued as chairman. After asking for Thompson's resignation, Williams took over as chairman until Wein's appointment. Neither Williams nor Wein attended any Cotesworth board meetings held between May 2000 and March 2001. 272

The Cotesworth syndicates wrote a variety of business. Two of them were named in the top 10 Lloyd's syndicates in the reporting period current at the time of the HIH acquisition. 273 The integrity of the reported results is, however, questionable for a number of reasons, including the impact of financial reinsurance contracts to which they were a party (see Section 13.1.12).

Between 15 March 2001 and 30 June 2002 the estimated reserve requirement for Cotesworth, as calculated by the KPMG actuaries, rose by over A$500 million to

The failure ofHIH Insurance 57

A$1 077 million. It is thought that a significant part of the increase in required net reserves was caused by the expected inability to collect reinsurance assets as a result of the previously undisclosed reinsurance arrangements (discussed below). 274

13.1.12 Financial rei nsurance arrangements

The financial reinsurance arrangements were between Cotesworth and Lureco and others and were established in 1991.275 During the negotiations Lureco described them as being a continuous contract in respect of losses occurring on or after 1 January 1991.276 They were also described by the parties as non-proportional reinsurance cover. Pursuant to its terms, Cotesworth would pay premium to Lureco every year. The premiums would be accumulated in an experience account, for which Lureco would issue an experience statement annually. That statement would reveal the balance of the income (all net premiums paid, all reinstatement premiums and a positive bonus on any credit balances) minus all paid losses and negative bonuses on debit balances. The positive and negative bonuses were calculated by reference to prevailing interest rates from time to time and were equivalent to interest on outstanding balances in the experience account.

There was no termination date under the contract but a right of cancellation was given to each party. If the reinsurer cancelled the reinsurance, it was required to pay Cotesworth 75 per cent of the positive balance held at that time in the experience account. If the cedants cancelled the reinsurance for reasons other than their ceasing to trade, they were required to pay the reinsurer any negative balance in the experience account. If, on the other hand, the cedants cancell ed the reinsurance by reason of their ceasing to trade, they had no obligation to pay the negative balance (if any) to the reinsurer.

The only sense in which risk was transferred to Lureco pursuant to this arrangement was the risk that, if the cedants cancelled the reinsurance as a result of their ceasing to trade when the experience account was in deficit, Lureco was responsible for meeting the shortfall in the account. Such a risk is known as counter-party risk, as opposed to normal risk transfer pursuant to a policy of reinsurance.

Before it entered into these reinsurance arrangements with Lureco, Cotesworth retained its auditor, Ernst & Young, to negotiate with the advisory department of Lloyd's in order to confirm that the arrangement was acceptable to Lloyd's.277 In early 1992 Lloyd's advised the auditor that the contract wording would not enable the syndicates concerned to comply with its regulatory guidelines on outwards reinsurance. 278 Lloyd's noted that the policy was not risk bearing and that there were difficulties accounting for the transaction appropriately. Lloyd's advised that it might be difficult for the underwriter to discharge its duty to its names in arranging reinsurance and allocating the risks and premium obligations under the contract

between various Lloyd's underwriting years, during which the extent of participation by the original members of the syndicate might have changed.

58 Unprofitable international operations

Representatives of Lloyd's, Cotesworth and their auditors and solicitors met in January 1992 to clarify the situation.279 Following that meeting, Cotesworth presented further submissions to Lloyd's?80 Lloyd's responded that its concerns had been satisfactorily responded to

281 , and Cotesworth's reinsurance arrangements with

Lureco continued substantially in this form until 1998.

During 1998 there was a significant variation to the Lureco reinsurance

arrangement, one that was to have an extremely detrimental effect on the ultimate financial deficiency of Cotesworth?82 The variation imposed an obligation on the cedants to pay the reinsurer any negative balance in the experience account if they ceased to trade. This removed the one and only risk that had been transferred under the original policy-the counter-party risk.

It is not clear what benefit might have flowed to Cotesworth from agreeing to this amendment. It is clear, though, that its consequence in 1998 was to greatly diminish the value of Cotesworth's reinsurance by eliminating the only risk it had transferred to the reinsurer under the original 1991 arrangements. There is no evidence of the

new arrangements being revealed to the board of HIH. Such a fundamental change should have been fully disclosed to HIH and should have been the subject of considerable attention and advice during the due diligence exercise. It is significant that the amending agreement put forward in evidence to the Commission was dated 6 August 1998283 -which was after the due diligence exercise had been completed

by HIH but before the Cotesworth acquisition had been approved by Lloyd's and finalised, in December 1998. The conditional heads of agreement between Lureco and HIH European for the acquisition of Lureco's shareholding in Cotesworth Capital was dated 7 August 1998.284

Conclusions The evidence does not permit any finding as to the state of knowledge of any of the relevant representatives ofHIH, Cotesworth, Cotesworth's auditor (Ernst & Young) or the auditor that undertook financial due diligence for HIH in connection with the acquisition (Andersen). The same applies to the actuarial and legal advisers

(Coopers & Lybrand and Clyde & Co.) who also undertook due diligence for HIH in connection with Cotesworth. The individuals concerned did not give evidence to the Commission, and much of the relevant documentation was not produced. It might be significant that the provider of the reinsurance, Lureco, was a shareholder in Cotesworth Capital before the HIH acquisition and that a director of Lureco, Copping, was also a director of Cotesworth Capital at the time. Copping resigned his directorship of Cotesworth Capital when HIH acquired Lureco's shareholding in that company in December 1998.

The information relating to this financial reinsurance became known to the Commission only at the end of its hearings. The Commission did not have the opportunity to inquire into the relevant parties' state of knowledge to determine whether HIH was the victim of a deceit perpetrated by third parties or whether it

failed to evaluate fully the information provided to it before the acquisition. As a

The failure ofHJH Insurance 59

result, it is not possible to make findings or recommendations in connection with this matter.

In Sections 14.2 and 14.3 and Chapter 16 I express concern at the use--or, more accurately, the abuse- of fin ancial reinsurance. Although I make no specific find ings in relation to the Lureco arrangements, they do provide a fu rther example that does little to allay my concerns.

13.1.13 Letters of credit

For a limited-liabi lity entity to participate as a member of a Lloyd' s syndicate­ called a 'corporate capital vehicle'-it must maintain funds at Lloyd' s as security to enable it to meet its liabilities as a member of the syndicate if the syndicate incurs losses. Any losses incurred above the assets of the corporate capital vehicle and the fund s at Lloyd's must be met by the Lloyd's Central Reserve Fund.285

HIH acquired the Cotesworth companies so that it would be able to write insurance business in the Lloyd' s market through a corporate capital vehicle. The funds at Lloyd's that Cotesworth was required to maintain varied, but at December 2000 amounted to £110 million. These funds took the form of letters of credit from banks, which in turn held security over assets owned by the HIH group of companies in Australia. Among those companies were HIH C&G, F AI General Insurance Company Limited and CIC. 286 The significance of those securities to the insurers' capacity to satisfy the net asset requirements of Australia's Insurance Act 1973 is discussed in Chapter 19.

The liquidators stated that by December 2002 all of the £11 0 million in letters of credit had been drawn down. 287

At a board meeting on 2 December 1998 Fodera advised that Westpac and SocGen had issued letters of credit totalling £84.16 million on behalf of HIH in order to secure sufficient funds to support Cotesworth Capital' s underwriting operation. The board resolved to ratify management's actions in securing those letters of credit. 288

As noted, on 2 November 2000 Fodera reported to the board that Cotesworth would have to be sold in two years ' time because it would demand substantial amounts of cash, preventing the use of that money elsewhere in the group. Further letters of credit would be required .Z 89 At a board meeting held on 29 November 2000, however, Fodera tabled letters from Westpac and SocGen setting out additional terms and conditions for the continuation of their banking support. 290 The board resolved that the letters of credit supporting Cotesworth be extended for another year.Z91 It was at this board meeting that Britten, who travelled from the United Kingdom to attend, spoke about the Cotesworth results, whereupon the board resolved that, subject to extension of the letters of credit by HIH's bankers, Cotesworth's operations would be supported into 2001. 292

60 Unprofitable international operations

13.1.14 The UK regulators

In the United Kingdom insurance regulation is the responsibility of the Financial Services Authority. Before 1999 it was the responsibility of the Department of Trade and Industry. 293

The FSA began regulating Lloyd's corporate capital vehicles on 1 December 2001. 294

The UK branch of HIH

HIH Casualty & General wrote business through its UK branch operations and was therefore subject to regulation by the FSA. 295 The normal regulatory requirement is that both the branch and the company globally should maintain a margin of solvency, as set by a statutory formula.Z 96

On 14 July 1997, at a meeting between representatives of the Department of Trade and Industry and Payne, Gordon McCall, Bird, Lusty and Piper, the department advised that the UK branch had 'one of the worst compliance records of any company currently underwriting'.297

At 30 June 1999 the FSA solvency margin requirement, as calculated by the UK branch, was net assets of £30 million. The UK branch did not hold that amount, so HIH in Australia earmarked £43 million to meet the shortfall. 298 Yet at October 1999 the negative net asset position of the UK branch persisted. 299 It therefore appears that between May and October 1999, no action was taken within HIH to remedy the deficiency. Williams, Payne, Sturesteps and Fodera had all been aware of the deficiency as a result of the May board meeting in the United Kingdom. For at least six months they did not communicate to the Australian board the serious

matter of a failure to meet solvency requirements. They should have. 300

At 30 June 2000 the required margin of solvency was £14.4 million, as a result of which HIH in Australia had to earmark £54.8 million to meet the shortfall of assets of the UK branch. 301 The UK branch submitted its June 2000 Global Insurance Return to the FSA on 31 January 2001. 302 The solvency deficiency at June 2000 was A$313.258 million, or £124.238 million. 303

The return was discussed at a meeting between representatives of the FSA and PWC and Wein, Lusty and Piper on 13 February 2001.304 The UK branch gave the FSA a copy of its forecast balance sheet as at 31 December 2000, which showed negative net assets of £36.9 million and earmarked funds in Australia of £51.4 million.305 The UK branch was required to submit the December 2000 return to the FSA by

15 February 2001. It did not do so. 306

The UK branch was also required to submit a 1 0-year plan to the FSA before Christmas 1999. It was prepared after the branch went into run-off in September 1999 to demonstrate the resources available to fund the run-off. 307 Large sums would be needed from HIH in Australia-£59.1 million directly and £29 million via Lureco. Lusty advised the UK board that the FSA 'seemed sympathetic towards the

Branch relying on the global solvency ofHIH C&G' ?08

The failure ofHJH Insurance 61

A meeting was held with the FSA in late 1999 to review the plan. The status of the run-off was then reviewed at a meeting on 17 March 2000309, and by the end of March 2000 the FSA had agreed to the UK branch's 10-year plan. 310

Cotesworth

All corporate capital vehicles writing business through Lloyd's syndicates were regulated by Lloyd's, which after 1 December 2001 was in tum regulated by the FSA. 31 1 Two reports prepared by Lloyd's in 2000 and 2001 were critical of Cotesworth & Co. and its syndicates. 312 The reports highlighted deficiencies in the company's management systems; non-compliance with Lloyd's regulatory requirements; inadequate attendance by Williams, and then Wein, as chairman of Cotesworth; an inadequate independent review function within Cotesworth; and a failure to improve the controls in respect of the binding authorities.

The Commission did not learn of the existence of these reports until the hearings were coming to an end; they were subsequently produced by Lloyd's. It is thus not possible to make any findings as to how or when HIH or Cotesworth saw copies of the reports. Both reports do, however, state that the findings and recommendations on significant matters identified during the review are ones that Lloyd's wished to bring to the attention of Cotesworth & Co. They also state they are confidential and intended for the sole use of the regulatory division of Lloyd's and of Cotesworth. 313

13.1.15 Possible confl icts of interest

Payne, Williams, Sturesteps, Geoffrey Cohen, Abbott and Stitt were directors of HIH and each held numerous private interests as individual names on Lloyd's syndicates. They did so in years for which the UK branch wrote direct reinsurance cover with those syndicates or was indirectly exposed through retrocession cover of

liabilities of the syndicates. The possibility for a conflict of interest thus arose? 14

Payne first became a name at Lloyd's in 1955 and gradually extended his interests over the years. 315 Williams had been a name since 1972. 316 Sturesteps was a name in the mid-1980s and again during the period 1993 to 1997.317 Stitt had been a name since 1980318 , Abbott since 1977319 and Cohen for approximately 20 years, from about 1979 until 2000. 320 None of those directors considered that a conflict of interest arose as a consequence of their participation as names at Lloyd's.

In 1995 the UK branch started writing inwards excess-of-loss reinsurance business and some quota-share reinsurance for a number of prominent Lloyd's syndicates. 321

It also wrote stop-loss insurance for Syndicate 386/683. Payne was the founder of the syndicate (otherwise known as Michael Payne & Others Syndicate) and an active underwriter from 1973 to 1991. 322 In 1993 Williams, Sturesteps, Cohen, Abbott, Stitt and Payne were all names in this syndicate.323

After HIH acquired Cotesworth, Cotesworth & Co. took part in the auction process to acquire additional capacity for Cotesworth Capital, particularly in Syndicate 535.324 Payne, Williams, Sturesteps, Cohen and Stitt were names in Syndicate

62 Unprofitable international operations

535 325 ; Williams and Payne acquired additional capacity in that syndicate in

1998 326; and Payne, Williams, Cohen and Stitt acquired further capacity in it at the auctions held in 1999. 327 After the 1999 auctions John Hale, Cotesworth's compliance officer, informed Payne that he considered there was a conflict of interest and advised Payne not to purchase further capacity. 328 Thompson advised Payne329, and thereafter Williams330, in similar terms.

Both Payne and Williams acquired capacity in Syndicate 535 at the auctions for the 2000 underwriting year331 , although Payne testified that the allocation for that year was a pre-emption whereby all existing names were offered additional participation at no cost. 332

In February 2000 Britten circulated a memorandum to Williams, Cohen, Payne, Stitt and Alan Bonner in relation to acquiring additional capacity in Syndicate 535 at the auctions to be held that year. 333 This presumably related to capacity for the 2001 underwriting year. 334 Essentially, the memorandum warned against such acquisition because of the consequences for Cotesworth: any further acquisition of capacity by directors would be counted towards the capacity of the managing agent, which if it reached 75 per cent would require Cotesworth Capital to make an offer at the

highest possible price for the remaining capacity.335

Reference was made in HIH's annual reports for the financial periods 1997, 1998 to 30 June 1999 and 2000 that controlled entities entered into reinsurance transactions with Lloyd's syndicates, some of whose directors were names?36

Conclusions There does not appear to have been an actual conflict of interest of any material nature. The evidence does not reveal that HIH suffered any loss or that the directors gained any profit through the preferment of their private interests to HIH. But there was undoubtedly the potential for a conflict of interest. It is not apparent that the directors concerned took any, or any adequate, steps to eliminate that potential. The question was specifically asked of Williams and Payne. Britten's memorandum of February 2000 should have alerted the other directors who received it to the possibility. They should have been aware of the potential, regardless of whether

they received the memorandum.

Conflict between interest and interest or between interest and duty is at the very heart of fiduciary relationships. The relationship between director and company involves a fiduciary obligation on the part of the former to the latter. The failure of the named directors to consider fully the potential for conflict of interest and to take steps to eliminate that potential constitutes an undesirable corporate governance practice. The board minutes of HIH do not reveal any appropriate discussion of the matter.

The finding of undesirable corporate governance is directed at Williams, Payne, Abbot, Stitt, Sturesteps and Cohen for their failure to take steps to reduce the potential for conflict of interest-in particular, through their private participation in

The failure ofHIH Insurance 63

syndicates managed by Cotesworth and in syndicates with whom HIH conducted significant reinsurance business.

Stitt gave evidence that he did not believe his underwriting membership of Syndicate 535 gave rise to any conflict between hi s personal interests and the interests of HIH. He said that at no time did he knowingly place himself in a position where there was a conflict between his personal financial interests and the financial interests of HIH and its shareholders. He did not believe that there was in fact any such conflict.337 Abbott gave evidence that if he had been alerted to a possible conflict of interest he would have disclosed it to the board, absented himself while it was discussed, abided by any decision of the board and, if necessary, resigned from the board. 338 Williams said that at no time did he set out to put himself in a position of conflict between being a Lloyd's name and being chief executive of HIH, and nor did he consider he had done so. He said a number of HIH directors-Cohen, Stitt, Sturesteps, Payne and Abbott- were names and this was known to the board. Further, he stated that none of those directors ever thought that a situation of conflict was created by the fact that they were names. 339 Payne considered that reference in HIH' s annual reports to controlled entities having entered into reinsurance transactions with Lloyd' s syndicates, some of whose directors were names, was disclosure, at least on his behalf, of his interest in Lloyd's syndicates. 340

In the circumstances, disclosure does not detract from a director's obligation to avoid circumstances that offer the potential for giving rise to a conflict of interest.

13.2 Operations in the United States

During its push into the US market in 1997 and 1998, HIH's US operations were conducted principally through HIH America Compensation & Liability Insurance Company and Great States Insurance Company.

The failure of the US operations can be attributed to three main factors:

• inadequate assessment of the pros and cons of re-entering the Californian workers compensation market

• inadequate action by the boards of the subsidiaries and of HIH when losses began to emerge

• the boards' failure to recognise and disclose the losses when they occurred.

These three failings contributed to the collapse of the US operations and ultimately ofHIH itself.

64 Unprofitable international operations

13.2.1 Losses resulting from US operations

It is estimated that, as at 30 September 2002, the losses resulting from HIH's operations in the United States amounted to approximately A$620 million. 341 The vast bulk of these losses can be attributed to the reacquisition of the Californian workers compensation business in 1997. This occurred without adequate due diligence and against the wishes and advice ofHIH's major shareholder, Winterthur. The decision to reacquire the business was made on the basis of anecdotal evidence that the Californian workers compensation insurance market was becoming profitable, having been through an unprofitable phase. But the reality was that after HIH's re-entry the market remained unprofitable. HIH continued to write business and expand its presence in the US market despite the fact that the business was

unprofitable.

13.2.2 Entry into the US market and the sale of Heath Cal

HIH started operations in California in 1987, when it established CE Heath Compensation and Liability Insurance Company, or Heath Cal, to write workers compensation business.342

Sturesteps was president of the group of US subsidiaries until his resignation on 12 September 2000. 343 Peter V McCarthy was manager of marketing and underwriting and was based permanently in the United States. 344 By the end of 1988 McCarthy had assumed the role of deputy chief executive.345

Before HIH moved out of the Californian market the workers compensation insurance industry was regulated by the state government, which prescribed a 'minimum rate' that could be charged to a policyholder. In the early 1990s legislation to deregulate pricing and create an 'open rating ' system was to be introduced, whereby insurers could charge what they wanted for policies.

346 The

imminent introduction of the legislation caused HIH to consider leaving the US market to other players.

Colin Richardson, of Hambros Corporate Finance Limited, advised HIH on the sale of Heath Cal. 347 He informed the Commission that he prepared for HIH a report that noted the Australian investment community's concern about the HIH business in the United States?

48

On 3 March 1994 HIH received from UniHealth America-then parent of CareAmerica Health Plans Inc.- a conditional offer for the purchase of Heath Cal. 349 In April 1994 the sale was ratified by HIH and CareAmerica; it was completed on 19 August 1994.350 HIH gave the vendor an indemnity against deterioration in outstanding claims.35 1

Heath Cal was sold for $118.6 million352 , which generated an abnormal pre-tax profit for HIH of $24.5 million. 353 The company subsequently changed its name to CareAmerica. 354

Th e failure ofHJH Insurance 65

13.2.3 Reacquiring the Californ ian business

In late 1996 HIH began negotiating to re-enter the US market through reacquiring CareAmerica, the (renamed) company it had sold in mid-1994.

On I January 1995 'open rating' had been introduced in California, removing most of the pricing regulation in the Californian workers compensation industry. 355 The effect was a dramatic increase in competition, which in turn caused a big reduction in premium prices and in the profitability of insurers writing this class of business.

356 Nevertheless, HIH management believed that the Californian workers compensation market had bottomed out after two years of deregulation and that it was a good time to re -enter the market. 357

The Hambros briefing paper: December 1996 In December 1996 Fodera circulated to the HIH directors a briefing paper prepared by Hambros and stated,

As addressed in the Hambros paper this is an opportune time to re-enter the Californian Workers Compensation market. The current owners are looking for 1.1 or 1.2 times net assets compared to our sale at 2 times net assets.rss

The briefing paper was based on documents provided to Hambros by CareAmerica adviser Dean Witter Reynolds in late November 1996. It noted a number of positive aspects of the reacquisition, including:

• consolidation of the existing HIH-Winterthur Hawaiian operations

• a significant increase in group premium in a single transaction and

diversification of risk base

• additional benefits of know-how and international presence through exposure to the United States-the predominant workers' compensation market in the world

• an increase in net assets of approximately US$13.2 million through the discounting of CareAmerica reserves under Australian accounting standards. 359

Hambros expressed the view that the fundamental consideration for the board in determining whether to proceed with the transaction was the prognosis for the Californian workers compensation market in the next five years. It referred to anecdotal evidence, supported by senior management of CareAmerica, that suggested improving market conditions. Forecasts provided by CareAmerica's senior management were said to offer attractive returns. 360

The forecasting model indicated that revenue and net earned premium would rise between 1997 and 2000 and that the number of claims would reduce as a proportion of premium income. 36 1 Reductions in underwriting expenses in each of those years were also forecast. 362 Hambros stated that the forecast model had not been audited or verified and, if the acquisition were to proceed, further due diligence would be necessary. 363 This did not occur.364

66 Unprofitable international operations

Historical financial information contained in the briefing paper indicated that during periods of deregulation in the workers compensation business, loss-ratio levels made it difficult for workers compensation insurers to generate profits.365 Hambros stated that evidence from other states where deregulation had previously occurred suggested that rates would rise in 1997.366 Deregulation in other states had resulted

in an initial dip in profits in the first two years but a stable environment thereafter367 , and HIH was expecting the same trend in the Californian market.

The briefing paper was tabled at the December 1996 board meeting and approval was given to Williams to enter into negotiations. The purchase offer was to be US$67 million. 368

The Winterthur representatives on the HIH board opposed the reacquisition.

On 13 December 1996 Hambros prepared a 'summary of offer' letter which showed that CareAmerica would accept an offer equivalent to 1.2 times its net assets. 369 It also stated that CareAmerica would attach considerable value to an offer that did not expose it to any residual risk after the sale was completed? 70 As a result, HIH did

not seek a reserve indemnity, which would have rendered CareAmerica liable for any inadequacy in the provision made for claims on business written in the two years the company was owned by CareAmerica.

At that time the HIH board was informed that HIH already had exposure for claims up to completion of the sale of Heath Cal in 1994. Therefore, it was said, by reacquiring CareAmerica the additional risk assumed related only to the two years following the sale.371 This was used to justify the absence of any indemnity against a deterioration in outstanding claims, in keeping with the vendor's wishes.

The reacquisition was subject to limited due diligence. HIH retained Timothy Perr of Timothy B Perr and Company, an independent actuary, and Andersen (US), HIH's external auditor, to conduct limited due diligence. Milliman & Robertson was the consulting actuary to CareAmerica and had carried out an actuarial assessment for it. The actuarial analysis of the loss and loss-adjustment expense reserves for

CareAmerica recorded that CareAmerica's reserves at 30 September 1996 were almost US$5 million greater than the estimates and that its carried reserves were US$115.2 million.372

This actuarial report was the subject of a desktop review by Perr. 373 He was not

asked to analyse the source material in order to give an opinion about whether the reserves were appropriate 374 ; instead, his review involved several readings of the

Milliman & Robertson report and conversation with its author.

No revised comprehensive actuarial analysis was sought by HIH: it simply relied on the Milliman & Robertson report.375 This was despite Hambros' report to the board recommending that no reserve indemnity be sought and that a 'full due diligence' be undertaken on the reserves.376 No appropriate due diligence was undertaken on the reserves, and that fact was not drawn to the board's attention.

Th e failure ofHIH Insurance 67

Perr reported that, because of time constraints, his report was limited to analysis of the indemnity benefit and medical benefit components, which accounted for 80 per cent of CareAmerica's total net reserves. Perr referred to the difference between the central estimate calculated by Milliman & Robertson (US$110.6 million) and the reserve carried by CareAmerica (US$ 115.2 million) and cautioned HIH by drawing attention to two particular features of the approach taken by Milliman & Robertson- the apparently low provision for outstanding claims and the decreasing loss ratios from accident years 1995 and 1996. These features did not accord with Perr' s understanding of the California workers compensation market. 377

The summary of the negotiations, prepared by Hambros, disclosed that the sale would cause a US$40 million loss to CareAmerica. The document recorded that no reserve indemnity had been sought and that Milliman & Robertson stood by its previous assessment of the level of reserves held by the company.378

On 23 December 1996 Williams advised the HIH directors that HIH had offered A$79.3 million for the company.379 He enclosed a paper from McCarthy that referred to CareAmerica' s business strategy and stated that the Californian workers compensation premium appeared to have bottomed out in 1996 and that the acquisition was most opportune and timely.380 McCarthy's basis for asserting that the market had bottomed out was conversations with people involved in that market. No study or test was done to determine whether this was in fact the case nor was

there any empirical evidence to the effect that rates had in fact hardened.

Conclusions HIH directors gave conflicting evidence in relation to the part Hambros played in the reacquisition?81 Some thought Hambros was advising on the desirability of the transaction generally382 , a proposition Richardson denied. 383 Others thought Hambros had a more limited role. 384 Whatever the actual terms of the Hambros retainer, the reports to the board that it produced managed to engender a belief that it had a general advisory role in relation to the transaction. Some of the directors testified that they were unsure of what due diligence had been carried out. 385 However, the documents provided to the board should have made all directors aware of the limited nature of the exercise: in addition to the inadequacy of the review of provisions for future claims, there was no review of US management's assumptions about future revenues, costs and claims. This was despite the fact that Hambros' earlier report to the board had foreshadowed that full due diligence

386 would be

carried out on those assumptions.

Both management and the board of HIH based their decision to reacqutre CareAmerica on the following points:

• HIH had made a profit from the company prior to its sale in 1994 and they thought this could be done again.

• CareAmerica had retained a number of personnel when it bought the company in 1994.

68 Unprofitable international operations

• Going by market history, two years following deregulation the market would stabilise and return to profit.

These are legitimate points. But they provide no justification for the inadequate due diligence and, in particular, the failure to fully test the reserves and to test the revenue and loss assumptions in the forecast model. Instead, the decision to reacquire was based on a superficial, inadequate review of the reserves and HIH's unsupported conviction that the Californian workers compensation market would

become profitable.

The acquisition and subsequent operating environment The sale agreement was dated 31 December 1996 but the acquisitiOn was not completed until 12 June 1997.387 The purchase price was $79.3 million. After the reacquisition the company was renamed HIH America. 388

Even before the reacquisition was complete, evidence emerged that the operating environment for workers compensation insurance in California was deteriorating rather than improving. 389 The management of HIH seems to have ignored this assessment: no steps were taken to ensure that the business was being conducted prudently in such an environment. HIH chose to reject this and subsequent

assessments and to rely upon its own optimistic outlook. It could, for example, have purchased additional reinsurance cover to protect against claims deterioration, but it did not do so at the time of the acquisition.

One important factor causing the deterioration in the operating environment for workers compensation insurers in California was a court decision in 1997 expanding the authority of 'primary treating physicians', who were given virtually unilateral authority to determine levels of disability. The courts applied the decision to claims

occurring before the date the legislation had taken effect, which caused claims costs to rise retroactively. 390 Insurers thus found they had underpriced workers compensation risks. Further, because it was long-tail insurance, the losses flowing from the contracts took years to emerge and , by the time they did , the extent of

competition in the marketplace and the consequent inability to recoup past losses via premiums on new business meant that little could be done.

13.2.4 Great States

The second major acquisition in furtherance of HIH's ambitions for growth in the US market was Great States Insurance Company. Great States specialised in workers compensation insurance and began operating in California in February 1989. 39 1 In February 1998 HIH America considered purchasing Great States so that the US operations could expand into Arizona and nearby states.392 HIH acquired Great States in July 1998 and the sale was completed on 31 October that year

393 , for

a purchase price of US$14 million. 394

The acquisition of Great States allowed HIH America to book a profit upon acquisition, which had the effect of improving the balance sheet of the US

Th e failure ofHIH Insurance 69

operations. HIH decided that the prudential margin Great States had applied was no longer required and could be released to profit. 395

HIH America retained Milliman & Robertson to perform the actuarial due diligence of Great States. No reserve indemnity was sought from Great States, although it was agreed that the reserve for claims losses and loss expense would be set at not less than the estimate of reserves at the end of the month preceding closing. 396 HIH booked reserves of US$70 million and planned to purchase reinsurance to protect against any blow-out. 397

13.2.5 Policyholder dividends

Great States did not have under-reserving problems as severe as those of HIH America. But there were other causes for concern, one of which was the treatment of policyholder dividends.

A policyholder dividend is a dividend paid to a policyholder who does not make a claim. The insurer has no legal obligation to declare such a dividend, but it is recognised as good business practice because it promotes customer loyalty. The dividend is paid from the insurer' s profits. It follows that if the insurer decides not to declare such a dividend the sum set aside for that purpose can be released to profit.

For the period ending 30 April 1999 Great States recorded US$11 million in earned but undeclared policyholder dividends in relation to Colorado and Arizona policyholders.398 It did not carry a reserve for payment of the dividends; that had been released to profit in 1998.

On 21 May 1999 HIH America management informed Waters that if Great States failed to declare the dividend for the year, its position in the market was likely to become tenuous. 399

Removal of Waters' internal audit report At Thomas Matson's request, management's view about Great States' treatment of policyholder dividends was recorded in Waters ' internal audit report 400

, which

noted,

If GSIC [Great States] fails to fulfil its obligation or is otherwise unable to manage its way through a factored PHD [policyholder dividend] strategy, GSIC's market throughout the Southwest region will likely

become tenuous. 4 1

On 14 June 1999 Waters sought confirmation that his internal audit report would be included in the board papers for the board meeting scheduled for 16 June. 402 On the day of the meeting, however, Terry Griffith, secretary and legal counsel for HIH America, told Waters that the report had initially been included with the board papers but that Williams had instructed that it was to be taken off the audit

committee agenda and so it was removed from the board packs. Instead, the audit report would be submitted at the September 1999 audit committee meeting. 403

70 Unprofitable international operations

On the day after the board meeting Matson advised Waters by email of Williams's concern that disclosure of the policyholder dividend situation in the internal audit report might compromise HIH America' s ability to achieve its June 1999 forecast results.4 04

Waters' report was removed from the agenda for the meeting because there was a fear that, if Andersen became aware of management's intentions, it might not be willing to sign off on the results reported as at 30 June 1999.405 A total of US$11 million would have been eliminated from the reported earnings. This would have appeared in Great States' results as at 30 June 1999 and consequently would

have affected the results reported by HIH on consolidation.

Williams's evidence Williams confirmed in evidence to the Commission that he gave the direction to remove the internal audit report from the board materials. 406 He asserted, however, that his primary motivation was concern about the discoverability of the report in any legal proceedings initiated by Great States policyholders challenging the decision not to pay policyholder dividends rather than any concern about its effect

on the June 1999 forecast results. 40 7 This explanation is troublesome. It is illogical because the discoverability of the internal audit report was unrelated to its presentation to the audit committee If it existed, which it did, its character as a discoverable document would not be affected (one way or the other) by its disclosure to the audit committee.

At the time he gave the direction to remove the report, Will iams was neither a director of HIH America nor a director of Great States.408 The only authority and capacity Williams had to order its removal was as a director and chief executive of HIH. He knew that allowing the report to go before the auditors could have had an adverse effect on the 30 June 1999 reported results, which on consolidation would have affected the accounts of HIH. In ordering the removal of the report, Williams engaged in an inappropriate exercise of his duties as an officer of HIH.

In closing submissions Williams stated that Griffith did not provide evidence before the Commission and that Williams was the only person to give evidence of his conversation with Griffith and the reasons for the removal of the report. He further stated that the Commission relied solely on the emails of Matson, who had also not provided evidence before the Commission.409 But Williams did admit directing

removal of the report. Although Matson did not provide evidence to the Commission, his email to Waters41 0 was a contemporaneous business document indicating that Williams said his reason for issuing the instruction was that there could have been an effect on the 30 June 1999 reported results.

Conclusion The question that arises is whether, in these circumstances, Williams might have breached s. 232(2) of the Corporations Law. I note that proceedings of thi s nature would involve taking evidence from overseas witnesses. I did not hear evidence from either Matson or Griffith, and their evidence is obviously of critical

The failure ofHIH insurance 71

importance. In the absence of that testimony, I make no finding that Williams might have contravened the Corporations Law.

The matter of policyholder dividends illustrates both Williams's determination to minimise the impact on the 30 June 1999 results and an inappropriate exercise of his authority in removing an audit report from review by the external auditor. The board of HIH America should have had that information. HIH America was an indirect subsidiary of HIH. In intervening as he did in the affairs of a subsidiary, Williams

involved himself in an undesirable corporate governance practice in relation to HIH.

Ultimately, the HIH board was made aware of the matter through Waters' internal audit activity report411 , which was tabled at the August 1999 audit committee meeting and stated:

The management of policyholder dividends represents a future risk for the Company, with approximately USDllM in earned but undeclared dividends for which GSIC [Great States] had no financial provision. The matter was to be subject to Board deliberations in June 1999. 412

On 16 September 1999 the board of Great States resolved to pay the policyholder dividends in part. 4 13 These two actions provide additional reasons for not making a finding that Williams might have contravened the Corporations Law and for not referring the possible contravention to an agency.

13.2.6 Internal audit and reporting of results

What follows is an account of the various auditing and regulatory procedures associated with the US operations, the less-than-satisfactory responses of US management, and the HIH problem areas disclosed by those processes.

Following the reacquisition of CareAmerica, Waters took control of the internal audits of the US operations.414 He prepared detailed analyses and internal audit reports on the individual US subsidiaries, and these were provided to the boards of HIH America and Great States. He also prepared internal audit activity reports, which summarised the main findings contained in the internal audit reports prepared during the preceding period; these reports were provided to the HIH board and audit committee for the purposes of reporting on the US operations. 415

The results of the US operations were reported back to HIH each quarter. They were then reported in HIH' s financial reports and presented at the audit committee and board meetings. The financial report for the period to 31 December 1997 recorded a core underwriting profit of$12 million for the US operations. 41 6

Internal audits: July 1997 and May 1998 Waters first visited the United States in' July 1997 and prepared an internal audit report on HIH America as at 30 June 1997. The report identified 25 problems that needed to be resolved. The main areas of concern were inadequate control over product pricing and inadequate discount rates being used by underwriters.

417

72 Unprofitable international operations

The second internal audit report was prepared as at 31 May 1998.418 Waters reported a significant adverse difference between the actuarial estimates and HIH America's net reserves. As at March that year, the difference between the actuarial estimate and the booked reserves was US$11.7 million.419 The increase in the loss ratio in

1997 was indicative of the emergence of an unfavourable trend, and Waters cautioned that the adequacy of the reserves would need to be closely monitored.420 He also reported that underwriting activities continued to cause concern because of a failure to adhere to standards established within the HIH America office. 421

The Andersen management letter Whilst conducting the May 1998 internal audit Waters became aware that Andersen had recently completed a review but would not be producing a management letter. Matson informed Waters that Andersen had not identified anything that warranted

attention and that it would therefore not need to issue a management letter. 422

On 22 July 1998 Waters asked Tim Mooney, chief executive of HIH America, why a management letter would not be provided by Andersen. 423 On 28 July Waters raised the matter again and informed Mooney that, although he was initially told that nothing of note had arisen from Andersen's audit, he subsequently learnt there was a question about loss reserves.424 Matson responded425, saying that the question centred on Andersen's tolerance of the carried reserve shortfall against the Milliman & Robertson central estimate. Andersen was concerned that the carried reserve

shortfall was more than 5 per cent of the estimate recommended by Milliman & Robertson.

HIH America had an informal agreement with Milliman & Robertson and Andersen that allowed it to be within 5 per cent of the actuarial central estimate. This meant that, as long as its booked reserves were within 5 per cent of the figures set by

Milliman & Robertson, the auditor would accept the figure. 426 In fact, carried reserves at the time were 6.5 per cent below the actuary's estimate. The problem was resolved when Fodera, on behalf of HIH C&G, issued $3 million of reinsurance cover. 427

On 17 August 1998 Mooney informed HIH America executive management that the overall result of the internal audit was negative and below the acceptable margin. As a result, a formal audit committee for the US operations was formed.428

The results of the May 1998 internal audit and the issue of the Andersen management letter were contained in an internal audit activity report429 tabled at the HIH audit committee meeting held in August 1998.430 Waters reported to the following effect:

• Andersen had not produced a management letter because year-end audit work was still the subject of ongoing discussions with the Andersen Sydney office.

• Many matters arising from the previous audit had been said to be resolved but actually remained unresolved.

The failure ofH!H Insurance 73

• Underwriting activities were not operating m accordance with standards established by HIH America management.

• There were accumulated backlogs and unresolved reconciliation problems in the accounting operations.43 1

In relation to claims reserves, Waters reported a widening difference between the figures calculated by HIH America management and Milliman & Robertson, with indications of a continuing deterioration in results. 432 At the time of the due diligence, Milliman & Robertson had claimed CareAmerica's reserves were US $115.2 million-some US$4.8 million more than the M&R recommendation. 433 After one year of control by HIH America, the booked reserve was US$11. 7 million below the actuarial estimate. 434

Developments in late 1998

A further internal audit report was prepared as at 31 October 1998.435 The results for 1998 had deteriorated in the third quarter as a result of the impact of reinsurance programmes. The mechanisms that had been introduced to protect the 1997 and 1998 years of account had been substantially used-only US$3 million and US $4 million respectively remained as at 30 September 1998. 436 The core earnings figure as at 30 June 1998 had disclosed a loss of US$800 000. Waters found that underwriting expenses were US$1.2 million in excess of budget and underwriting results were US$7.2 million below budget; this represented a core underwriting loss of 107 per cent against a break-even budget. Core earnings results were US $7.6 million below budget. Waters noted that, were it not for the reinsurance programmes, the losses sustained would have been far more significant.437

In December 1998 Matson advised Fodera that Andersen was uncomfortable with HIH America' s carried reserves as a result of the completion of Andersen's actuarial work on the September study. Andersen was concerned that the problems encountered in 1998-a US$30 million deterioration-would recur. 438

HIH' s financial report for the year ending 31 December 1998 was presented to the HIH board in February 1999. It recorded a $900 000 profit for the US operations.439 But the deterioration in reserves for the period to 30 June 1998 was $34 million, and there was a further deterioration in pre-acquisition claims reserves of $12 million recorded against goodwill. The combined deterioration- $46 million- represented the assessment of under-reserving at the time of acquisition. 440

Waters tabled an internal audit activity report to 31 December 1998441 at the February 1999 audit committee meeting. 442 His report recorded a deterioration in HIH America's results because of the cost of reinsurance programmes and adverse developments in outstanding claims . . He further reported that HIH America's operations were being closely monitored and there was a need for caution in the assessment of the operation. 443

74 Unprofitable international operations

Developments in 1999

In mid-April 1999 the international rating agency AM Best advised Matson of the downgrade of HIH America's rating to B++ because of poor underwriting performance, a deterioration in capital and strained liquidity. The poor underwriting performance was caused by the significant deterioration in claims reserves recognised in 1998.444

Waters carried out an internal audit review in May 1999 to assess the adequacy of action taken in response to the previous audit. 445 He found that no internal compliance systems existed to ensure that claims operations were being internally managed. The claims area continued to operate under severe pressure because of the strong focus on the Californian operations, with no corporate control of remote

locations. 446

An internal audit report of HIH America as at 31 July 1999 noted that losses were continuing and the company's liquidity was deteriorating. 447 The financial position of the US operations was looking grim, and US$45.8 million of the HIH America group's

448 investments had been sold to reduce the amount of outstanding claims liabilities. HIH America management had advised that the group no longer had any free investments available to liquidate and was relying on approval from the Californian Department oflnsurance to free up US$15 million of pledged assets. 449

Waters noted that HIH America had received a capital contribution of US$25.4 million from HIH in June 1998. 450 Notwithstanding this, the net equity position had deteriorated since then from US$82.4 million to US$74.3 million as a consequence of losses sustained in the year to June 1999, by which date the losses

amounted to US$20.3 million. 451

The cash flow analysis prepared by Waters disclosed that in the 18 months to 30 June 1999 there was a deterioration of US$ 54 million of cash flow, such that by June 1999 cash outflow exceeded cash inflow by US$43.6 million. 452

As at 30 June 1999 the California company had sold US$32.2 million of investments.453 Loans from Great States and HIH Hawaii totalled US$23.7 million. Together with the capital tnjection of US$25.4 million, approximately US$81.3 million had been contributed to HIH America. The operation made a loss ofUS$23.6 million, further eroding the net asset position of the company.

454

An internal audit activity report dated 17 August 1999 was tabled at the audit committee meeting later that month. 455 The report recorded that HIH America continued to present difficulties.456 There had been a lack of preparation in advance of Waters' visit to the United States and the managers responsible had not been available. Previously identified problems remained unresolved.

457

At 31 December 1999 HIH America carried net reserves of US$157 .6 million. The actuarial analysis estimated reserves at US$165.8 million. The difference of US$8.2 million represented the margin of 5 per cent permitted by Milliman & Robertson and Andersen. 458 HIH's financial report for the half-year ending

The failure ofHJH Insurance 75

31 December 1999 recorded a core underwriting loss of $9.6 million for the US operations. Core earnings were lagging behind budget, but written premium had exceeded budget in California. 459

The internal audit activity report to 31 December 1999, tabled at the HIH audit committee meeting held in February 2000460, commented that the 31 July 1999 audit had been conducted during a difficult period for HIH America-a period characterised by continuing underwriting losses, liquidity problems, restructuring and downsizing of activities, and changes in senior management.461 Whatever optimism the directors might have felt about the prospects of the US operations should have been dispelled by these events.

Continuing deterioration in 2000 Waters conducted the final internal audit review of the US operations as at 29 February 2000.462 He concluded that HIH America continued to experience core earning and underwriting losses, which in tum caused decreases in capital and shareholder funds. There was uncertainty as to the adequacy of the increased premiums to support the significant claims volatility.463 At that date HIH America had booked US$173.4 million in claims liabilities.464

The HIH financial report for the year ending 30 June 2000 recorded a $53 million core underwriting loss for the US operations.465 The report noted, 'Management have continued with the United States accepted practice of being within 5% of the actuarial central estimate of reserves'.466 The 'US accepted practice' was to be within 5 per cent of Milliman & Robertson's central estimate. In fact, on this occasion it was David Slee, the HIH group consulting actuary, who had conducted the actuarial assessment of HIH America' s reserves, not Milliman & Robertson.467 On that basis, the statement in the financial report was not accurate. HIH America had chosen to disregard Milliman & Robertson ' s analysis as at 31 March 2000 and instead had Slee provide an estimate as at 30 June 2000. This matter is discussed in some detail in Chapter 15.

Milliman & Robertson' s 31 March 2000 analysis indicated a shortfall in HIH America's reserves of US$55 million. Slee's 30 June analysis identified the difference between his figure and that booked by HIH America as US$20 million. 468

The internal audit activity report to 30 June 2000 tabled at the September 2000 HIH audit committee meeting469 reiterated that HIH America continued to experience core earning and core underwriting losses. 470 HIH America' s books showed a loss of US$59.2 million for the year to 30 June 2000.471

In his address to the 2000 annual general meeting of HIH, held in October, the chairman, Cohen, acknowledged that oqe of the specific factors that had caused HIH financial harm was its re-entry into the US market. 472 At that stage HIH America had been placed into run-off. 473 Cohen explained that HIH had sought to buy back an interest in California at the bottom of the market, but that there had been no improvement in the market and premium as a percentage of payroll was the lowest it

76 Unprofitable international operations

had been in a decade. In addition, claims costs had risen by over 50 per cent since 1995. 474

Andersen's examination of the financial models was hopelessly inadequate. This is not a criticism of Andersen. It did what it was asked to do; the problem was with the brief. Perr's desktop review of the claims reserves was seriously inadequate. Again, this is not a criticism of Perr: he did what he was asked to do. The decision not to seek an indemnity against adverse claims development was flawed. The

performance of management and the board in relation to these matters was not what it should have been.

Conclusion

HIH's internal audit reports and internal audit actlVlty reports illustrated the continual deterioration in reserves and the persistent failure of US management to book reserves in accordance with actuarial recommendations.

Each of the di rectors of HIH- being in receipt of the internal audit activity reports- was on notice as to the problems identified by the internal auditor in respect of the US operations. The directors admitted they knew of the problems stemming from the US operations.475 It was their responsibility to make inquiries and to take steps to resolve the problems, yet no structured plan was developed or

implemented.

For two years the board of HIH was told every quarter that the US operations were losing money. The losses were attributed to the fact that claims arising from past business were continually underestimated, and there was a recurrent need to increase provisions for those claims. The HIH board failed to come to grips with the obvious fact that there was a systemic flaw in the assessment of claims arising from past business. It also failed to direct that steps be taken to ensure that the provision

made for fu ture claims arising from past business was adequate, once and for all. Instead, it allowed the US business to die the death of a thousand cuts as the inadequacies in past provisions became more and more evident with the passing of each quarter.

13.2.7 Sale proposals-2000

Because HIH America and its management did not inform the HIH board of the existence of the 1999 report of the Californian Department of Insurance, and of the Milliman & Robertson476 and Tillinghast477 reports of 31 March 2000, HIH was denied the opportunity to consider closing down the US operations earlier than it did and to report fu lly the financial consequences of the US operations. Again, this is discussed in Chapter 15 .

In mid-2000 HIH decided to try to sell the US operations478 , and by October there had been negotiations between HIH America and Alaska National Insurance Company. As of 30 October, HIH America ceased to incur underwriting risk on

The failure of HJH insurance 77

workers compensation insurance in the United States and had negotiated with Alaska to transfer to it a portion of the HIH America business and assets. 479

The US operations announced the sale to Alaska pursuant to a resolution of the HII-1 board on 31 October 2000. Because of the lack of profitability, it was recommended that the company leave the US market. 480 The renewal rights to HIH's Californian workers compensation insurance business went to Alaska, while Argonaut Insurance Limited bought the smaller and scheme business. 481

13.2.8 US regulators

The Californian workers compensation market was regulated by the state's Department of Insurance, which conducted triennial audits of insurance carriers, with each examination taking about 18 to 24 months to complete. The department audited HIH America as at 31 December 1997. This involved a review of HII-1 America's practices, procedures and management records; tests and analysis of detailed transactions; evaluations of assets; and a determination of liabilities. 482

The Department of Insurance report The Department oflnsurance report, completed on 22 October 1999, detailed many disturbing facts about the business-including that HIH America was under­ reserved by US$57. 75 million as at 31 December 1997. 483 It also revealed that I-III-I America's policyholder surplus had deteriorated from approximately US$62 million to US$15.3 million484 between 2 July 1993 and 31 December 1997. 485 There were four important findings:

• In 1998 HIH America had entered into three retroactive reinsurance contracts that were used to reduce losses by US$33 million but had failed to account for the contracts in the manner prescribed by the National Association of Insurance Commissioners Accounting Practices and Procedures. It should have accounted for them as deposit transactions, not as reinsurance contracts. 486

• In April 1999 HII-1 America had applied to withdraw funds on deposit with the Department of Insurance. The deposit was a safeguard in the event of insurance companies needing funds to pay claims. At that time HIH America had an excess of US$22.5 million on deposit. The department allowed it to withdraw US$1 0 million of that amount. 487 HII-1 America subsequently sought to withdraw the remainder of the funds .

• I-III-I America had ceded a portion of its workers compensation business to Reliance Insurance Company under the terms of a reinsurance contract. Reliance was fronting for Unicover, a third party manager of reinsurance pools. There were allegations of misrepresentation and non-disclosure by Unicover, and Reliance claimed damages and specific performance of the contracts. Re liance had entered into settlement discussions by October 1999. 488

• HIH America had entered into a reinsurance-pooling arrangement with its affiliates HIH Hawaii, HIH Illinois and Great States. The agreement provided

78 Unprofitable international operations

that net premium written in 1999 and beyond would be pooled and reallocated on a set of percentages agreed to by the Department oflnsurance.489

HIH America's response to the department On 17 February 2000 HIH America wrote to the Department of Insurance raising a number of concerns and seeking corrections to the department's report. 490 It claimed the report was flawed because it included developments during 1998 and 1999, without a consideration of the reinsurance contracts HIH America had entered into. In its response, HIH America made the following points:

• It asserted that the statement in the department's report regarding the policyholder surplus was misleading. It claimed the policyholder surplus was correct at the date of the examination, based on the data available at that date and was supported by the opinions of Andersen and Milliman & Robertson.49 1

• It agreed that the way in which the three reinsurance contracts were recorded was incorrect and that they were retrospective in nature and should have been accounted for under the guidelines specified in the accounting standards. 492

• It stated that at 31 December 1998 its deposit reflected US$30 million in reinsurance credits recoverable from General Reinsurance Corporation and Zurich Re under separate aggregate excess-of-loss contracts. 493 It was prepared to accept the department's position on how to account for the contracts, although it did not consider this was relevant to the deposit. 494

Conclusion The 1999 Department of Insurance report was the product of an investigation by the Californian regulatory body. It was an important document that should have been made available to the HIH board, regardless of what US management thought about

its accuracy. It would have been of great interest to the board in considering the viability of the US operations. The board relied upon US management to inform it of developments in the US businesses. The non-executive directors of HIH were not told of the report's existence. Again, this is a matter discussed in Chapter 15.

As part of the arrangements for Alaska to take over some of its business, HIH America made the necessary application to the Department of Insurance. In October 2000 the department allowed the business to be placed into run-off and HIH America ceased to write any new business.495

The department's triennial report, completed as at 31 December 2000, recorded a loss and loss-adjustment expense figure of US$81.3 million for the US group. The increase in loss and loss-adjustment expense reserves reflected continuing adverse loss development in the US operations' 1999 and previous accident years. 496 The department found that HIH America was insolvent under the definition of the Californian Insurance Code. 497 In late March 2001 it took over what remained of HIH America. 498

Th e failure of HIH Insurance 79

13.2.9 The ultimate deficien cy

As at 30 September 2002 the estimated deficiency of the US operations of HIH, as assessed by the California Insurance Guarantee Association, was US$31 0 980 031 for HIH America499 and US$19 998 972 for Great States. 500

A summary of the financial reports of HIH for the period 31 December 1997 to 31 December 2000 501 indicated, in respect of the US operations, a core earnings loss of A$135.8 million. 502

As at the date of liquidation, the KPMG actuaries assessed the liabilities of HIH's US operations at A$613 million, together with a required prudential margin of A$80 million. This is in contrast to the amount of A$387 million that appeared in HIH' s books for the US operations as at 31 December 2000.503

In December 2002 the KPMG actuaries provided a fresh overview of its actuarial assessment504 and confirmed there had been a significant but unquantified deterioration in the Californian workers compensation book. As a result, the claims liabilities for the US operations were likely to be substantially more than A$600 million. 505

13.3 Operations in Hong Kong and elsewhere in Asia

As with its other international operations, HIH incurred losses-albeit more modest-as a result of its operations in Hong Kong and elsewhere in Asia. These operations further demonstrate the problems that arose from HIH's strategy of expansion into unfamiliar markets and lines of business. Inadequate attention was given to the problems that were encountered, and there was a lack of reaction to the losses that were incurred.

HIH envisaged that its Hong Kong operations would provide a platform for expansion throughout Asia. 506 Several acquisitions and joint ventures were considered and pursued in the region between 1995 and 2001. 507

HIH's efforts to establish a broad Asian operation were made against a backdrop of consistently critical internal audit reports. These reports focused particularly on the lack of internal controls and the unreliability of reported results. There was little critical analysis or testing of the company's expansion programme or its acquisitions. The financial consequence of carrying on business in this fashion was ultimately a $235 million excess of liabilities over assets, as assessed by the provisional liquidators in 2002.

13.3.1 Group stru ctu re

HIH's Hong Kong operations, acquired in 1986, consisted of a portfolio spread over a number of countries in the region, primarily Thailand, Malaysia and the Philippines. 508 HIH saw Asia as a market suited to expansion and contemplated

80 Unprofitable international operations

several acquisitions and joint ventures between 1995 and 200 I. It pursued business opportunities in China, Thailand, the Philippines, Malaysia, Guam, Singapore, Taipei, and even as far afield as Kaliningrad in Russia.509

In March 1999 HIH Holdings (Asia) Limited acquired Forex Insurance Company Limited, which was subsequently renamed HIH Insurance (Asia) Limited.510 Upon the sale of HIH Holdings (Asia) Limited' s subsidiaries to HIH lnsurance (Asia) Limited, HIH' s Asian operations were structured as shown in Figure 13 .1.

Figure 13.1 HIH's Asian operations

I

HIH Overseas Holdings

I Limited •

I

HIH Holdings (Asia) Limited I

• HIH Insurance (Asia) Limited (formerly Forex Insurance Comoanv Limited) I • + + + + + HIH Casualty FAI First Asian Area HIH Forex HIH & General Pacific Reinsurance Underwriting Insurance Management Insurance Insurance Co. Limited Services Company (Asia) Limited (Asia) Limited Company (Asia) Limited (Macau) Limited Limited 13.3.2 Operations HIH' s Asian operations, including potential and actual acqmsthons and joint ventures, were described in various annual reports of the compan/11 , as well as in the 1998 offering memorandum at the time of Winterthur's sell-down. 51 2 The operations were invariably described in positive terms, with an optimistic assessment of the potential that expansion in Asia represented. There was minimal disclosure of the problems experienced by the Asian subsidiaries, as identified in internal audit reports and other business and accounting records of HIH. Expansion in the region was pursued for most of the period between 1998 and late 2000, when it was decided that the Asian operations should be sold. The justification for the decision to sell was that the company had ' different priorities for its capital'.513 This was despite the favourable assessment of the Asian operations contained in the 2000 annual report.514 Th e failure ofHIH In surance 81

13.3.3 Internal audit reports

As with elsewhere throughout HIH's operations, Waters prepared internal audit reports periodically. Internal audit activity reports were presented to the audit committee half-yearly in relation to the Hong Kong operations.

The reports relating to the Hong Kong business during the period 1998 to 2000 revealed a number of corporate governance deficiencies that were not adequately remedied by management. It was the practice for all directors to attend audit committee meetings, so all directors were aware of the problems experienced in the Hong Kong operations.

Internal audits: 1998 onwards An internal audit report in August 1998 warned that HIH was attempting to expand its Hong Kong business in order to obtain the necessary 'critical mass' in an extremely difficult environment. That environment had affected the performance of the Hong Kong operations, with actual results falling short of budget and deteriorating underwriting results and claims development. Waters described the operations as being ' at an operational precipice'. 515

The May 1999 internal audit report also identified a number of concerns, among them doubts about the adequacy of back-office infrastructure to support the business ' s growth, the need to adhere to underwriting strategies, and the need to establish appropriate levels of management controls to ensure self-management going forward. 516

In late 1999 and early 2000 the internal auditor advised that urgent consideration should be given to redressing these deficiencies. Warnings were sounded about the lack of financial integrity in the reported results. 517 Substantial variance was noted between results reported by the branch and those used for the group consolidation. Generally, the operations were characterised by an undisciplined pursuit of growth without due regard to profitability and probity. 518

The failure to respond to damning internal audit reports was said to be an indictment of Hong Kong management, which did not acknowledge the deficiencies identified. The problems arising from management's inaction were exacerbated by the continued financial deterioration of the Hong Kong branch.

5 19

The Hong Kong operations' expansion activities were said to be diverse and substantial and would not provide a return to the HIH group for some period. The need to maintain financial control over the regional activities was emphasised. 520

One internal audit report also disclosed the Hong Kong operations' entry into financial reinsurance contracts with General Cologne Re. These contracts were executed on 28 and 29 June 2000 and provided an immediate $20 million boost to HIH Insurance (Asia) Limited's profit-and-loss account for the year ending 30 June 2000. No risk was transferred to the reinsurer.521

82 Unprofitable international operations

Discussion notes prepared by the internal audit department in November 2000 stated that profitability for HIH Insurance (Asia) and HIH C&G (Asia) continued to deteriorate. 522 The integrity of the financial results was again raised as a concern. When the Insurance Commissioner for Hong Kong imposed a HK$300 million cap on premium production for HIH Insurance (Asia), management's response was to reallocate premium from that company to HIH C&G (Asia). 523

Throughout this period the internal audit reports on the Asian operations of HIH repeatedly highlighted concerns about the inadequacy of controls over the business and the unreliability of the financial statements.

13.3.4 Proposed initial public offering or sale

During the course of the FAI takeover, Abbott and Wein formulated a strategy for the float of HIH's Asian businesses. New insurance companies were to be acquired and fresh equity was to be introduced by outsiders. This strategy stands in contrast to the growth plans HIH had elsewhere, which involved wholly-owned subsidiaries of HIH. In the new group there were to be businesses in each of the principal countries in East Asia. 524

The proposed strategy During the first half of 1999 Abbott and Wein promoted the Asian strategy. A network of Asian operations was to be built through a process of acquisition. 525 Abbott was engaged as a consultant for the project, through his company Ashkirk Pty Limited. 526 The remuneration terms of this consultancy and the contingency fee arrangement are described in Chapter 23.

At a meeting ofHIH directors in June 1999, Wein was authorised to proceed with a project to raise capital to fund the development of a major insurance operation in Asia. 527 Following the meeting, Merrill Lynch was approached with a view to becoming involved in an underwriting or book-build exercise to raise the necessary capital. It advised that US$1 00 million was a suitable level of capital to be raised.

The proposal was advanced over the following 12 months, with consideration being given at one point to floating the US and UK operations in combination with the Asian business. 528 In May 2000 an information memorandum was issued, offering to sell HIH Asia through a three-phase process that included investors subscribing capital and the company growing to a value of no less than US$1 billion, at which

point there would be a 25 per cent float. 529 The memorandum contained very little analysis of the actual financial position of HIH Asia or its historical results; instead, the focus was on the potential of the business.

A change in the proposal Between June and October 2000 the proposal for restructure changed to one involving all of HIH- not just the Asian operations. A recapitalisation proposal from KoreaOnline was considered as a possible alternative to placing the personal

lines of business into a joint-venture vehicle with Allianz Australia Insurance

The failure ofHJH Insurance 83

Limited. The KoreaOnline proposal was the subject of several amendments at the behest ofHIH. 530 Ultimately, it involved Asia Insurance Partners (AlP) subscribing for a maximum of 65.4 million shares in HIH, with an agreement to underwrite a rights issue so as to raise a further $300 million. 53 1 The final proposal was presented to the board meeting on 12 September 2000, at which the Allianz transaction was approved, despite strident opposition from Abbott and Wein. 532 This is described in greater detail in Chapter 1 7.

After the board of HIH had approved the Allianz transaction it pursued a proposal for the sale of the Asian businesses. The main potential bidders who emerged were Russell Miller Asian Advisers, KoreaOnline and Wahfu. 533

In December 2000 another information memorandum relating to the prospective sale was issued. A three-phase process was proposed, involving the acquisition of 80-100 per cent of HIH Asia, followed by an injection of US$20-50 million of new capital, and then the listing of HIH Asia on either the Singaporean or Hong Kong stock exchange. 534

HIH was having difficulty selling its Asian operations. Hong Kong investment banks were reluctant to canvass the market for prospective bidders. 535 W ein blamed the run-off ofF AI First Pacific Insurance Company Limited and the unfavourable claims experience of the construction book in HIH C&G (Asia) for some of the problems encountered. 536

After reviewing HIH Asia for about two months, Russell Miller advised Abbott in early December 2000 that it would not proceed with its bid. The consensus among its lawyers, actuaries and accountants was that the shortcomings identified in the business made the investment too risky. 537

No agreement for the sale ·of HIH Asia was reached before HIH went into liquidation on 15 March 2001. The Asian operations' asset deficiency of A$234.5 million and its external administration have eliminated the possibility of any further sale of the business. 538

13.3.5 HIH investments China

The HIH excursion into the People's Republic of China is a case study of HIH's lack of control or critical analysis in its overall expansion in Asia. HIH sought to establish a joint-venture operation in the People's Republic of China through a domestic insurer, Yong An Insurance Company. It is clear that the lure of the Chinese market, at a time when all of Asia was enjoying economic prosperity and expansion, prompted HIH's move into the area.

The proposed joint venture Under the joint venture, HIH would provide information technology and technical insurance services to Y ong An. 539 The Chinese promoter and joint venturer in the project was Wang Jiancai. The venture was a failure. HIH managed to recoup its capital investment when the Chinese banking authorities were appointed

84 Unprofitable international operations

administrator of Yong An, but it had incurred costs of some $7.5 million through its participation in the enterprise and these were not recouped. The loss was subsequently capitalised in the HIH accounts. The basis for treating those costs as a capital asset was dubious to say the least.

The investment in Y ong An was made through a company incorporated in the Cayman Islands. HIH' s legal advisers questioned the use of this company because of the passive nature of its investment. No response to this qu ery has been identified. 540

The investment in Y ong An occurred pursuant to a resolution taken by the HIH board on 6 March 1996. It was noted that Winterthur supported the Yong An strategy, despite the commercial risks involved.541

Difficulties encountered with the joint venture During the course of negotiating the joint venture many amendments were made to its terms. The ultimate structure and nature of the operation were significantly different from that approved by the board.542 Furthermore, the investment was approved and executed without any due diligence being performed by HIH. Indeed,

when Wang became suspicious that HIH was investigating him, Williams reassured him that HIH had no knowledge of, and had not authorised, any such

investigation. 543

Instead of performing any substantial due diligence investigations, HIH relied upon the perceived personal integrity of Wang: his attitude, influence and performance were thought to be critical to the success of the joint venture.544 It seems that the board' s perception of him was coloured by his ability to arrange luxury vehicles accompanied by military escorts to transport the HIH directors and to organise state

banquets involving senior Chinese military personnel when the directors were visiting China. 545

From the outset the joint venture suffered from operational difficulties. Despite th e original intention, HIH was excluded from contributing its expertise to the operations. 546 Anecdotal evidence suggests that Y ong An had neither the systems nor the insurance accounting expertise to complete the basic tasks necessary to

. 547

operate an msurance company.

HIH made requests of Wang to remedy the operational and management deficiencies, but these were not remedied to HIH's satisfaction. 548 Because of the multitude of problems encountered and the failure over time to act to redress them, Williams arranged a meeting with Wang in Hong Kong in October 1997. At that

meeting he terminated HIH' s involvement in the operation and asked that its US $13.6 million capital contribution be repaid. 549

An HIH memorandum on the Yong An project detailed various allegations of misfeasance by Wang, whom HIH ultimately came to view as a fraudster. 550

The failure of HIH Insurance 85

Disclosure in the 1997 annual report

HIH's annual report and financial statements for the period ending 31 December 1997 referred to the investment in the People's Republic of China in the following terms:

The company continued to pursue a range of initiatives in China, although some difficulties were experienced with our joint venture in Xian . Our relatively modest current investment in China is an extremely worthwhile deposit for the future .55 1

This disclosure of the difficulties with Yong An was inaccurate and misleading: it did not provide shareholders with an accurate representation of the extent of the problems experienced.

The annual report was approved by a resolution on 4 March 1998, at an HIH board meeting that had previously been adjourned. Upon reconvening, Cohen, Terence Cassidy, Fodera, Stitt, Sturesteps and Williams were present. 552 Neville Head, Cohen, Stitt, Cassidy, Fodera and Alexander Gorrie were present at the audit committee meeting held two days beforehand, when the annual report was approved subject to minor amendments. 553 The directors present at those meetings had attended previous board meetings at which the problems with the Y ong An project had been discussed. Having been present at those meetings, and knowing the position with respect to Y ong An had not been resolved, those directors should not have approved the reference to the joint venture in the annual report.

The decision to capitalise the $7.5 million in expenses incurred through the Yong An project caused some consternation for HIH's auditor. At the audit committee meeting held in February 1998, Andersen questioned the appropriateness of carrying forward this expenditure as an asset in the books. 554 Management argued that what had been learnt would be useful for future activities in China and that, rather than being wasted, the expenditure was a deposit for the future. The audit committee preferred management's position that the expenditure should be capitalised in the accounts; Andersen held the contrary view. 55 5 Ultimately, Andersen acceded to management's view despite the absence of any specific benefit flowing from the expenditure.

13.3.6 Provisional liquidation

Provisional managers were appointed to HIH Insurance (Asia) Limited, HIH C&G (Asia) Limited and FAI First Pacific Insurance Company Limited on 2 April 2001. Each company was placed into provisional liquidation on 19 April2001. 556

13.3.7 The financial position ..

Table 13.1 shows the provisional liquidators' estimation of the financial position of the three main subsidiaries ofHIH in Hong Kong as at 31 August 2002.

86 Unprofitable international operations

Table 13.1 Hong Kong operations: estimated financial position, 31 August 2002

Company HK$m AU$m

HIH Insurance (Asia) Limited 557

470.9 109.1

HIH Casualty and General (Asia) Limited 55 8

376.2 87.2

FAI First Pacific Insurance Company Limited 559

164.9 38.2

Total 1012.0 234.5

Note: A$1 equals HK$4 .3 149, as at 30 August 2002 .

Before the provisional liquidators were appointed, the financial statements of these companies reported a position substantially more favourable than that estimated by the provisional liquidators. The main reasons for the large deficiency were the write­ down in the value of HIH' s investment in subsidiaries, inter-company indebtedness, and adjustments made to the provision for outstanding claims. 560 Table 13.2 shows

the variation in the provision for outstanding claims.

Table 13.2 Hong Kong operations: provision for outstanding claims, March-April 2001 and 31 August 2002

Provisional liquidators'

Financial statements Directors' statement estimate of financial Company as at 31.03.01 of affairs at 02.04.01 position at 31.08.02

A$ HK$ A$ HK$ A$ HK$

HIH Asia 561 80.2 346.0 80.5 347.2 101 .3 437.2

HIH C&G 562 38.6 166.6 74.0 319.3 95.4 411 .8

FAI First Pacific563 24.0 103.6 29.4 127.0 39.6 170.7

PricewaterhouseCoopers Actuarial Pty Ltd conducted for the provisional liquidators a further review of the provision for outstanding claims for the HIH Hong Kong group as at 31 December 2000; the results were published on 18 June 2002. Table 13.3 summarises the results.

Table 13.3 Hong Kong operations: provision for outstanding claims, 31 December 2002 564

AdditionaiiBNR

Gross recommended Estimated

reserve per March 2001 Revised gross reserves

accounts (Crespoint) reserves (PWC) Difference

Company (A$'000) (A$'000) (A$'000) (A$'000) (A$'000)

HIAAsia 66 330 17740 84170 92 268 8 098

HIH C&G 27 257 11143 38 400 74 595 36 195

FAI First Pacific 23 524 0 23 524 28 711 5187

Total 117 211 28 883 146 094 195 573 49 479

Th e f ailure of HIH Insurance 87

13.4 Conclusion

The losses stemming from HIH' s various international operations share a common pattern-expansion with little regard to its financial consequences, bad decision making and business judgment, adverse insurance market conditions, and a failure to react adequately to contain and minimise the losses that consequently flowed.

The UK operations were a major contributor to the collapse of HIH. Section 13 .1 reveals a number of reasons for the operations ' lack of success. There was an obvious lack of knowledge about the new areas of business HIH entered. There was repeated failure to control properly the underwriters' activities. There was rapid growth and further expansion into a market that failed to harden. Finally, there was a failure to comprehend the financial reinsurance arrangements to which Cotesworth was party and that rendered this reinsurance an illusory asset when Cotesworth went into liquidation.

Some of these criticisms are equally applicable to HIH's operations in the United States and Asia. After inadequate evaluation HIH re-entered the United States, moving into an adverse workers compensation market that, despite HIH's continued optimism, failed to improve. HIH was repeatedly warned to improve its controls over the US business, but it ignored the warnings and took inadequate steps to resolve the problems that led to losses.

Similarly, HIH' s ill-fated venture into China, although not resulting in a major loss in itself, illustrates the inadequacy of HIH's management controls and a failure to take proper steps to protect the business of the company when moving into areas in which it had limited experience. It is also indicative of HIH's overall strategy of expansion- promoted principally by Williams. That strategy proved disastrous in most of the international markets to which it was applied.

In essence, Williams's thirst for continual expansion, both internationally and in Australia, was a primary cause of the collapse ofHIH.

4

88

United Kingdom: approximately A$1.7 billion (WITS .0295.001 at 009 par. 32). United States: approximately A$600 million (CIGA.0001.002 at 004; WITS.0296.001 at 004 to 005 par. 13(f)). Hong Kong - Asia: approximately A$200 million (HIHK.0005 .001 at 018 to 020).

WITS .0295 .001 at 009 to 010 par. 32.

WITS.0165 .001A at 021A pars 68 and 69.

WITS.0165 .001A at 022A pars 77 to 80.

WITS .O 176.001 at 003 par. 4. WITS .0165 .001A at 023A to 024A pars 82 to 84.

Unprofitable international operations

WITS.0165.001A at 026A par. 95; WITS.0156.001A at 016A par. 2.3.

T14153/54 to Tl4154/4; WITS.0156.001A at 028A par. 104.

WITS.Ol56.001A at 030A par. 3; WITS .0165.001A at 028A pars 104 to 107 ; Tl4154/24 to Tl4154/30. 10 WITS.0129.001 at 007 par. 37. II

WITS.OOOI .001 at 003 to 004 pars 7(a), 7(b) and at 011 pars 39 to 41. 12 WITS.0283.001 at 010 par. 49 and at 022 par. 108. 13

WITS.0165.001A at 036A par. 134. 14 BRD.002.100 at 109 to 111. 15

WITS.0165.001A at 034A par. 125; WITS.0040.001 at 015 par. C1.2; Tl2042/39 to Tl2042/53; WITS.0129.001 at 007 par. 37; T9707/48 to T9708/26; WITS.0121.001 at 021 par. 103. 16

WITS.0040.001 at 016 pars C1.7, Cl.8 and C1.1 0; WITS.0176.001 at 026 pars 50 and 51. 17 WITS.0153.517 at 521; SBB.019.196_001 at 196_005; see also WITS.Ol53.015 at 049

par. 129. 18 The Winterthur reports are at SBB.020.574_001, SBB.020.575_001, SBB.020.576 _ 001, SBB.020.577 _ 001, SBB.020.578 _ 001, SBB.020.579 _ 001 ;

WITS.Ol65.001A at 097A par. 345. 19 WITS.0165.001A at 097A to 098A par. 346; WITS.0153.547; see also WITS.Ol53.015 at 050 to 051 par. 134. 20

WITS.0153 .589; see also WITS.0153 .015 at 051 par. 135. 2 1 WITS.0153.609 at 612 to 613; ROY.0098.0031 at 0034; see also WITS.0153 .015 (annexure B) at 052 to 053 par. 142. 22

SBB.020.573 001. 23 SBB.020.573_001 at 573_011 to 573_012; SBB.063.324_001; T9712/8 to T9712/29; T9716/3 to T9716/39. 24

SBB.086.578 001 at 578 002 . - -25 T9718/4 to T9718116. 26 PA YN.0012.001 ; SBB.063.324_001; WITS.0165.001A at 098A par. 349; T14223; WITS.OJ56.001A at 165A par. 9. 27 AUDC.007.001 at 003; AUDC.008.101 at 104. 28 WITS.0165.001A at 099A par. 351. 29 ROY.0098.0124; PAYN .0005.022; see also T9721 to T9724. 30 ROY.0098.0124 at 0128. 3 1 ROY.0098.0124 at 0129 to 0130. 32 ROY.0098.0 124 at 0133 . 33 ROY.0098.0124 at 0127. 34 AUDC.009.001 at 003; BRD.026.050 at 060; AUDC.009.091 at 101; WITS.0168.001 at 065 pars 187 to 190. The failure of HI H Insurance 89

35

SBB.021.152 001. 36 Waters gave evidence that, to be described as 'bad', it had to have very significant problems; T9727/ 13 to T2927117. 37

SBB.021.153_001 at 153_003; PAYN.0004.015 at 015; T14233 to T14234. 38 WITS .Ol65 .001A at 099A par. 353. 39

PA YN.0004.007; WITS.0165.001A at 100A par. 354 (Payne). 40 PA YN:0004.008; WITS.0165 .001A at 100A par. 355; T14230/21 to T14232/33. 41

PA YN.0004.007. 42 T11090/37 to T11090/41. 43

PAYN.0004.007; PAYN.0004.008. 44 T11239/ 19 to T11239/28; WITS.0165 .001A at 100A par. 357; T\1243/56 to Tl1244/ 12. 45

PA YN.0004.001; Tll091/32 to Tl1091/45; WITS .0165.001A at IOOA par. 357. 46 WITS.0165.001A at 100A par. 357; T14234. 47

CIV.001.999 160at999 210and999 178. - - -48 AUDC.OI0.001 at 003. 49 AUDC.010.083 at 085 and 087. 50 SB0.005.224 001 and SB0.005.225 001. - -51 HIUK.0006.036; RIDD.0003.130 at 133. 52 SB0.005.223 001. 53 RIDD.0003.126 at 127. 54 SB0.005.224_001 at 224_002, 224_004, 224_005, 224_006, 224_009 and 224_014 (in particular). 55 SB0.005.225 001 at 225 002. - -56 SB0.005.225 001 at 225 004. - -57 SB0.005.225 001 at 225 014. - -58 PA YN.OOOI.086; see also T11240/28 to Tl1241 /34. 59 T 11 094/36 to Tl1 095/7. 60 RIDD.0003 .052 at 052. 61 RJDD.0003.088; RIDD.0003.095 at 098 to 099; Tl1114/16 to T11114/26. 62 SBB.015.866 001 at 866 002. - -63 AUDC.012.001 at 003; AUDC.012.138 at 141. 64 RIDD.0003.126 at 127 to 128 . 65 WITS.0129.001 at 005 pars 24 and 25. 66 SB0.005.253 001 A at 253 003A and 253 025A. - - -67 T9736/40 to T9736/42 . 68 AUDC.012.001 at 003; AUDC.012.138 at 140 and 143 . 69 WITS.0129.001 at 005 to 006 pars 25,26 and 27 . 90 Unprofitable international operations

70

RIDD.0003.033 at 034 . 71 SBB .021.156 OOIA at 156 OOIA. - -72 SBB.021.156_001A at 156_003A; W1TS.Ol68 .001 at 112 to 113 par. 346. 73 SBB.021.156_00IA at 156_007A; T9737/46 to T9737/54. 74 AUDC.Ol3 .001 at 002; AUDC.Ol3 .106 at 109 ; BRD.048.166 at 169 . 75 RIDD.0003 .019 at 019 to 020 ; T14273 /17 to Tl4273/36. 76 WITS.OI79.001 at 067 par. 317; WITS.OI79.907 at 972 and 979 to 980. 77 T2 22. 78 WITS.0165 .001A at 075A to 076A par. 282; WITS.OOOl.OOl at 004 par. 7(f) (Riddell). 79 RIDD.0003 .065 at 068 . 80 RIDD.0003 .007 at 007; Tl4283/33 to Tl4284/ 12 ; RIDD .0003 .065 at 065. 81 TI3487/37 to TI3488/2. 82 RIDD.0003 .007 at 008 . 83 SBA.194.092 001 at 092 002. - -84 SBA.l94.092 00 I at 092 002 and 092 003. - - -85 SBB.021.158_00IA at 158_003A; T9740/35 to T9740/45. 86 SBA.194.092 001 at 092 003 . - -87 TI4291135 to Tl4292/03 . 88 TI1132/ l 0 to Tlll32/29 . 89 SBA.l74.522 001. 90 TI2065/51 to T12066/ ll. 91 RIDD.0003.062 at 063 ; see also SBA.I94.094_001. 92 T9739/51 to T9739/53 . 93 SBB .021.158 OOIA. 94 T974119 to T9741/10. 95 SBB.021.158 OOIA at 158 007A. - -96 AUDC.OI6.001 at 002; AUDC.Ol6.049 at 053. 97 SBB.021.159_001; WITS.0129.001 at 007 pars 43 and 44; T9745/39 to T9745/50 . 98 SBA.209.539 001 at 539 002 . - -99 AP.0002.0010.0027 at 0037 and 0039; see also Tl2093/41 to Tl2093/54 . 100 Tl3493/6 to Tl3494/33. 10 1 RIDD.0003 .062 at 062 to 063 . 102 SBA.209.539 001. 103 RIDD.0003 .058 at 058 and 060; HI.0016.0041.0020. 104 SBB.044 .806 001 at 806 003 and 806 005. - - -10 5 WITS.0040.001 at 020 par. Cl.29. 106 AUDC.Ol7 .031 at 033 ; AUDC.Ol7.001 at 001 to 002 . 107 T12113/53toT12114/02 . Th e failure ofHJH Insurance 91

108

APRA.0102.0077 at 0110,0114 and 0116. 109 BRD.062.000 at 000 001. 110

BRD.064.000 at 000 003 . 111 WITS.Ol79.001 at068to069par.325. 112

WITS.O 179 .999 _104; WITS.Ol79.001 at 069, par. 327 (Abbott). 113 WITS .OI79 .999 _ 107 ; WITS.Ol79.999 _112; WITS.0 179.999 _ 113 ; WITS.O 179.00 1 at 069 par. 328. 114

Tl2115/31 toT12115/49. 115 WITS.Ol89.001 at 168 to 172 pars 572.1 to 572.6. 116

WITS.OOO 1.001 at 017 par. 69; and at 021 to 022 pars 83 to 90. 117 WITS.OI65.001A at 035A par. 129. 118

WITS.O 165 .00 lA at 035A par. 129; and at 052A to 054A pars 197 to 206. 11 9 WITS.Ol65.001A at 035A par. 130; and at 055A to 057A pars 211 to 2 19. 120

WITS.0165.001A at 051A, par. 195. 121 WITS .OI65 .001A at 048A to 064A pars 179 to 245; WITS.0040.001 at 039 to 040 pars C8.7 to C8.13 . 122

SBB.020.573 001 at 573 011 and 573 013 . - - -123 SBB.020.573 001 at 573 042. - -124 ROY.0098.0124 at 0129. 125 AARA.0273.00 10 at 0021. 126 SBB .Ol6.274 001 at 274 013 . - -127 AARA.033 7.0 139; SBB.003.837_001. 128 BRD.029.019 at 029. 129 BRD.036.032 at 041. 130 RIDD .0003 .076 at 079. 13 1 SBA.l74.522 001. 13 2 WITS .OOOI.OOI at 021 pars 83 to 86. 133 WITS .OOOl .OOI at 022 par. 89. 134 WITS .0001.041 at 042; T237/37 to T238/54. 135 WITS.OI75.001 at 001 to 007 pars A, Band D; WITS.Ol63.001 at 004 par. C. 136 BRD.017.061 ; WITS.Ol53 .655; WITS.0153.015 at 054 par. 148. 137 BRD.017.060; WITS .Ol53.655; WITS.0153.015 at 054 par. 148 . 138 BRD.017.000 at 000_003; WITS.Ol53.015 at 054 pars 149 to 150. 139 WITS.Ol53 .594. 140 WITS.0153.596. 141 BRD.Ol8.001 at 003; WITS.Ol53.015 ai 055 par. 153. 142 BRD.Ol8.001 at 003; BRD.020.000 at 000_002. 143 T I II 11 /3 2 to Till 11 /3 7. 92 Unprofitable international operations

144

HIUK.0004.110 at 110. 145 SBA.202.360 002 at 360 003. - -146

WITS.O 175 .001 at 006 to 007 par. D; Tl1111 /7 to Tl1111/20 ; RIDD.0003 .007 at 00 8; SBA.J94.092_00 l at 092_002 ; Tl4290/33 to Tl4290/52 ; WITS.O 175 .0 I 0; WITS.O 175.011. 147

RIDD.0003 .007 at 008. 148 SBA.194.092_001 at 092_002; Tl4290/33 to Tl4290/52 . 149

WITS.0175 .001 at 007 par. D. 150 WITS.0153.684; WITS .0153.015 at 055 par. 154. 151

WITS.Ol53.685 ; WITS.Ol53.015 at 055 par. 155 . 152 WITS.Ol53.691 ; WITS .Ol53.015 at 055 to 056 par. 156. 15 3

WITS.0153.692 ; WITS.0153.015 at 056 par. 157 . 154 WITS.0176.001 at 028 to 029 par. 56. 155

WITS.0165 .001A at 064A to 065A par. 248 . 156 WITS.0165.001A at 050A par. 190; Tl4237. 157

PAYN.0015.020at020. 15 8 Tl1229/44 to Tl1229/48. 159

WITS .Ol65.001A at 050A par. 192 . 160 WlTS.0165 .001A at 118A par. 427 . 16 1

WITS.0165 .001A at 118A par. 428 ; BRD.059.000 at 000_005 . 162 WITS.0 165.132; WITS.0165.00 1A at 031 A pars 115 and 116 . 163

WITS.OOOl.OOI at 018 par. 70. 164 WITS.0001.001 at 017 par. 69 . 165

WITS .OOO 1.001 at 018 par. 74. 166 WITS .OOOI.001 at 019 par. 76. 167

WITS.0001.001 at 019 par. 78 . 168 WlTS.0295.001 at 005 to 006 par. 19. 169

WITS .0295.001 at 006 par. 20. 170 WITS.0165.00 1A at 067A par. 252 . 17 1

WITS.Ol65 .001A at 067A par. 252 ; T14251/28 to Tl4251/42; Tl427l/2 to T14271 / 12. 172 PAYN.0002.042; Tl4251/28 to T14251 /42. 173

SUBP.0005.001 at 023 par. 73. 174 SBA.202.360 002 at 360 006. - -175

TlllOl/52 to Tlll01/59; PA YN.0010.016 at 016; Tl4266/5 to Tl4266/56. 176 RIDD.0003.052 at 053 to 054. 177

WITS.0040.001 at 036 pars C7.9 and C7 .10. 178 RIDD.0003 .082 at 083 ; SBB.201.271 _001 ; Tllll7/58 to Tllll8/2; WITS.0040.001 at 036 par. 7.11.

Th e failure ofHIH Insurance 93

179

SBA.194.092 001 at 092 002. - -180 HIUK.0004.148 at 148. 181

RIDD.0003.005 at 005 to 006. 182 WITS.O 130 .001. 183

WITS .Ol30.001. 184 WITS .Ol30.001 at 001 to 002 ; T9793/ 15 to T9793 /20. 185

WITS.Ol30.001 at 002. 186 T98 04/35 to T9804/40. 187

WITS.Ol30.001 at 002; see also T9793/29 to T9793/45 , T9794/7. 188 Tl5051 /48 to Tl505 1151 ; see also Tl5050/36 to Tl5050/46. 189

Tl5052/4 to Tl5052/8. 190 Tl5050/48 to Tl505112. 191

T9808/27 to T9808/37 ; HIHP.0023.001. 192 Placement was in the form of a line slip, which covered a number of films known as a 'slate'. Each film within a slate would then be declared individually under the line slip

when production commenced: HIHP.0013 .010A. 193 HIHP.0013 .010A. For an illustration of the fronting arrangement, see WITS.Ol24.019. 194

HIHP.0025 .308 . 195 T9807/33 to T9807/53. 196

T15052/57 to T15053/6. 197 BRD.014.133 at 141. 198

BRD.014.133 at 592. 199 BRD.Ol4.133 at 594; T9735/2 to T9735/8. 200

T9794/27 to T9794/31 . 20 1 T9795/19 to T9795/25. 202

T9795/27 to T9795/34. 203 T9796/24 to T9796/29. 204

T9798/2 to T9798/ 12. 205 T9806/20 to T9806/22. 206

SUBP.0038.001 at 002 to 003 . 207 T9809/50 to T981 0/3. 208

TI5050/ 13 to Tl5050/20. 209 Tl5050/36 to Tl5051/2. 210

Tl5052/49 to Tl5052/52; Tl5055/ 17 to T15055/27. 21 1 T15053/27 to T 15055/31. 2 12

SUBP.0038.001 at 001. 213 Tl5053/37 to Tl5054/6. 2 14

SUBP.0038.001 at 001.

94 Unprofitable international operations

215

SUBP.0038.001 at 002. 216 T9803/34 to T9803/36 (Small), HIHP.0025.214 at 220 to 221. 217

T9803/38 to T9803/43 . 218 SUBP.0038.001 at 003 (Small). 219

Tl5054/37 to Tl5054/45 . 220 Tlll19/50 to Tll119/58. 221

WITS.OI77 .001 at 001 par. 4. 222 SBB.073 .938 001. 223

The lists are at SBB .073 .938_001 ; HIHP.0003 .184 ; SBB.023.954_001 ; SBB.073 .902_001 ; SBB.023 .937 _001 ; SBB.023.938_001 ; SBB.023 .921 _001 ; SBB .023.922_001 ; SBB.023 .923_001. 224

HIHP .0013 .010A. 225 HIHP.0013 .010A at 016A; see also WITS .0124.001 at 010 to 011 pars 40 to 43 . 226

HIHP.0013.079A; T9819/46 to T9819/49 . 227 T982 1/53toT982 1/57. 228

HIHP.0023.510; HIHP.0023.511 at 512. 229 Lureco is owned by General Electric Company of the United States, a publicly li sted company: WITS .0175 .001 at 007 par. E. 230

WITS.OI63.001 at 004 par. C. 231 WITS .01 75.001 at 006 par. D. 232

WITS.0175 .001 at 006 par. D; WITS .0163 .001 at 004 par. C. 233 WITS.0165.156 at 165 to 166. 23 4

WITS.0295.001 at 010 to 011 par. 36 . 235 WITS.0001.001 at 028 par. 120. 236

RIDD.0002.151 ; RIDD.0002.156. 237 WITS.0 175 .001 at 006 par. D. 238

WITS.OOOI.001 at 028 par. 118 ; WITS.0295.001 at 011 pars 37 and 38. 239 RIDD.0002 .157 . 240

Tl2022/50 to Tl2022/55 ; WITS.0189.001 at 109 par. 345 .10; CIY.OOI.999_ 160 at 999 17 1. 241 WITS .Ol63 .001 at 003 par. C. 242

BRD.027 .000 at 000 005 . 243 WITS.0156.001A at 065A par. 4. 244

WITS.0040.00 1 at 033 to 034 pars C6.3 and C6.4. 245 SBB. I75 .282_001 ; SBB.175 .283 _001. 246

ROY.0141.0015 . 247 WITS.0162.001 at 048 par. 6.3.2. 248

BRD.030.000 at 000 004. 249 ROY.01 4 1.0015.

Th e failure ofHIH Insurance 95

25 0

SBB.Ol 3.779 001. 25 1 BRD.031.000 at 000_002; BRD.03l.016 at 021. 252

SBB.l75.2 03 _001. Note that it had been signed by Thompson on 21 April1998: SBB.175 .212 00 I. 253 SBB .175 .205 001. 254

WITS.O 180.296. 255 WITS.0040.001 at 034 pars C6.5 and C6.6. 256

SBB.104.719 _001; SBB.l04. 720_001 ; ROY.0142.0137; ROY.0143 .0040; ROY.0144.0001. 257 BRD.033.000 at 000 001. 258

SBB.l04.712 001 at 712 002. - -259 SBB .1 04.709 001. 260 HI.0006.0002.01 85; RlDD.0003 .038 at 051. 26 1 WITS .0179.001 at067par.316. 262 WITS .OJ75 .001 at 004 par. D; SBB.000.069 _001. 263 SBB.000.069 001 . 264 RlDD.0003 .016 at 016. 265 WITS .O 175 .001 at 004, par. D; SBB .000.069 _ 00 I (Piper). 266 BRD.035 .033 ; BRD.035 .043 . 267 BRD.035.000 at 000 001 to 000 002 . - -268 RIDD.0003 .026. 269 SBB.105 .019 001. 270 This is discussed further in Section 13 .1.13 . 27 1 BRD.041 .000 at 000 002. 272 WITS .0165.001A at 073A to 074A pars 275 and 276; CMC0.0061.271 at 277. 273 WITS.0189.001 at 108 par. 345 .8; WITS .0165.001A at 074A to 075A par. 278; WITS .0003.001 at 008 to 009 par. 21(i). 274 WITS.0295 .001 at 008 to 009 pars 28 to 31 . 275 WITS .0307.02 7. 276 WITS .0307.004. 277 WITS.0307.012. 278 WITS.0307.014. 279 WITS.0307.017. 280 WITS.0307.023. 28 1 WITS.0307 .026. 282 WITS.0307.035. 283 WITS .0307.035 . 284 SBB.000.069 001. 96 Unprofitable international operations

285

WITS .0001.001 at 005 par. 9 and 006 par. 15 . 286 WITS.OOOl.OOl at 006 pars 13 and 14; WITS.0295.001 at 002 par. 5. 287

WITS.0295.001 at 002 par. 6. 288 BRD.041.000 at 000 002. 289

BRD.062.000 at 000_001; WITS.Ol68.001 at 165 par. 518. 290 This is discussed further in Chapter 17. 291

BRD.064.000 at 000 004. 292 BRD.064.000 at 000 003. 29 3

WITS.Ol65 .001Aat046Apar.l72. 294 WITS .OOOl.OOl at 009 par. 27 and 011 par. 37. 295

WITS.OOOl.OOl at 009 par. 28. 296 WITS.OOOl.OOl at 009 par. 30. 297

HIUK.0005.057. 298 WITS .OOO 1.001 at 009 to 010 par. 31; Tll128/6 to Tlll28/24. 299

SBA.l74.519 001 at 519 002. - -300 This is discussed in detail in Chapter 23. 301 WITS.OOOl.OOl at 010 par. 32 . 302 APRA.0047.0040 at 0040; APRA.0047.0042; APRA.0047.0043; APRA.0047 .0052 . 303 APRA.0047.0043 at 0047. 304 HIUK.0005 .026. 305 HIUK.0005.026 at 027 . 306 SBB.041.479 001. 307 WITS.Ol75.001 at 008 par. F. 308 RIDD.0003 .062 at 063 to 064. 309 SBB.051.015 001. 3 10 RIDD.0003.058 at 060. 311 WITS.OOOl.OOl at 011 par. 37; CMC0.0061.252 (footer). 3 12 CMC0.0061.257; CMC0.0061.271. 3 13 CMC0.0061.257 at 259 and CMC0.0061.271 at 273. 3 14 WITS .OOOI.OOl at029par.l26andat030par.l32. 3 15 WITS.Ol65.001A at 014A to 015A pars 39 to 46 and 48 and at 016A to 017A pars 50 and 51; WITS .Ol65 .127. 3 16 WITS.Ol56.001A at 169A par. 9; Tll909/27 to T11909/31. 3 17 WITS.Ol76.001 at 027 pars 52 to 54. 3 18 WITS.Ol89.001 at 118 to 119 pars 383 to 390. 3 19 WITS.Ol79.001 at 115 pars 532 to 534. 320 WITS.Ol68.001 at 216 par. 702. 32 1 WITS.0165.001A at lOlA par. 361. The failure ofHIH Insurance 97

322

WITS.0165 .001A at 103A par. 369; Tl4190/26 to Tl4190/46. 323 Tl4193/4 to Tl4193/9. 324

Tlll05/16 to Tll105/19. 325 WITS .0165 .001A at 075A par. 280. 326

T14193/38 to T14193/49. 327 WITS .0168.001 at 216 par. 702; WITS.Ol89.001 at 115 par. 378.10; WITS.OI65.001A at 018A par. 59; Tll921/28 to Tl1922/3. 328

WITS.Ol65.001A at 019A par. 63. 329 WITS .Ol65 .001A at 019A par. 63; Tl4208/34 to Tl4209/35. 33 0

Tl1258/33 to Tll258/37 ; Tl1919/35 to Tl1920/ 11. 33 1 Tl1921 /55 to Tl1922/3. 332

Tl4206/23 to T14206/31. 333 COHE.0003.001 ; COHE.0003 .002. 33 4

WITS.Ol68.001 at 218 par. 707. 335 COHE.0003 .002. 336

CIV.001.764 at 828 ; CIV.001.838 at 911; CIV.001.926 at 995 . 33 7 WITS .0189.001 at 117 pars 380 and 381; see also SUBP.0062.001 at 035 to 036 pars 136 to 145. 338

WITS.0179.001 at 115 par. 536; SUBP.0063 .001 at 058 to 059pars 255 to 264. 339 WITS .0156.001A at 169A; SUBP.0055 .001 at 112 to 119 pars 516 to 543 ; SUBP.0005 .001 at 030 to 031 pars 93 to 97. 340

Tl4354. 341 Using an exchange rate of $0.5500. CIGA.OOO 1.002 at 004 and 005 . 342

CIV.001.999 160 at 999 212. - - 343 WITS .0176.001 at 007 to 008 par. 13 and 068 par. 133. 344 WITS.01 76.001 at 007 par. 12. 345 WITS.0176.001 at 007 to 008 par. 13. 346 WITS.0157 .001 at 002 pars 3.2 and 4.1. 347 WITS.Ol76.001 at 013 par. 23; WITS.0155 .001 at 023 par. 96 . 348 WITS.0155.001 at 023 to 024 par. 96. 349 SB0.007.449 001 at 449 00 I. - -350 ROY .0077.0107 at 0114; ROY.0077.0129 at 0134. 35 1 ROY.0077.0129 at 0134. 35 2 CIV.OOI.999 160 at 999 212. - -353 BRD.002.020 at 033. 354 CIV.Ol5.001 at 004. 35 5 WITS .Ol57.001 at 002 pars 3.2 and 4.1 (Castro). 356 CIV.OI5.011 at 237 and 238. 98 Unprojilable inlernalional opera/ions

357

WITS.0176.001 at041 to042par. 79. 358 WITS .0153.695 at 695 . 359

BRD.020.040 at 042. 360 BRD.020.040 at 042 . 361

BRD.020.040 at 047. 362 BRD.020.040 at 047. 363

BRD.020.040 at 046. 364 T10902/33 to Tl0902/36, Tl5699113 to Tl5699119, Tl5094/52 to Tl5094/56, Tl2126/2 to Tl2126/30. 365

BRD.020.040 at 060. 366 BRD.020.040 at 046. 367

BRD.020.040 at 063 to 065 . 368 BRD.020.000 at 000 003. 369

BRD.020.030 at 030. 370 BRD.020.030 at 030. 37 1

BRD.020.030 at 030. 372 SBB.008.125 001at125 004. - -373

BRD.021.022. 374 Tl5098/2 to Tl5098/41 (Sturesteps ). 375

T15097/37 to Tl5098/41. 376 BRD .020.030 at 030 to 031. 377

BRD .021.022 at 022 to 023 . 378 BRD.021.007 at 009. 379

US$59 million: BRD.021.006; CIV.001.999 160 at 999 213. 380 BRD.021.010 at 011. 381

Tl4678/44 to Tl4678/57 , Tl441 7/45 to T14417/55, Tl5094/52 to Tl5096/55, T12118/16 to T12118/46, Tl5697116 to T15697/33. 382 Tl4678/47 , Tl4417/45 to Tl4417/55 , T15697/25 to Tl5697/33. 383

Tl1600/2 to Tl1600/44. 384 Tl2118116 to Tl2118/46. 385

T14679/30 to Tl4680/4, Tl5516/2 to Tl5518/24, Tl4418/59 to Tl4419/4, Tl5094/52 to Tl5095/28 , Tl5460/ 11 to Tl5461 /23. 386 BRD.020.040 at 044. 387

AARA.0338 .0219 at 0219. 388 US$59 million: BRD.021.006; CIV.001.999_ 160 at 999 213. 389

SBB .201.327_001 ; see also SBB .Ol3.790_001. 390 WITS .Ol57.001 at 002 to 004 pars 4.2 to 4.5. 39 1

CIV.016.010 at 011.

Th e failure ofHJH Insurance 99

392

BRD.028.029 at 030. 393 CIV.016.379 at 499. 394

BRAM.OO 11.03 7. 395 AND.l387.0031.0003 at0015. 396

BRAM.0011.037 at 037. 397 BRD.032 .006 at 006. 398

AND.1671.0001.0292A at 0294A. 399 ROY.0096.0008 . 400

AND.1671.0001.0292A at 0294A. 401 AND.1671.0001.0292A at 0294A. 402

ROY.0096.00 10. 403 ROY.0096.001 0. 404

ROY.0096.001 1. 405 T9771 /48 to T9772111; WITS.0228.001 at 004 par. 3.6. 406

Tl2175/ 11 toT12175115. 407 Tl2175/ 17 to Tl2175/26; Tl2667 /50 to Tl2667 /53. 408

SBA.062.497 _ 00 1A at 497 _ 007 A. Williams was appointed a director of Great States at the 16 June 1999 board meeting. 409 SUBP.0053 .00 1 at 032 pars 78 to 81. 4 10

ROY .0096.001 1 4 11 AUDC.O I4.001 at 002 . 412

AUDC.OI4 .055 at 062. 413 SBA.062.474 001 at 474 003 to 474 007. - - -414

WITS.0097.001 at 011 par. 64. 415 An example is at SBB.021.825_001A and BRD.026.150. 416

BRD.029.128 at 150 . 417 SBB.021.825 001A. 4 18

SBB.021.167 OOIA. 419 SBB.021.167 OOIA at 167 007A. - -420 SBB.021.167 OOIA at 167 008A. - -42 1 SBB.021.167_001A at !67_01IA to 167_013A. 422 WITS.0228.00 1 at 002 par. 2.3. 423 WATS.0018 .001. 424 BRAM.0003 .094 at 095 . 425 BRAM.0003 .092 . 426 HIH.0259.0309. 427 BRAM.0003.092; T9749/58 to T9750/3. 428 SBA.062.482 001 at 482 001. - -100 Unprofitable international operations

429

AUDC.012.138. 430 AUDC.012.001 at 003. 431

AUDC.012 .138 at 142 . 432 AUDC.012.138 at 142. 433

SBB .008.125 001 at 125 004. - -434 SBB.021.167 001A at 167 007A. - -435

SBB.021.168 001 A. 436 SBB.021.168 001A at 168 004A. - -437

SBB.021.168 001A at 168 009A. - -438 SBB.152.1 07 001 . 439

BRD.048.114 at 131. 440 BRD.048.114at117. 441

BRD.048.166; AUDC.013.096. 442 AUDC.013.001 at 002. 443

BRD.048.166at171. 444 SBB.002.261 001. 445

SBB.133.621 001 at621 002. 446 SBB.133 .621 001 at 621 004. The other locations were Illinois and Hawaii. - -447

SBB.021.170 001A at 170 003A. - -448 The HIH group consisted of California, Great States, Illinois, Hawaii and HIH Insurance Services. 449

SBB.021.170 001Aat 170 016A. - -450 SBB.021.170 001A at 170 007A. - -451

T9754/3 to T9754/15; SBB.021.170_001A at 170_007A. 452 T9754/45 to T9754/54; SBB.021.170_001A at 170_014A. 453

SBB.021.170 001A at 170 015A - -454 SBB.021.170_001A at 170_015A to 170_016A; T9755/16 to T9755/50. 455

AUDC.014.001 at 002 . 456 AUDC.014.055 at 057 . 457

AUDC.014.055 at 061 . 458 SBA.337.808 001 at 808 006. - -459

BRD.056.135 at 152. 460 AUDC.016.001 at 002. 461

AUDC.016.049 at 052 . 462 SBB.021.174 00 1A. 463

SBB.021.174 00 1A at 174 004A. - -464 SBB.021.174 001Aat 174 008A. - -465

BRD.059.071 at 086.

The failure of HJH Insurance 101

466

BRD.059.071 at 086 . This was the practice permitted HIH America by both Milliman & Robertson and Andersen. 467 As di scussed in Chapter 15 . 468

As discussed in Chapter 15. 469 AUDC.017.001 at 002. 470

BRD.059.134 at 138; AUDC.017.088 at 092. 471 AA.0003.0001.0037. 472

PA YN .0005.001 at 003 . 473 WITS.O 157 .001 at 014 par. 11.6. 474

PAYN.0005.001 at003. 475 Tl2154/46 to Tl2160/46; Tl2864/36 to Tl2865/31; T15706/2 to Tl5707/57. 476

HIHP.0030.00 1. 477 HIHP.0033 .001. 478

T11616110 to T11616/25. 479 SBB.055 .989 001. 480

BRD.061.000. 481 WITS.0176.001 at 044 to 046 par. 84; SBB.055 .989 _001. 482

SBA.187.865 001 at 865 004. - -483 SBA.187.865 001 at 865 028. - -484

SBA.187.865 001 at 865 005 . - -485 SBA.187.865 001 at 865 004. - -486

SBA.187 .8 65 001 at 865 006 and 865 029. - - -487 SBA.187.865 001 at 865 007. - -488

SBA.187.865_001 at 865_007 to 865_008. 489 SBA.187 .865 001 at865 008. - -490 HIHP.0003.228. 491 HIHP.0003 .228 at 229. 492 HIHP.0003.228 at 230. 493 HIHP.0003.228 at 230. 494 HIHP.0003.228 at 230. 495 WITS.O 157.001 at 014 par. 11.6. 496 EBSW.0001.051 at 065. 497 EBSW.0001.051 at 054. 498 WITS.0 157.001 at 014 par. 11.6. 499 CIGA.0001.002 at 004. 50° CIGA.0001.002 at 005 . 50 1 BRD.029.128 at 150; BRD.048.114 at 131 ; BRD.054.053 at 069; BRD.056.135 at 152 ; BRD.059.071 at 086; SBA.ll4.075_001 at 075_017. 102 Unprofitable international operations

so2 Because no earnings figure for the six months to 30 June 2000 was available this figure was calculated by deducting the loss for the six months ending 31 December 1999 from the loss for the year ending 30 June 2000. so

3

APRA.0027.0034 at 0039.

so 4

WITS.0296.001 at 004 to 005 par. 13(f).

sos No precise figure was given.

so 6

CIV.OOI.999 160 at 999 215. - -so 7

WITS.Ol76.001 at 016 par. 31 and 030 to 035 pars 60 to 68.

so 8

CIV.OOI.999 057 at 999 099 . - -so 9

WITS.O 176.00 I at 030 to 035 pars 60 to 68. sJo AND.5617.0081 at 0082 and 0092.

SJJ See, for example, CIV.OOI.838 at 848 (the 1999 annual report) and CIV.001.926 at 930

(the 2000 annual report). sJz CIV.001.999 160 at 999 215 . - -

s13 CIV.OO I.926 at 930.

s14 CIV.OOI.926 at 930. sJs AND.6030.0002 at 0005.

516

SBB.022.186 001 at 186 003. - -s17 W ATS.0020.059 at 061 to 062.

s18 WATS.0020.001 at018.

s19 SBB.Ol9.843 001. s2o WATS.0020.109atll3.

s21 WATS .0020.1 09 at 174 to 176.

s22 WATS.0020.191 at 193.

s23 WATS.0020.191 at214to215.

s24 ABB0.0913 .167 . s2s BRD .050.0 19 at 030.

s26 ABB0.0313.063.

s27 BRD.050.000 at 004.

s28 ABB0.0313.085.

s29 BRD.057.005. s30 WITS.0179.001 at 074 to 077 pars 348, 349, 350, 354 and 358 to 361.

s3 1 BRD.059.000 at 000 003.

s32 ABB.Ol20.012 at 013 to 014. s33 BRD.061.000 at 000_001; ABB0.0613 .001; ABB0.0513 .116; ABB0.0513.113. 534 ABB0.0513 .123.

s3s BRD.064.119.

s36 BRD.066.000 at 000 003 .

The failure of HIH Insurance 103

537

ABB0.0513.068. 538 HIHK.0005 .00 1. 539

CIV.001.999 160 at 999 215. - -540 SBB.l30.172 001. 541

BRD.Ol2 .000 at 000 001. 542 SBB.l30.311_00l at 311_001 to 311_002; SBB.130.324_001 at 324_004 and 324_009. 543

SBB .l31.892 001. 544 SBB.l32.267 001. 545

SBB .130.324 001 at 324 003 to 324 007 . - - -546 SBB.130.078 001. 547

SBB.l30.047 001 at 047 004 . - -548 SBB.130.311_001 at 311_008; SBB.130.814_001; Tl4409/58 to Tl4410/ 15; SBB.130.810 _001; SBB .l30.481 _001. 549 SBB.l30.492 001. 550 SBB.130.323_001 ; SBB.130.324_001. 55 1 CIV.001.764 at 774 . 552 BRD.029.000 at 000 005 . 553 AUDC.011.001 at 002. 55 4 AARA.0249.0135 . 555 Tl7401 /47 toT17402/31. 556 WITS .0002.154 at 159; WITS.0002.168 at 173 ; WITS.0002.175 at 181. 557 HIHK.0005 .001 at 018 . 558 HIHK.0005 .001 at 019. 559 HIHK.0005.001 at 020. 560 HIHK.0005 .00l at 018 to 020. 56 1 HIHK.0005.001 at 018. 562 HIHK.0005.001 at 019. 563 HIHK.0005 .001 at 020. 564 PWCC.0002.001 at 003. 104 Unprofitable international operations

14 The impact of the FAI acquisition

The adverse financial and managerial consequences of HIH's takeover of FAI Insurances Limited in 1999 were a substantial contributing cause and an important circumstance surrounding the collapse of HIH. Some of those adverse consequences are explored in Section 14.7. Others are dealt with elsewhere in this report. Because of these consequences various events and circumstances leading up to and surrounding HIH's takeover ofFAl were the subject of inquiry by the Commission.

F AI Insurances commenced operations as a general insurer in 1953 and was listed on the Sydney Stock Exchange in the same year. Following a number of insurance company acquisitions in the late 1960s and early 1970s, F AI thereafter conducted its general insurance business through a wholly-owned subsidiary company called F AI General Insurance Company Limited. In 1983 F AI entered the life insurance

industry through a subsidiary company called Falkirk Assurance Society which later changed its name to F AI Life Limited.

F AI was later listed on a number of overseas stock exchanges and it diversified into a range of other financial services including stockbroking, investment advising, superannuation and mortgage and other lending through subsidiary and associated compames.

In 1990 F AI acquired a number of substantial assets of the collapsed Bond group of companies. These assets were acquired following the default by Bond on loans which had been made by F AI to Bond group companies. These assets included the St Moritz Hotel in New York, the coal mining operations of Oceanic Coal Australia Limited and the property development site on the former Emu Brewery premises in Perth.

In 1992 FAI's subsidiary company FAI Life was listed on the Australian Stock Exchange. F AI sold approximately 57 per cent of the issued capital ofF AI Life for $45 million and retained approximately 43 per cent. At about this time F AI also commenced operations in a number of Asian and Pacific general insurance markets

14.1 FAI provisioning

The primary harm done to HIH as a result of its takeover of F AI arose from the unexpected losses it incurred as a result of under-provisioning in F AI. It is necessary then to have some understanding of the F AI insurance businesses, how they operated, and the accounting systems and practices relevant to claims estimates.

The failure ofHIH Insurance 105

14.1.1 The FAI insurance businesses: an overview

During the 1990s F AI carried on insurance business in Australia and to a lesser extent in New Zealand, Papua New Guinea, Hong Kong and Fiji. It also participated in some Lloyd's syndicates in the United Kingdom. The main operating entity that held an authorisation under the Insurance Act 1973 was FAI General Insurance Company Limited. Two other companies-F AI Reinsurances Pty Limited and F AI Traders Insurance Company Pty Limited-had authorisations but were of much less significance. Most of the international business was written by the Australian authorised entities, particularly F AI General.

FAI General had three main divisions managing a number of portfolios:

• The general insurance division managed risks in respect of compulsory third party insurance, smaller public liability risks, workers compensation and other lines of business known as 'personal lines' - fire, household insurance, motor vehicle, and so on.

• The inwards reinsurance division managed risks that were reinsured by F AI.

• The commercial and professional insurance division managed, among other lines of business, larger commercial public liability risks and professional indemnity risks written internationally and in Australia.

At all times that are relevant to the Commission' s inquiry Rodney Adler was the chief executive officer of F AI and a member of the board of directors. Timothy Mainprize was the finance director. From 1988 Niranjan Peiris was employed in a variety of accounting roles; from 1995 he was group financial controller.

In the mid-1990s the general insurance division was managed by Daniel Wilkie and the commercial and professional insurance division was managed by Angus Maciver. At that time Ashraf Karnha was the general manager of the latter, reporting to Maciver. Early in 1997 Maciver resigned and was replaced by Karnha. In late

1997 both divisions came under the control of Wilkie.

In the ordinary course of events, case estimates for all notified claims were set by F AI claims officers and were recorded in the F AI mainframe database, known as Aegis, which had been established in the early 1990s. Among other things, Aegis recorded the claim number, the policy number, the case estimate that had been set, changes to the case estimate during the term of the claim and all payments paid or received in respect of the claim.

There was an interface between the case estimates recorded on Aegis and the F AI general ledger system; in other words, a change to case estimates within Aegis resulted in an automatic corresponding_ change to the level of reserves recorded in the F AI general ledger. It was the data recorded in the general ledger that were the subject of actuarial analysis. If there was a variance between the reserves recorded in the general ledger and what was recommended by the actuary, there would be an adjustment to the reserves. The level of reserves in the general ledger was the basis

106 The impacl of/he FA/ acquisilion

for setting the outstanding claims provision in F AI' s published accounts. 1 In such a system there was potential for under-provisioning if the case estimates were not accurately recorded in the Aegis system: if the case estimates were wrong the general ledger would not reflect the appropriate level of reserves and the actuary would be receiving and relying on inadequate or inaccurate data.

The method by which the F AI general ledger was adjusted to reflect the actuarial analysis was as follows. As noted, there was an interface between Aegis and the general ledger. F AI created a series of templates or spreadsheets that were populated with general ledger balances as at the period in question for the class of business to which the templates related. The balances obtained from the general ledger were obtained through general ledger inquiry and the figures were then manually entered

into the templates. The balances obtained from the general ledger represented the total claims provision for each class of business. The templates for the long-tail classes were forwarded electronically to the external actuaries, who then populated a number of fields, including the IBNR field, and returned the templates to F AI. After

the templates were returned a journal entry would be processed and posted to the general ledger, extracting the figures provided by the external actuary?

Depending on their experience and seniority, claims officers or managers in FAI had authority to set case estimates within a range. A claims officer who recommended a case estimate beyond his or her level of authority was required to obtain approval from the relevant claims manager or, depending on the amount of the case estimate, from the divisional head and beyond that the board.

From 1995 or thereabouts any case estimate increase of more than $1 00 000 required the preparation of a major loss report by the claims officer, setting out in detail the justification for the increase. This would be passed to the relevant claims manager, who would authorise, reject or vary the suggested increase? At the end of June each year there would be a claims file review, in which all claims would be reviewed by claims officers. This was normally, but not always, done by an officer

other than the officer with responsibility for the file. 4 During claims file reviews, case estimates were reconsidered and altered as necessary.

Within F AI there was a research department and a management information services department (also known as the information technology, or IT, department). The research department was staffed with actuaries and other personnel, and one of its roles was to produce data that assisted F AI in estimating outstanding claims

liabilities. In 1995 the research department established a data warehouse that contained, among other things, case estimates of outstanding claims, past claims payments and numbers of claims. From this information it was possible to extract data relating to a portfolio at a particular time, in whatever format was required for

analysis. 5 Each month the data were downloaded from the Aegis system to the data warehouse. The research department then analysed the data for the general insurance division. The commercial and professional insurance division, or CPID, used the IT department, rather than the research department, to analyse its data. This

arrangement continued until late 1997, when Wilkie assumed control of both

The failure ofHIH Insurance 107

divisions. 6 From about 1996 the research department provided data from the data warehouse to the external actuaries for their analysis.

One line of business managed in the CPID was an international professional indemnity line known as Minet's International Professional Indemnity, also known as MIPI or the Big Six. This business was written by F AI from the late 1980s until the mid-1990s in respect of the six largest accounting firms in Australia. The CPID also managed an international line of business known as ALAS (the Attorneys' LiabilityAssurance Scheme), which dealt with American attorneys ' liabilities; this was written by F AI from the late 1980s until the early 1990s. A further line of business written out of Australia was in respect of the Medical Defence Union (also known as UMD, UMP or AMIL); this was written from about the mid-1980s to the late 1990s, and from 1997 it was placed in FAI's inwards reinsurance portfolio.

The policies underwritten by F AI in respect of the Big Six and ALAS were excess­ of-loss policies in which a pool of insurers covered the risk at various layers. 7 In the case of ALAS, in some years F AI was a direct insurer and in some years a reinsurer. Thus certain years of the business were managed by the CPID and other years by the inwards reinsurance division. F AI' s participation varied depending on the layer and quantum of cover.

In respect of the Medical Defence Union, F AI basically operated as a reinsurer. The cover provided varied widely from year-to-year but contained two main elements.

From 1984 to 1990 it consisted of a combined excess (for MDU) and limited cover (for F AI), for both individual claims and the total cost of claims. From 1991 the cover was split into two parts, with an individual claim excess and limit and a stop­ loss excess and limit applying to claims below the individual excess; from 1984 the

indi vidual excess of loss contained a complex experience-rating mechanism for determining premiums. F AI was the sole underwriter on these contracts until 1994, after which time the experience-rated excess-of-loss contract was led out of London and F AI' s share declined to 50 per cent. F AI took out no outwards reinsurance in relation to the Big Six, ALAS or MDU.

For certain lines of business, the broker would produce bordereaux. These were usually sent to F AI quarterly and recorded, among other things, the relevant layer, the claim, payments to date, outstanding or case estimates, together with F AI' s participation. The bordereaux for the Big Six and ALAS were signed off by the lead underwriter. In the case of the Big Six, the bordereaux excluded IBNR

8 ; in the case

of ALAS, it included an adverse development factor similar to IBNR. Because F AI was not involved directly in managing the claims, the bordereaux were material documents in the setting of reserves.

PricewaterhouseCoopers provided actuarial services for the CPID and fo r the general insurance division. This included the preparation of formal reports for which PWC undertook an independent assessment of the data provided to it by employees ofF AI, to give statistical central estimates of the claims outstanding at the end of a reporting period. PWC also reviewed reports prepared by FAI's internal actuaries

108 The impact of the FAI acquisition

and prepared forms to be submitted to the Insurance and Superannuation Commission for some portfolios.9 In the main, Christopher Latham and Phillip Chappell performed the F AI actuarial work on behalf of PWC.

From 199 5 to early 1998 Antony Boulden, the financial controller of the CPID, was the primary point of contact within F AI for the provision of CPID data to PWC. A number of PWC's reports in this period note that they were compiled at the request of Boulden. Latham confirmed that Boulden gave instructions to PWC in

relation to the CPID public liability portfolio, professional indemnity insurance portfolio and inwards reinsurance. 10

The general insurance division had within it a research department. Late in 1996 Ian Heppell became research manager, reporting to Wilkie. Heppell was responsible for all products within the division. After Wilkie became chief operating officer m November 1997, the research department covered the products of both divisions.

F AI also retained internal actuaries. Peter McCarthy was engaged in July 1995 as national CTP product manager; he also had several other positions, including that of research manager. Another internal actuary was Rodney Hoskinson; he prepared reports in respect of a number of portfolios, including compulsory third party, workers compensation and public liability. These internal reports were often

reviewed by the external actuary. In February 1998 Geoffrey Trahair started providing actuarial services to F AI, particularly in relation to the international professional indemnity portfolio.

14.1.2 Problems with provisioning practices

After FAI had been taken over by HIH in early 1999, several reports prepared by the internal audit division ofHIH noted deficiencies in FAI's provisioning practices- in particular, portfolios.'' It could be said that HIH itself was not the epitome of best practice when it comes to provisioning, so these reports should not be relied on. On the other hand, at least two independent reports prepared during this time were

critical ofFAl's practices in this regard; they are a better guide.

During 1997 the Swiss Re insurance company had been in discussions with officers ofF AI concerning reinsurance treaty renewals. On 27 October 1997 Swiss Re sent F AI a report covering various matters and containing the following critical comments about FAI's provisioning practices:

• failure to monitor claims below $100 000

• claims review staff reviewing their own claims files

• no methodology for the claims-handling process

• claims-reserving guidelines that were rudimentary

• claims registered at nominal amounts and no realistic reserve raised for some years later

The failure ofHIH Insurance 109

• lack of clarity in the method by which claims estimates were monitored

• absence of a claims manual detailing basic processes and controls.

In April 1998 General Cologne Re Australia Limited conducted basic due diligence on F AI' s outstanding claims liabilities as part of the negotiations for a reinsurance contract. It determined that, for the Big Six scheme, F AI received quarterly bordereaux detailing individual losses by accountant and layer (without IBNRs) but that the documentation was not processed at all. Rather, F AI' s system was to reserve only as paid. This meant that FAI's Big Six loss reserves were booked only after a monthly or quarterly statement of account was received and paid. The result was that no case reserves were taken up. 12

14.1.3 Hoskinson's review

I propose to spend some time examining events surrounding an actual review of the public liability portfolio in 1997. I am doing so not because I have decided there might have been breaches of the law but because it illustrates a quite extraordinary attitude to the transmission of data within and between F AI and its external actuaries. It also reveals tensions within F AI and an approach to resolving tensions that was curious. These matters are examples of reserving practices that were not conducive to identifying and confronting problems.

In about March or April 1997 Boulden instructed PWC to carry out a valuation of FAI's public liability portfolio. Hoskinson, an internal actuary in FAI's research department, initially reported to McCarthy and then to Heppell. 13 In

about March 1997 Kamha approached McCarthy and asked that the research department carry out an actuarial review of the CPID's public liability portfolio. McCarthy agreed and instructed Hoskinson to do the work. 14

Hoskinson started work in relation to the portfolio in about March 1997.15 He was not told of PWC's involvement16, although he understood his report would be reviewed by FAI's external actuary, PWC. 17 On 14 July 1997 Hoskinson emailed Boulden to inform him that he expected to have his draft report ready to give to PWC 'within the next day or so' .18 On the same day Kamha forwarded an email to Hoskinson (copied to McCarthy and Boulden) in which he instructed Hoskinson that no data were to be sent to PWC, nor were there to be any discussions with them regarding the CPID's public liability portfolio, without Kamha's permission. He also asked Hoskinson to treat the matter as 'very serious'. 19

On or about 25 July 1997 Hoskinson became aware that PWC had been retained to carry out an independent actuarial review of the CPID's public liability portfolio. 20 By that time Hoskinson had concluded that the portfolio was required to book about $40 million by way of reserves. PWC was recommending a figure of about $28 million. 21 On 28 July 1997 Kamha, in an email copied to Boulden, advised Hoskinson that he was no longer retained to perform any further work for the CPID and that he had no responsibility or accountability for advising or recommending on the level of CPID's loss provisions. Kamha also instructed Hoskinson that the

110 Th e impact of the FA! acquisition

information and data in his possession as a result of the work he had done 'should not be released to or discussed with any person other than Tony Boulden or myself without my prior approval ' .22

Hoskinson forwarded a copy of Kamha's email to Heppell on 31 July 1997 .23 Following a meeting with Hoskinson, Heppell emailed Kamha on 31 July 1997 stating:

Rodney [Hoskinson] and I have discussed the professional issues made by the withdrawal of your request for him to recommend provisions for CPID's PL portfolio.

Tomorrow morning we will send you and Tony Boulden a draft report containing Rodney's advice. It will be a separate report. It will not be attached to or mentioned in the report to the General Insurance Division. Without further discussions with you it will remain a draft .

. . . We still feel that the most appropriate course of action would be to hold discussions between Research, [PWC] and yourself so that the differences in view [sic] are exposed and examined.

While we encourage you to seek more than one opinion we are disappointed that we were not informed that [PWC] were working on this project. Appropriately timed discussions between us and them would have improved the quality of advice of both parties and benefited F AI. Also, we feel that they have been denied access to the previous work that we have done and the detailed data that we have prepared and examined. In

Rodney's opinion, if you set the provisions on [PWC's] advice the portfolio will be under-reserved. Further, if you were to set premiums in line with that advice, losses would result. Rodney and I are employees of F AI General Insurance Company Limited and, I believe, carry some responsibility to the shareholders to press this view. We would like to provide a copy of the report to Tim Mainprize as well and Tony and yourself to clearly satisfy this responsibility.

You asked Rodney not to release any of the information or data that he has gained through doing the work to any person other than Tony or yourself without your prior approval. May we have your permission to provide a copy of the draft report to Tim Mainprize? 24

Heppell's email raised a number of important matters that warranted consideration. Hoskinson's review had concluded that the level of reserves required was higher than that identified by PWC, so it was appropriate that the disparity be the subject of debate and analysis. It was management' s responsibility to ensure that the actuaries had all material information and data to assist them in recommending reserves; in the circumstances, this would have been facilitated by a free and frank exchange of

views between the internal and external actuaries. The email also confirmed that PWC did not receive detailed data that had been prepared and examined by the research department. Hoskinson testified that the additional data to which he had access probably made his estimate more reliable than that ofPWC.

25

Th e failure ofHIH Insurance 111

On 1 August 1997, following a meeting between Kamha and Adler during which Heppell' s email of 31 July 1997 was discussed, Adler forwarded to Heppell a memo in the following terms:

I am in receipt of the E-mail that you sent Ashraf Karnha, copied to Rodney Hoskinson, regarding CPID' s Public Liability portfolio dated 31 July 1997 .

. . . The Research Department has my absolute support as it is one of the most valuable tools of our organisation. However, we must work as a team and as various Executives/Product Managers approach you for advice/help and your opinion-whether or not that advice is taken up must be up to the executive who originally requested that information. If you abuse that trust they will naturally not approach you.

Whether you are right or wrong is somewhat immaterial at this juncture, as it is not the point that I am making. Please work within the parameters that you have been set and if you feel strongly about any issue (which we respect and appreciate will happen as we have a very strong group of Executives), then you will always have the opportunity to speak to Daniel [Wilkie] , Tim [Mainprize] or myself. But we do not need to recei ve your opinion in writing .

. . . There, of course, will be no di scussion between Research, Coopers & Lybrand, Ashraf/Tim, but we will always appreciate your views and counsel. 26

There had been no discussion between the research department and PWC in relation to Hoskinson's draft report, nor were the additional data relied on by Hoskinson made available to PWC. Had such discussions taken place and all material data been made available to the external actuaries, PWC would have taken such matters into

h

. h . . 27

account w en prepanng t e1r report.

Kamha denied having any role or involvement in determining what data were supplied to PWC. He also denied that he had prohibited the release of information to PWC: he said the instruction that his sign-off was required before the release of information was because he wanted PWC to arrive at its conclusions independently of Hoskinson. In the event, Kamha' s sign-off was not requested, and Kamha submitted that the direction not to discuss the matter with PWC came from Adler, not him. The 31 July 1997 email from Heppell to Kamha, however, referred to· PWC having been denied access to the detailed data prepared and examined by the research department. Both Kamha and Boulden denied being aware of that. 28 It is nevertheless likely that they were aware: there would have been little utility in giving the instruction not to disseminate the data if the data were already in the possession of PWC.

I can understand (but do not necessarily accept) that had Hoskinson communicated his conclusions to PWC before it had reached its own conclusions, PWC' s independence might be said to be compromised. But refusing to allow PWC to receive the data on which Hoskinson's views were based is a different matter.

112 The impact of the FAJ acquisition

I cannot see how the transmission of data to PWC could possibly be said to affect its independence.

Further, the evidence establishes that by at least 25 July 1997 PWC was recommending a figure of about $28 million in respect of the CPID' s public liability portfolio reserves. 29

In those circumstances there was no reason for prohibiting any exchange of views or data between the actuaries on the basis it may compromise the independence of PWC's conclusions. The instruction from Kamha had the consequence that PWC was not aware that Hoskinson had carried out a valuation of the portfolio, nor was it aware of the data upon which the valuation was based. Such a prohibition went beyond seeking to protect the independence of PWC's conclusions. It is possible to deduce that Kamha was motivated by a desire to ensure that PWC was not made aware of Hoskinson's findings or of the data he had relied on, so that PWC would not reconsider its estimate.

Kamha' s evidence was that he terminated Hoskinson's services in connection with the CPID ' s public liability portfolio because Hoskinson had become emotional and accused him of compromising his (Hoskinson' s) professional integrity by preventing him talking with PWC.30 The email terminating Hoskinson' s services was sent on 28 July 1997.3 1 The termination presumably had the intended effect­ that the CPID would not receive Hoskinson' s draft report. Kamha gave no

satisfactory explanation as to why, even accepting his evidence about the reasons for terminating Hoskinson's services, the CPID should not receive the benefit of Hoskinson's work, which was by then complete. The email was written within a few days of Hoskinson having become aware that PWC had prepared a valuation some $ 12 million lower than his own. 32 There is nothing in the contemporaneous material

that gives support to Kamha's reasons for terminating Hoskinson's services, nor is his action consistent with the matters raised in Heppell's email to Kamha of 31 July 1997.

If Hoskinson's report had been delivered to the CPID it seems likely that the CPID would have been required to deal with the conclusions expressed in it. In my view, Boulden and Kamha had decided to accept the PWC valuation, regardless of Hoskinson's conclusions.

This finding is reinforced by the contents of and the events following Heppell ' s 31 July 1997 email to Kamha. That email refers to Heppell's decision to provide Hoskinson's draft report to Boulden and Karnha following Kamha's withdrawal of instructions to Hoskinson. I am satisfied that this prompted Kamha to meet Adler for the purpose of having Adler instruct Heppell in the terms set out in Adler's

memo of the same day. Adler's instruction had the effect of prohibiting the dissemination of the report. Kamha testified that it was because of Adler's instruction that he subsequently withheld Hoskinson's draft report from PWC. 33

In

my opinion, Kamha had decided on that course before his meeting with Adler.

In evidence, Adler asserted that during his meeting with Kamha on 1 August 1997 there was no discussion of the extent of the discrepancy between Hoskinson's calculation and that of PWC?4 Kamha, however, said he told Adler of the

Th e failure ofHIH Insurance 113

discrepancy. 35 I do not need to resolve this dispute. But there are two matters concerning these events that are striking.

First, Heppell gave evidence that shortly after his (Heppell' s) 31 July email Karnha told him, 'There are some things it is better for [Adler] not to know ' .36 Heppell took this comment as an attempt to dissuade him from delivering a copy of Hoskinson ' s draft report to the CPID. But it is consistent with the inference that Karnha was selective in the information he passed on to his superiors.

The second matter relates to the research department itself. It had been created and championed by Wilkie, with Adler's support and encouragement. There was tension and rivalry between Karnha and Wilkie. Karnha was reluctant to use the services of the research department at all, but Adler had insisted that Karnha do so in relation to the public liability portfolio. 37 Karnha's view was that Adler's pressure on him to use the research department had resulted only in problems and aggravation. 38 When Karnha came to see him about Heppell ' s email, Adler considered that the research department had become a 'political football ' between Karnha and Wilkie. Adler said that his motivation in agreeing to Karnha' s request and in issuing his direction of

1 August 1997 was to defuse the tension and to preserve the role of the research department. This is a strange way to resolve a ' management problem' in circumstances where the portfolio was to be assessed by the external actuary and in circumstances where Heppell ' s memorandum warned of under-reserving.

The 1 August memo from Adler contained a number of unsatisfactory elements. First, it advocated a prohibition on Hoskinson's report being forwarded to the CPID-contrary to the suggestion of Heppe!!. In doing so, Adler adopted Karnha's position of depriving the CPID of Hoskinson's views; those views would have been of assistance in determining the appropriate level of reserves. Second, the memo prohibited any discussion between the research department and PWC in relation to Hoskinson' s draft report. A full and frank discussion between the actuaries would have been helpful in determining an appropriate level of reserves and may have caused Hoskinson or PWC to adjust their figures in the light of such an exchange. Third, although Adler informed Heppell he would always have the opportunity to speak with Wilkie, Mainprize and himself, he instructed Heppell that he was not to express his opinions in writing. This suggests that Adler was concerned lest there be a written record of Heppell's views on the matter, which might be the subject of examination at a later date. Finally, Adler directed that there be no discussion with the chief financial officer ofF AI (Mainprize): no rational explanation was given for that instruction.

Counsel assisting submitted that Karnha, Boulden and Adler might each have breached the Corporations Law in relation to these events. In the end I have decided not to make that finding. The consequence of Karnha' s actions was to deprive the actuaries (and therefore the auditors) of the benefit of Hoskinson's work. That was unfortunate. Boulden supported Karnha' s position. That too is unfortunate.

Adler was the chief executive ofF AI at the time of the events under discussion. In that role he should have been conscious of the need to ensure that FAI's actuaries

114 Th e impact of the FAJ acquisition

were provided with all material information and that there was free and frank discussion between management, the research department and the external actuaries, to ensure the level of reserves held by the company was adequate. Adler did not take these steps. He supported a course of action that prevented the proper exchange of views between the actuaries. Such an exchange of views was likely to have improved the quality of the information provided by the external actuaries to FAJ .

I find that Adler's conduct did not measure up to what was required of him in the proper discharge of his management functions to lead, guide and direct the senior executives of F AI. In this respect he was involved in an undesirable corporate governance practice. So too were Kamha and Boulden.

14.1.4 Under-reserving at 31 December 1997

In May 1997 Boulden observed high levels of payments in relation to the international professional indemnity portfolio when compared with case estimates.39

Following a discussion with Kamha, he began a review of the portfolio. In September or October of that year he reported to Kamha that there was significant under-reserving in relation to the portfolio.4° Kamha said that a short time later he informed Adler of Boulden' s findings and told him that the shortfall was likely to be significant.41 Adler instructed Kamha to report the final valuation to

him as soon as possible and to keep the matter confidential.42 Adler said Kamha told him the shortfall was in the order of $5 million to $10 million and agreed such an amount was material in relation to the reported profit of the F AI group as at 31 December 1997.43

On or about 2 December 1997 Boulden informed Kamha that he had calculated $112 million shortfall in reserves as at 30 June 1997, excluding IBNR.44 A prominent cause of the shortfall was the failure to ensure that case estimates contained in bordereaux forwarded to F AI were included on Aegis. The reasons for that failure are discussed below. Boulden set out his work and conclusions in a spreadsheet that was password protected.45

Kamha said that shortly after his meeting with Boulden he had a meeting with Wilkie. As noted, in November 1997 Wilkie had been appointed chief operating officer, assuming responsibility for both the general insurance division and the CPID. Wilkie said he did not assume effective control of the CPID

until February 199846, but on his appointment he became Kamha' s immediate supervisor. Kamha said he informed Wilkie of Boulden's findings. 47 Wilkie denied that48 and submitted that at no time did he refer to a reserve shortfall figure of $112 million when dealing with any person. He pointed out, for example, that in his dealings with Stuart Stow from Willis Faber and Dumas Limited (who was helping F AI obtain reinsurance) no such figure was mentioned, and he submitted that he was aware of a potential problem broadly in the region of $65 million. He further submitted that had he been informed about under-reserving in accounts other than

The failure ofHJH Insurance 115

those of the Big Six and ALAS he would have taken steps to have the problems quantified and dealt with at the same time.

I prefer the evidence of Kamha- that he informed Wilkie of Boulden' s findin gs in December 1997- for the following reasons. Kamha did not attempt to hide from Adler the fact that Boulden had in September or October 1997 discovered a si gnificant shortfall in reserves, al beit then unquantifi ed. He had brought the matter to the attention of Adler (his immediate superior at the time), so there was no rational r:eason why he would not follow through by reporting the results of the investigation to Wilkie in December, since Wilkie was then the person to whom he reported. Further, Adler had instructed Kamha to report Boulden's findings to him. If he had failed to report the matter to Wilkie, Kamha would have risked being criticised for failing to comply with Adler' s instruction.

Wilkie was unable to explain why Kamha would refrain from informing him of such a matter- other than suggesting his relationship with Kamha was strained. 49 Wilkie' s evidence does not explain why a senior executive of FAI would not inform his superior of serious matters that were material to, among other things, an assessment of F AI' s so lvency. Certainly, Stow was informed of a shortfall in

re serves in respect of the Big Si x and ALAS 5°, but it does not follow that Wilkie was unaware of the figure of $112 million in December 1997. A possible explanation for Stow not being told of the higher figure is that it raised concerns about the solvency of F AI. This may have limited the prospect of obtaining remsurance.

In about February or March 1998 Boulden updated his results in the light of more recent bordereaux received from brokers. He concluded that at 28 February 1998 there was A$1 02 million disparity between the bordereaux and the case estimates recorded on Aegis. 5 1 Once again, Boulden reported his fi ndings to Kamha, who said he in turn reported them to Wilkie. Kamha also said Wilkie instructed Kamha that Trahair, the actuary newly employed by F AI, would take responsibility for the matter and should be provided with Boulden' s file .52 In due course this occurred.53 Wilkie denied any knowledge of the matter54 , but for the reasons already given I prefer Kamha's evidence.

Boulden was not the only person who raised questions about reserving. In 1996 Robert Spratt, the national claims manager of F AI, retained Peter Moran as international claims manager to review the international professional indemnity portfolio. 55 In late 1997 Moran sent to Spratt a series of emails advising that certain lines of FAI business were under-reserved by more than $70 million. 56 He concluded there had been a fai lure to reconcile case estimates in the bordereaux with Aegis. 57

While the evidence established the failure to record in Aegis the case estimates found in the bordereaux, the reasons for that failure are not as easy to discern. Given the passing of time, failings of memory and the lack of documentation, the evidence in this regard was not easy to follow. It does not permit a finding that the failure was the result of a conscious decision on the part of F AI senior management, although

116 The impact of the FA! acquisition

that was the opinion of Moran and Spratt. 58 It was more probably the result of a failure within F AI to ensure that Aegis properly reflected case estimates recorded in the source documentation.

The process appears to have worked adequately between 1992 and November 1995 . During that time F AI had retained Mark Bonnar as international underwriting manager and he was responsible for this work. Bonnar left FAI in November 1995. His replacement, Colin Russell, had no experience of underwriting or managing

long-tail portfolios. 59 He did not reconcile the bordereaux with Aegis60, assuming that this was being done by others.6 1 But there is no evidence that the task was in fact being done by anyone in FAL62 This was confirmed by Moran, who concluded during his investigation that no one appeared to have been allocated the task of

reviewing the bordereaux and ensuring that the recommended case estimates were reflected in Aegis. 63 This leads me to conclude that F AI management failed to ensure that, following Bonnar's departure in November 1995, bordereaux were reconciled with Aegis. By 1997 there was a significant discrepancy between the case estimates identified in the bordereaux and what was recorded in Aegis.

In about late 1997 Boulden retained PWC to conduct a valuation of FAT's international professional indemnity portfolio as at 31 December 1997. He had discussions with PWC and forwarded data to enable it to complete its review and prepare a report. 64 He was aware that if PWC was not informed of the results of his

investigation it was likely to underestimate outstanding claims liabilities. 65 He was also aware that the data forwarded to PWC were misleading because they did not include the shortfall identified by him as a result of his own review. 66 Boulden's explanation for not informing PWC of his investigation and conclusions was that he

had informed his superiors and did not think it necessary to take the matter any further with PWC. 67 But he was the financial controller of the CPID and had retained PWC to carry out a review of the reserves held by that division: it was incumbent on him to ensure that as far as possible all material data were provided to PWC, so that an accurate estimation of outstanding claims liabilities could be carried out by the external actuary. In the absence of such data, there was a serious

risk that the provisions held by F AI would be understated and that as a consequence reported profit would be overstated and the balance sheet misleading.

On 20 January 1998 Phillip Chappell ofPWC advised Boulden that $38.509 million could be booked in reserves for the international professional indemnity portfolio as at 31 December 1997.68 On 21 January, in an email to Marc Gross and copied to Kamha, Boulden instructed Gross to book the amount recommended by the

actuary. 69 The instruction was issued in the knowledge that the recommended amount seriously underestimated F AI' s outstanding claims liabilities as determined by Boulden's investigation. 70 Boulden conceded that because of the instruction he gave to Gross it was highly likely F AI' s published accounts as at 31 December 1997 would be misleading. 7 1

I am satisfied that in late 1997 Boulden forwarded, or was privy to the forwarding of, data to PWC that he knew to be misleading for their interim valuation of the

The failure ofHJH Insurance 117

international professional indemnity portfolio as at 31 December 1997. I am also satisfied that on 21 January 1998 he instructed the booking of reserves that he knew to be understated by reason of the misleading data forwarded to PWC. Further, I find that Boulden gave such an instruction in the knowledge that reported profit was thereby likely to be overstated in the accounts ofF AI as at 31 December 1997 and the balance sheet misleading.

I also find that, although Boulden was aware that the auditors had been given the same datil as PWC and that the auditors had been misled as to FAI's outstanding claims liabilities, he took no steps to remedy the situation. 72 He said he would wait until he was asked before providing information to Andersen. 73 In my opinion, that is a misunderstanding of the relationship between management and the auditors: if he was aware the auditors had been provided with misleading data he should have taken steps to advise them of the problem, rather than wait for the auditors to raise the matter with management. In this respect Boulden contributed to undesirable corporate governance practices.

After the departure of Maciver in early 1997 and before Wilkie had assumed responsibility for both the general insurance division and CPID in late 1997, Karnha reported monthly to the board in respect of the CPID. His November 1997 report does not refer to under-reserving in the CPID. He had been aware of the problem since at least September or October 1997 by reason of his discussions with Boulden.74 In my opinion, Karnha's failure to report such a matter to the board had the consequence of misleading the board in relation to the profitability of the CPID and the solvency of the group's operations. Any instruction from Adler to keep the matter confidential did not excuse Karnha from informing the board of that matter­

unless he had satisfied himself that others had taken steps to inform the board.

The question arises whether Karnha's conduct in omitting from his November 1997 report to the board the information passed to him by Boulden about 'significant under-reserving' in the professional indemnity portfolio rendered the report false and misleading. If he knew it to be false and misleading, this might be a

contravention of s. 1309(1) of the Corporations Law; if he failed to take reasonable steps to ensure that it was not false and misleading, it might be a contravention of s. 1309(2).

When Karnha was preparing the report to the November board meeting the extent of the under-reserving had not been quantified, although further work was in progress. The board should have been told. On the other hand, Karnha had passed the information to Adler, the chief executive and a member of the board. Adler had instructed him to keep the matter confidential.

Wilkie was aware of the level of un9er-reserving within F AI, as identified by Boulden in early December 1997. Wilkie knew the accounts published as at 31 December 1997 were misleading. He was the person in charge of reporting the performance of the insurance divisions within F AI, and he must have known the company did not take up the shortfall in reserves that Boulden had identified only a few weeks before the end of the half-year. This was confirmed when Wilkie was

118 The impact of the FAJ acquisition

subsequently informed of Boulden's updated report in early March 1998. However Wilkie took no steps to inform the board of the level of under-reserving at the board meeting in February 1998.75

The board report suggests a certain level of under-reserving. By the time he came to prepare the report to the February board meeting, Wilkie knew of Boulden's estimate that the professional indemnity portfolio had a reserve shortfall exceeding $100 million. There is no reference to that figure in the report. In my view, this would be sufficient to render the report false and misleading in a material particular. Unlike the situation with Kamha's report to the board in November 1997, Wilkie

was reporting to the board at a time when the under-reserving had been quantified­ if not finally at least with some degree of precision.

In September or October 1997 Adler became aware, as a result of a conversation with Kamha, that there was a possibility that under-reserving within F AI amounted to between $5 million and $10 million. 76 Kamha said he told Adler that any identified shortfall was likely to be significant. 77 Adler asked Kamha to prepare a

report an d come back to him in relation to the matter as soon as possible. 78 Subsequent to this meeting, Wilkie was appointed chief operating officer ofF AI and assumed responsibility for both the GID and the CPID. In the months following September-October 1997 Adler did not ask Kamha or Wilkie what

investigation had occurred, what was the progress of the investigation, what was the magnitude or nature of the problem, how it was to be dealt with, how the problem had arisen, and why it had to wait until February before it could be dealt with by an actuary. 79 Further, in none of his reports to the board in late 1997 or early 1998 did Adler make reference to matters raised by Karnha or Wilkie in the months following September-October 1997. He submitted that he preferred to go to the

board only after he had a final report of the investigations. Yet this decision kept the board in ignorance of material information, and, as it turned out, the report that was eventually prepared was concealed from the board.

14.1.5 Manipulation of case estimates on 2 January 1998

Kamha gave evidence that in late December 1997 he was out of Sydney, on holidays with his family. He said that whilst on holidays he received a telephone call from Wilkie--either shortly before Christmas or between Christmas and New Year. Wilkie informed him that, as a result of a claims review, adjustments had been approved and a list setting out the changes was being sent to the claims department. He asked Kamha to ensure that list was processed before the close of the accounts for the month. 8° Karnha testified that, after that conversation, he had a conversation

with Ronald Shorter, a claims officer, in which he advised Shorter that an approved list of adjustments was being sent to him and asked that the adjustments be processed before the close of the accounts for the month. 8 1

Karnha did not see the

list at the time, nor did he personally arrange for the list to be sent to Shorter. 82

Shorter said he received the instruction from Kamha during a telephone conversation on 2 January 1998. 83 The list required case estimates to be reduced by $24.78 million. 84 Shorter and an assistant made the alterations on 2 January 1998. 85

Th e failure ofHIH Insu rance 119

Alterations to case estimates on Aegis on 2 January 1998 would be reflected in the accounts ofFAl as at 31 December 1997.86

None of the people called to give evidence in relation to this matter attempted to justify the adjustments made on 2 January 1998. Shorter said he knew the reductions had nothing to do with prudent management of files and assumed it had something to do with interim reporting.87 Similar evidence was given by Spratt and David

Kendrick, a professional indemnity claims manager within F AI. 88 In my opinion, the evidence . establishes that the $24.78 million reduction in case estimates on 2 January 1998 was improper and was done with the intention of inflating reported profit as at 31 December 1997. It is difficult to say who was responsible for the manipulation of the case estimates.

Kamha denied that he understood the purpose of the adjustments was to inflate profit artificially.89 He said it was not until he was preparing his evidence for the Commission that he saw, for the first time, the list of adjustments provided to Shorter on 2 January 1998. When he did see the list, he formed the view that the only rational explanation for the reductions was to improve reported profit as at 31 December 1997.

90 I accept Kamha' s evidence that he did not see the list at the time the adjustments were made and was not aware at that time of the improper nature of the adjustments.

Kamha also gave evidence that, on either the day of his conversation with Wilkie or the following one, he received a telephone call from Mainprize. He said Mainprize asked him whether the adjustments had been made. Kamha responded he had spoken to someone in the claims department to ensure that the adjustments were made before the end of the month. 9 1 Mainprize denied having such a conversation with Kamha. He said he was in French Polynesia without convenient telephone access from 24 December 1997 until 6 January 1998, although the hotel reception did have a telephone.92 Mainprize also denied having seen the list provided to Shorter on 2 January 1998. 93 I prefer the evidence of Mainprize in relation to this matter. If Kamha had had a conversation with Mainprize in the terms alleged by Kamha he must have already spoken with Shorter-namely, on or after 2 January 1998. Yet Kamha testified that he spoke with Mainprize on either the day of his conversation with Wilkie or the following one. He placed his conversation with Wilkie in the period before Christmas or between Christmas and New Year. The evidence does not permit a finding that Mainprize was aware of the adjustments.

Wilkie denied having a conversation with Kamha in late 1997 in which he asked Kamha to process adjustments to the accounts to the month

ending December 1997.94 He denied asking anyone at FAI to make such adjustments or ever having seen the list provided to Shorter on 2 January 1998. 95

Wilkie submitted that the only evidence he was in any way involved in this instance of manipulation of case estimates was that given by Kamha and there was no contemporaneous document or other material that would confirm that fact. He also submitted that other F AI personnel must have been involved in the preparation of

120 Th e impact of the FA! acquisition

the list that was forwarded to the claims department and that if he had been responsible for the list he must have been in contact with those people. Leaving Kamha's evidence to one side, there is no evidence that Wilkie spoke to anyone about the preparation of the list, its contents and effect or the means by which it was forwarded to Shorter or the claims department.

The evidence as to what occurred after 2 January 1998 is also inconclusive. Shorter said he had no further conversations with Kamha about the matter and nor did he discuss it with Wilkie. On 15 January 1998 he was contacted by John Byrne, a senior underwriter, and asked to prepare the monthly report for Wilkie. Later that

day Shorter discussed the issue with Spratt. They decided to append the 2 January reductions to the monthly report. This was done and Shorter forwarded the monthly report to Kamha so that it could be passed on to Wilkie.

There is no evidence of any discussion in January 1998 between Kamha and Wilkie in relation to the report. It might be expected that if Wilkie had then seen the adjustments for the first time he would have taken the matter up with Kamha. This did not occur.

The evidence is that between January and June 1998 there was an upward movement in the claims estimates included in the list, effectively cancelling out the 2 January 1998 decreases.96 But this is a reconstruction carried out by an expert retained by the Commission. There is no evidence as to how these further adjustments came about or on whose instructions they were made.

The 2 January 1998 adjustments are a very serious matter. They are indicative of a manipulative approach to case estimates. But in the light of the dearth of evidence as to the provenance of the list, who directed its preparation and how it was dealt with, I am unable to apportion responsibility.

14.1.6 Reductions in the general ledger

In an email dated 23 January 1998, from Boulden to Gross and copied to Kamha, the following is recording:

Please reduce the following case estimates in the Ledger

State 22 Liability State 21 PI State 22 Marine (OMM)

Please reduce the following lBNR's in the Ledger

State 13 Liability Total P&L mvt

entire balance $1.4m by $1m by $1.5m

by $1.6m is $5.5m

Advice from Graeme King re PI reserves should be forthcoming ASAP

All the adjustments should bring the result to under $20m ... [emphasis added] 97

The failure ofH!H Insurance 121

The corporate management accounts for December 1997 projected a year-to-date underwriting loss for the CPID of $19.676 million. 98 In a 23 January 1998 email from Gross to Boulden the following is recorded: 'I will organise these [sic] for these journals to be posted. As a matter of urgency could you please provide explanation for these journals so that they can be attached to the journals' .99 In a reply email on the same day, Boulden stated, 'Reduction in reserves due to claim review in December 1997 not processed in Aegis due to staff resourcing issues.

Should be processed in AEGIS in January 1998'. 100

Boulden testified that it was most unusual to alter case estimates in the general ledger. 101 He also said that if the adjustments had been made in accordance with his email there would have had a positive effect on reported profits. 102 He denied that a reference in his email to bringing the result to under $20 million was anything other than a comment relating to the outcome rather than the purpose. 103 In my opinion, however, Boulden's evidence does not sit well with the plain wording of his email, especially having regard to the corporate management accounts for December 1997, which projected a year-to-date underwriting loss for the CPID of just under $20 million. Further, Boulden's explanation for the adjustments, as expressed in his 23 January 1998 email to Gross, does not accord with the evidence. The evidence of

Shorter and Kendrick was that they were not aware of any claims review within the CPID in December 1997. 104 Nor was Boulden able to explain why 'staff resourcing issues' were such that the adjustments could not be processed until

23 January 1998 .105

Boulden claimed that by January 1998 his role in preparing the accounts for the CPID had diminished. 106 In my opinion, however, that does not accurately reflect his role when one considers his direction to Gross in the email of 23 January 1998. Of note here is Peiris's evidence that Boulden had prime responsibility for the accounting of the results for the CPID as at 31 December 1997, subject to Kamha' s supervision. 107

The adjustments listed in Boulden's email to Gross dated 23 January 1998 cause me concern because of the instruction that 'the adjustments should bring the result to under $20m', bearing in mind that the corporate management accounts for December 1997 projected a year-to-date underwriting loss for the CPID of just under $20 million. Nevertheless, lam not satisfied there might have been a breach of the law in respect of the matter. Section 1309(1) of the Corporations. Law provides that it must be an officer 'of the corporation' who makes available or furnishes information to directors 'of the corporation'. Boulden was an employee of FAI General Insurance Co. Ltd 108, but there is no evidence that by reason of the email instruction the directors of that company were provided with information that was false or misleading.

122 Th e impact of the FA! acquisition

14.1.7 Deletions of claims from large claims data

Another matter that arose during the inquiry was whether Boulden instructed Hoskinson to delete certain claims in large claims data that were forwarded to PWC so that it could estimate the value of outstanding claims as at 31 December 1997.

Heppell and McCarthy each gave evidence of discussions with Hoskinson in which they were informed of that instruction.109 Hoskinson testified that Boulden told him to withhold two large claims from the December 1997 data forwarded to PWC 110 and that the data were to be deleted because the claims were not due to be paid until the following year. This, of course, would not be an appropriate reason for deleting the data. Boulden said, although the data may have been removed from the large claims data, he did not believe they had been removed from the data provided to PWC.111

14.1.8 Hong Kong adjustment

A further matter that was the subject of inquiry was an adjustment made to the Hong Kong professional indemnity or medical malpractice gross IBNR reserve for the period ending 31 December 1997. On 23 January 1998 Boulden sent to Graeme King, manager of First Pacific's operations in Hong Kong, an email suggesting that,

in order to improve First Pacific's underwriting result for the half year, the IBNR be maintained at the June 1997 level. 112 The email suggested two options: that the IBNR be reduced in the books of First Pacific and subsequently built up over the next six months; or that Gross or Peiris reduce the IBNR on consolidation in the group's books with the reduction to be reversed in January 1998. Boulden said he gave the instruction to King because the increase in the IBNR in the period from June 1997 to December 1997 was not an accurate reflection of the IBNR position. 113

Mainprize said that because of a computer problem the IBNR had been incorrectly calculated and adjustments had had to be made. 114 He asserted that the computer system in Hong Kong automatically calculated an IBNR from July onwards such that by December there had been a large build-up on IBNR.

11 5 In a 23 January 1998

email to Boulden, King expressed his reluctance to reduce the December 1997 IBNR figure, because of the requirements of the Hong Kong Insurance Authority, and concluded that the adjustment would only 'add to the various problems we are -h • h . h" fi . 1 '

116

1acmg ere m t IS manc1a year .

14.1.9 The supply of data to PWC

In late April1998 McCarthy met with Wilkie and Trahair. According to McCarthy, during the meeting he wrote on an electronic whiteboard in Wilkie's office that estimated under-reserving totalled approximately $250 million. McCarthy testified that during the meeting Wilkie said something to the effect that 'it's all over' and

'we are creators of our own demise' .11 7 Trahair recalled meeting with McCarthy and Wilkie in April or May 1998 and that McCarthy informed them of under-reserving

Th e failure of HJH In surance 123

problems. He recalled that the numbers were written up on a whiteboard and they added them up. The total was at least $100 million but he could not remember if it was as high as $2 50 million. 118 Wilkie had no recollection of any meeting in April or May 1998 with Trahair and McCarthy during which McCarthy identified the magnitude of under-reserving on a whiteboard. 119 He said he would have placed no reliance on McCarthy's figures in any case, because he did not believe McCarthy had the information or the expertise in relation to the CPID and considered McCarthy 's initial position was to overreact. 120

Heppell also gave evidence that in or about March or April 1998 he had a discussion with Trahair in which Trahair advised him that the PWC estimates were low by about $150 million and that Hoskinson believed the CTP was under-reserved by up to $30 million. 121 Heppell went on to say that following that conversation, he had a discussion with Wilkie in which he asked Wilkie whether he was aware of A$180 million shortfall in reserves. Heppell gave evidence that he wrote detailed figures on the whiteboard in Wilkie's office, working from memory of what Trahair

had told him. 122 Heppell said he was subsequently informed by Trahair that the matter would be dealt with by Trahair, Peiris, McCarthy and Wilkie. 123 He also said that Wilkie subsequently told him the information about the level of under-reserving should go no further. 124 I accept the evidence of McCarthy and Heppell in relation to their conversations with Wilkie about under-reserving within F AI. Their evidence was credible and not motivated by self-interest.

Probably in late May or early June 1998 McCarthy and Trahair met to discuss the supply of data to PWC so that PWC could prepare its reserve valuations as at 30 June 1998. McCarthy and Trahair were faced with a difficult situation. They were aware of significant under-reserving within F AI and knew that unless PWC was informed of that it would probably recommend an inadequate IBNR provision. They agreed that, if one or other of them attended a meeting with PWC, they would inform PWC of the under-reserving.

After that meeting, McCarthy and Trahair said they had a meeting with Wilkie. They told Wilkie they would alert Latham (of PWC) to the problem with the international professional indemnity data on Aegis if they were put in the position of discussing the outstanding claims reserves for the CPID. McCarthy said Wilkie informed them they were to refer PWC to Kamha if it asked any questions about the CPID, and if Karnha was not available PWC was to be directed to Wilkie.

125

McCarthy said his meeting with Wilkie led him to believe that the data forwarded to PWC in respect of the international professional indemnity portfolio would be provided by Karnha and that such data comprised the case estimates on Aegis, which had not been updated to reflect the bordereaux.126 Trahair was of the same understanding. Wilkie said he had no recollection of giving Trahair instructions on the data that should or should not be provided to PWC and denied that he instructed either Trahair or McCarthy not to disclose to either PWC or the auditors information that would reveal the shortfalls in the case estimates.

127

124 The impact of the FAJ acquisition

In a memo dated 9 June 1998 and addressed to Wilkie, McCarthy and Trahair canvassed a number of matters relating to under-reserving and manipulation of case estimates. The relevant parts of the memo follow:

What does Chris [Latham] need to know? What does he suspect? Will he react badly if he is now told something that he only previously suspected?

Other than basic underpricing of risks the main cause of under-reserving is MANIPULATION OF CASE ESTIMATES.

We are becoming uncomfortable with the approach below as MIPI and ALAS are not the only significant manipulation of case estimates that has occurred-UMD - $20m, PI - $20m, Crane - $17m to $24m.

This memo sets out the points for discussion with Chris Latham prior to the 30 June 1998 valuations.

General Issues

We need to make sure that all FAI staff dealing with Coopers are seen to have the utmost credibility (in Coopers eyes), and it may be possible to sell Tony Boulden as the 'scapegoat' (although this begs the question as to whose directions he was carrying out). IT IS VITAL THAT DANIEL

[Wilkie] MAINTAINS CREDIBILITY IN COOPERS EYES. If Daniel or Research's credibility is in doubt in Cooper's eyes then we are really stuffed ...

Be upfront as much as we can with Chris-found significant problems with CPID, ( eg manipulation of case estimates) ... We cannot increase the reserves to their required level in one step-it will take a few years to get there.

For the CPID portfolios Ashraf Kamha will be the first contact (not Geoff Trahair nor Tony Boulden). Geoff Trahair, Tony Boulden are to refer all Coopers phone calls to Ashraf.

1. MIPI, ALAS, triggering of the GCRA reinsurance contract.

At present Chris is aware of MIPI but not ALAS ... F AI has not raised case estimates to reflect advised outstandings and the shortfall is around $65mAUD.

We have three broad alternatives:

o Say nothing and let Chris comment on the development ofMIPI.

o Disclose only the MIPI problem.

o Disclose both MIPI, ALAS and GCRA contract.

The problem with the first alternative is that the development ofMIPI over the last 12 months has been so dramatic that we won't retain any credibility if we don't talk about it ... Also, if we are not up-front with Chris about it then we could reasonably expect him not to give us a fair hearing on the issue.

The failure ofHIH Insurance 125

In terms of triggering the OCR contract, we will be better placed if Chris accounts for it in his actuarial report, rather than us making an adjustment later. If we make the adjustment then ibecause [sic] of its potential size it simply draws attention to the contract ( eg ISC and auditors) and raises the likelihood that we won't be able to rely on it. To persuade Chris to rely on the OCR contract requires disclosing its nature to Chris and again leads to the conclusion that we should be up-front.

Following this line of logic means that to trigger the OCR contract for ALAS requires that we disclose it to Chris and get him to value it in total ...

We do have at least one argument in our favour. Coopers have valued MIPI on the basis of looking at our data alone. They have simply followed the practice that was adopted by the previous actuary, Jim Gould. They have never asked about our exposures to the various accounting firms and layers. This is not really as thorough as they should have been, given that Coopers are recognised as experts in MIPI reserving. Had they examined our exposures, they would have realised the extent of the under-reserving before now. Chris should be amenable to some form of face-saving solution.

2. Supply of data and past PI problems .

. . . The manipulation of case estimates by CPID is a big issue to discuss with Chris-Daniel can safely say that he does not condone the practice. If he tells Chris we have not got to the bottom (WHICH WE HAVE NOT BUT PROBABLY HAVE IDENTJED MOST OF THE

UNDERRESERVING) of it Chris will feel uneasy about any valuations he does when he knows the extent of some manipulation (eg MIPI, ALAS) ... The extent of the manipulation is very substantial (UMD - $20m, PI -at least $20m, ALAS+ MIPI- $m, Crane - $10m, etc).

Tony Boulden directly manipulated some of the Australian PI data that was supplied to Coopers at December 1997. Several claims worth around $2- $3m were taken out of the data. we [sic] need to confirm that they have all been restored to their correct value .. .

We believe that TB [Tony Boulden] does not have much credibility in Coopers' eyes and we should therefore not involve him in any interaction with them, although he may still have a role in preparing the data. We should emphasise the fact that Research wi II prepare the data and Ashraf Karnha will it to Coopers and that the only 'changes ' will be for erroneous claims. 1 8

The tone and the can dour of the memo are striking. A reader would have understood that Trahair and McCarthy considered F AI to be under-reserved to the extent of about $125 million. It is also apparent from the memo that Trahair and McCarthy sought to persuade Wilkie to 'be up front as much as we can with Chris [Latham]', although less admirable alternatives are·also put forward. Wilkie testified he had no recollection of gaining access to the memo, which was password protected, and no recollection of having read the memo before presenting his evidence at the Commission. 129

126 The impact of the FAJ acquisition

Wilkie gave several reasons to support his claim that he did not read the memo. First, he said his time had been taken up with reinsurance arrangements he was negotiating, which were later to become the National Indemnity contract. 130 He said that contract was a significant matter that Adler had directed him to devote all his time to because it related to the reported profit of the companies as at

30 June 1998. 131 The matters raised in the email from McCarthy and Trahair were, however, likely to have a significant impact on the level of reserves as at 30 June 1998 the actuaries would recommend. Thus the memo of 9 June 1998 related directly to the very matter Adler had asked Wilkie devote all his time to.

Second, Wilkie dismissed McCarthy's views in relation to under-reserving by describing him as a 'noted alarmist' and not an expert in the corporate professional area. 132 He did not, however, characterise Trahair as an alarmist and placed some confidence in him. 133

Since his employment with F AI in February 1998, Trahair had gained some familiarity with the problems of under-reserving in the CPID. The memo he co­ authored with McCarthy suggests a well-reasoned position expressing the concern of two experienced actuaries. Even if Wilkie had reservations about McCarthy, he should have given the memo thoughtful consideration. Third, Wilkie maintained it was his view that internal actuaries had a tendency to come up with higher figures for reserves because if their estimates proved to be wrong it would be better to be

higher than lower. 134 In my opinion, even if this were a bona fide view it was not a reason for dismissing the contents of the memo. The memo's tenor was serious; it called for full and frank discussion between the internal and external actuaries, auditors and the board. None of those discussions took place.

I am satisfied Wi lkie gained access to the memo forwarded to him on 9 June 1998. On that day McCarthy told Wilkie the password for gaining access to the memo was 'Wilkie'. In a return email to McCarthy, Wilkie advised that he was unable to access the memo. This prompted a further email from McCarthy, advising Wilkie that the

password was 'Trahair'. In reply, Wilkie thanked McCarthy. 135 This establishes that

Wilkie both received the memo and had the capacity to open it. I also think it likely Wilkie proceeded to read the document, given the importance of its subject matter to the work Wilkie had been instructed by Adler to do. It follows, in my opinion, that Wilkie became aware not later than 9 or l 0 June 1998 that McCarthy and Trahair

believed FAI to be under-reserved to the extent of about $125 million. The evidence of events that later occurred suggests to me that Wilkie did not follow the memo's advice that PWC be informed of the under-reserving.

Other evidence also suggests that Wilkie read the memo on about 9 June 1998. On 10 June 1998 John Tuckfield sent an email to Geoff Bromley saying,

Spoke to Daniel today and in a nutshell the problem is:

Current account is fine

Corporate and professional has blown out

The failure of HlH Insurance 127

Needs a betterment of about 30 mio in or about 1997/98 to be paid back over the next 5-10 years

Thereafter wants to recognise about 100 mio in total additional unreserved loss in respect of continued deterioration in Angus account over next 10 years ... [emphasis added] 136

The reference to the 'Angus account' was a reference to the Big Six and ALAS accounts managed by Maciver. The combined figure of $130 million referred to in Tuckfield's email to Bromley is materially the same as the total figure for under­ reserving in McCarthy and Trahair's memo to Wilkie dated 9 June 1998. On

11 June 1998, in an email to Bromley, Tuckfield stated,

The propping up of the 1997/98 year would seem easily achieved by the back to back reinsurance approach being considered for New Cap.

The second task may be more difficult. The 100 mio has been arrived at by taking various actuarial views of individual portfolios and accumulating them at the Daniel [Wilkie] level. As such none of the individual line managers (incl. Ashraf [Kamha]) are aware of the full picture and this needs to remain so. Major contributors are Public and PI/D&O. F AI will cease to write PL from today when the underwriter will be sacked. PIID&O tends to come from Angus' bottom drawer and related to old years from about 1988/89. There is thus a time before settlements are made. The auditors [sic} (Coopers) are to be brought in. Daniel meets with them today to explain to them their responsibility for the state of affairs. They may be asked to kick the can but at any rate are going to be really on side with whatever we eventually put up which must be some advantage. The budget for C&P was set at zero and A65 mio loss is now best guess (and this does not include the additional 100 mio). [emphasis added] 137

Wilkie denied informing Tuckfield of the matters referred to in the italicised portion of the email. At this time, he was corresponding with Tuckfield and Bromley in relation to arranging reinsurance. 138 But the details of the email are generally consistent with the memorandum to Wilkie from McCarthy and Trahair. This suggests that Wilkie was conscious of the matters raised in the memo from McCarthy and Trahair at the time he spoke with Tuckfield. I also note in passing that in evidence it was established that a meeting between Wilkie and PWC had been set for 11 June 1998 (the date referred to in Tuckfield's second email), although the meeting was subsequently postponed to 23 June 1998. 139

There was no dispute that the data forwarded to PWC to enable them to prepare their valuation of the international professional indemnity portfolio as at 30 June 1998 did not contain genuine case estimates. Nor is it disputed that Karnha sent the data to PWC. But there was a conflict in the evidence as to whether Kamha was aware that the data did not include genuine case estimates.

Trahair could not recall specifically mentioning to Karnha that the shortfall in the international professional indemnity case estimates had been accounted for in the data the research department had prepared and that were being sent to Karnha. 140

But Trahair assumed Karnha knew because the information was in the same format

128 Th e impact of the FAI acquisition

as that in which it had been supplied in December 1997. 141 Trahair had previously discussed, with both Wilkie and Karnha, the case-estimate shortfall he (Trahair) had identified. 142

Kamha instructed Trahair to supply the data as previously agreed. 143

Karnha submitted that he had no motive for deliberately sending inaccurate data to PWC because by then he was planning to leave F AI. His evidence was that he left F AI in September 1998, by mutual agreement with the company. He had told Wilkie in August 1998 that he intended to initiate legal action against F AI for constructive dismissal, having been gradually and consistently stripped of his authority since Wilkie's appointment as chief operating officer. 144

There is sufficient uncertainty to cause me to hesitate before deciding that Kamha knew the data sent to him by the research department did not contain the identified shortfall in case estimates. I do not find that Kamha might have contravened the law in this respect.

As noted, I am satisfied that Wilkie gained access to and read the memorandum McCarthy forwarded to him on 9 June 1998. This must have put him on notice-if he was not previously aware- that McCarthy and Trahair believed F AI to be under­ reserved to the extent of $125 million. The memorandum also counselled Wilkie to

' come clean' with PWC in relation to the under-reserving identified by them. Contrary to that suggestion, Wilkie informed McCarthy and Trahair they were to have no direct contact with PWC and that Kamha was to be the F AI point of contact for any dealings with PWC. By reason of that direction, the data forwarded to PWC to enable them to prepare their valuation for the international professional indemnity portfolio as at 30 June 1998 were misleading in that they did not contain genuine

case estimates. This was likely to cause PWC to underestimate the required reserves for IBNR.

Wilkie submitted that responsibility for liaising with the external actuaries fell to others within F AI. This is correct as far as it goes, but it does not deal with the evidence of McCarthy and Trahair, who informed Wilkie orally and in writing in April and June 1998 of significant under-reserving within FAI and Wilkie's rejection of their proposal that PWC be informed of the situation. Wilkie also submitted that, regardless of who provided data to PWC, the evidence clearly showed it was not him. He further submitted there was no evidence to suggest he was responsible for the preparation or forwarding of data to PWC, nor was there any

evidence that he knew the information provided to PWC was not complete.

As stated, I accept Trahair and McCarthy's evidence that they informed Wilkie of serious under-reserving within F AI in or about April 1998 and that in a memo dated 9 June 1998 they identified under-reserving in certain F AI portfolios and suggested that PWC be informed of that matter. I also accept Trahair and McCarthy's evidence that they told Wilkie they would inform PWC of the situation if PWC raised the subject with them. It is my opinion that, faced with this situation, Wilkie directed

that Karnha was to be the point of contact with PWC and that all data would be forwarded to PWC via Karnha. Each of those matters is consistent with Wilkie

The failure of HIH Insurance 129

being aware the data were incomplete: otherwise, there was no purpose in putting in place the procedure by which material was to be transmitted to PWC.

Trahair and McCarthy acknowledged they were aware that the data forwarded to PWC were misleading. They were in a difficult position. They had told a senior executive in F AI of thejr views about under-reserving in the company and had suggested that the external actuaries be informed of the matter. Wilkie rejected that advice. McCarthy agreed he would have been acting contrary to Wilkie's

instructions if he had told the truth to PWC. 145 He was also mindful of the fact that a term of his employment contract prevented him from approaching PWC in relation to the matter without his employer's consent. Trahair said he was unhappy with the situation because he regarded dealing with PWC in that way as dishonest but was being 'over-ridden'.146

It is never easy to resolve such situations. As I said before, calling in aid an order or instruction from a superior will not always be a complete answer; nor will resort to a confidentiality clause in an employment agreement suffice when duties and obligations come into conflict. In my opinion, although McCarthy and Trahair might have dealt differently with the situation, they did attempt to persuade management to act appropriately. They then distanced themselves from the conduct that misled the external actuaries. In those circumstances I do not find there might have been a breach of the law in relation to the conduct ofMcCarthy and Trahair.

There was evidence that in June 1998 Peiris attended a meeting with Wilkie, Trahair and McCarthy at which the question of under-reserving in the order of tens of million s of dollars was raised. Peiris said he understood the reserving shortfall had been taken up in the work the actuaries have done on the professional indemnity business. 147 There was also evidence of a file note, created by Robert Martin in relation to a meeting with Peiris on 31 October 2000148 , that recorded Peiris having informed Martin he understood there may have been a reserve shortfall of between $50 million and $60 million. Peiris was not cross-examined in relation to that meeting, but he did provide a statement in which he said he told Martin, during a meeting with him in late October or early November 2000, that he was unaware of any under-reserving in the June 1998 accounts ofF Al. 149 He said he also told Martin that F AI booked reserves according to the actuaries' advice and he had no specific knowledge of under-reserving after closing the 30 June 1998 accounts.

Any finding against Peiris would rely significantly on the contents of his conversation with Martin and his state of knowledge of the under-reserving in F AI as at 30 June 1998. The question of what was discussed at the meeting between Martin and Peiris was not fully explored in the oral evidence. In those

circumstances, I do not find that Peiris might have contravened a law in this regard.

14.1.10 Non-disclosure of reserve shortfalls to the board

Shortly after 13 March 1998 Wilkie reported to Adler that the reserves F AI held were inadequate to the extent of $50 to $60 million. 15 0 Adler did not raise this

130 Th e impact of the FAJ acquisition

shortfall at the 7 April1998 board meeting.151 He testified that he did not inform the board of the matter because he had received from Wilkie an assurance that reinsurance would resolve the problem, so he considered there was no need to bring the matter to the board's attention. 152 By March 1998 Wilkie had, to Adler' s knowledge, begun negotiations with General Cologne Re for certain reinsurance cover. Adler agreed that the underwriting loss for the CPID, as referred to in his reports to the board in March and April 1998, did not take into account the shortfall and that to that extent the monthly accounts for the CPID misrepresented the current performance of the division; he said he was aware of that at the time.153 Although Adler considered the shortfall would be covered by the GCRA contract, he agreed that the problem had not been resolved during the period covered by the management accounts for the months of March and April 1998. 154

Adler submitted that he considered the problem was a developing one and thought that a reserve shortfall had not yet crystallised; in other words, it was Adler' s understanding that there was a risk of future adverse claims development, which could be remedied by means of a reinsurance solution. In a report he presented to the board on 10 June 1998 in respect of the month of April 1998, Adler stated,

'There is little doubt that F AI, as a company, will continue to suffer from the pre-1991 written business due to the claims experience that is developing' .155 He said that he intended to convey the idea that F AI would suffer losses post 30 June 1998 as a result of the development of claims experience for business written before

1991. 156 In those circumstances, Adler must have understood F AI was likely to suffer losses as a result of development of claims experience because F AI did not hold adequate reserves in respect of pre-1991 business. In a related submission Adler claimed that in late March 1998 he understood the problem was a prospective

one rather than historical. I confess I do not understand Adler's point because what is being considered is the adequacy of reserves for outstanding claims liabilities. By their very nature, outstanding claims liabilities are prospective-in the sense that payments must be made in the future in respect of events that have occurred in the

past. I think Adler's evidence, properly understood, was to the effect that he was aware that F AI' s reserves for business written before 1991 were inadequate to the extent of $50- $60 million and that losses would result. It was this that he failed to bring to the attention of the board following Wilkie's advice in March 1998.

Adler submitted that there was substantial over-reserving in other portfolios in the company, most notably the compulsory third party portfolio. He stated there was a buffer in excess of $30 million in the reserves for that portfolio--a statement supported by a memorandum from Andersen dated 27 August 1998.

157 In my

opinion, even if Adler believed there was an excess of reserves in the CTP portfolio, he had no grounds for assuming there was no under-reserving in any other area of FAI's operations. As a consequence, he had no grounds for assuming the $50-60 million problem about which Wilkie had informed him could be partially offset

by the over-reserving in the CTP portfolio.

Adler submitted that in late March 1998 Wilkie advised him the $50-60 million problem could be resolved by entering into an appropriate reinsurance agreement

The failure of HJH Insurance 131

and that the deal with GCRA was all but done. Adler further submitted that it was reasonable for him to treat the issue as one that was unlikely to become a serious problem because it would probably be fully resolved within a matter of days or weeks, but it had the potential to develop into a broader problem if not resolved. He submitted that the situation called for the exercise of judgment and discretion when reporting to the board. He also referred to the evidence of John Landerer (chairman of the F AI board), who agreed that Wilkie was giving information to the board 'that things are bad, we could see they were bad, but our staff were working on ways to try and stem that loss' .158

Adler also referred to the evidence of Mainprize, who said that in the period leading up to June 1998 Adler informed the board that losses were being incurred as a result of the long-tail nature of the CPID business. He submitted that it would not be in accordance with either the law or the facts then prevailing to have expected him to say more to the board about these matters because that would necessitate going beyond the realm of the legal obligations of a director and into the area of the exercise of commercial judgment. Finally, Adler submitted that the most serious allegation that could be made against him was that he ought to have realised that the April1998 management accounts that were provided to the board were misleading because by then he had been told about the shortfall and should have realised the problem was not yet solved because the GCRA contract had not been signed. In responding to that allegation, Adler stated that he believed the shortfall had in fact been taken up in the March, April and May 1998 accounts. He did not specifically check whether that had occurred, but this was not a matter with which a chief executive need concern himself.

In my opinion, the matters Wilkie brought to Adler's attention required Adler to put the board on notice. It was for the board to consider how the problem should be dealt with and for it to determine whether disclosure needed to be made to the market and how the inadequacy of reserves might affect reported profit. In failing to bring the matter to the board's attention Adler denied the board the opportunity to consider those matters. In his reports to the board dated 7 April and 10 June 1998, Adler referred in general terms to deterioration in claims experience. The board was generally aware that claims experience was deteriorating in the commercial lines, but Adler did not adequately disclose the existence of a significant issue that he had been aware of since at least March 1998. In his 7 April 1998 report to the board, he advised that there was no need to make any disclosure to the market and confirmed his 1998 profit forecast of $20 million. 159 That report must have suggested to the board that, although there was deterioration in the claims experience for the commercial portfolios, it was not such that it would affect the confirmed profit forecast for 30 June 1998. The GCRA contract was not entered into until 6 May 1998 . At the time of writing his to the board, Adler could be confident that the forecast 1998 profit of $20 million would be achieved only if he ignored the problem Wilkie had drawn to his attention.

In my opinion, it is possible to make a finding that Adler should have informed the board of the $50- 60 million shortfall in reserves in the period March to June 1998

132 Th e impact of the FAI acquisition

without intruding into an area that is no more than the exercise of commercial judgment. A reserve shortfall of that magnitude would probably have had major ramifications for F AI: it was plainly a matter that would have been of interest and concern to the board.

It is my view that Adler refrained from informing the board of the matter because he was conscious of the impact such a shortfall may have on the company's reported profit. He might also have been embarrassed-as he testified- at constantly having to go to the board, alerting them to a problem but adding that it was in hand, then

later having to go back to the board to tell them the problem was worse than anticipated or the putative solution had not been effective. But that is not relevant. If information is material it should be disclosed to the board, and this information was plainly material.

The conclusion that Adler was motivated by his appreciation of the effect of the shortfall is reinforced by the evidence in relation to a request by the board chairman, John Landerer, at the 10 June 1998 board meeting. He asked for a detailed report on the results of the CPID together with a forecast for the development of the tail. 160 In

response, Adler advised the board that, as part of the remedial action to improve the CPID's performance, an actuary specialising in long-tail classes had been employed. This was presumably a reference to Trahair, who had been retained by F AI in February 1998. Adler conceded that he failed to tell the board an actuary had

reported in March 1998 on the forecast development of the tail. 161 He also agreed that, following the meeting of 10 June 1998, he did not ask Wilkie for a copy of Trahair's report or ask Wilkie to provide a copy ofthe report to the chairman, as had been requested at the meeting of 10 June 1998.162

At the next board meeting, on 14 July 1998, Adler informed the board the report on the tail was still being prepared 163 , and Wilkie said a report would be made available when it was completed. 164 Adler agreed it was misleading to advise the board in July 1998 that the report requested in June 1998 was still being prepared, without disclosing that it had been finalised in March 1998. 165

The evidence relating to the chairman's request for a report and Adler's failure to disclose its existence points to Adler withholding important information from the board, thereby denying the board the opportunity to consider the contents of the report, its conclusions, and what steps should be taken to deal with the inadequacy of reserves. Had Adler understood the problem in the way submitted by him, there would have been no reason to withhold the report or its conclusions from the board. In the absence of a rational explanation as to why Adler and Wilkie withheld from

the board information about the $50-60 million shortfall in reserves

during March to June 1998, I infer it was because of the likelihood that such disclosure would have had significant consequences for the accounts and reported profit as at 30 June 1998.

Wilkie's reports to the board between March and July 1998 made no reference to the level of under-reserving of which he had been aware since late 1997. 166 Wilkie submitted he believed Adler had informed the board of the level of under-reserving

The failure ofHIH Insurance 133

and that there was no basis to suggest that Wilkie knew Adler had not informed the board of the true position during 1998.

I am not persuaded by those submissions. First, the fact Adler told Wilkie that he (Adler) would deal with the matter was no justification for failing to inform the board of material information. Second, it is difficult to understand how Wilkie could have reported to the board properly on the CPID's performance without revealing the level of under-reserving that had been reported to him. Third, Wilkie attended board meetings to report on the performance of both the general insurance division and the CPID. If Adler had informed the board of the levels of under-reserving one would have expected the directors to question Wilkie in relation to that matter and its effect on the performance and profitability of the CPID. None of the board minutes reflects any such discussion, and the F AI directors who testified said no such discussion took place and they were ignorant of the matter. Fourth, as noted, on 10 June 1998 the chairman of the board asked for a report on the results of the CPID, together with a forecast for the development of the tail. At the board meeting on 14 July 1998 Wilkie informed the board that a report would be made available when it was completed. Yet the evidence established that Wilkie was aware Trahair had prepared such a report in March 1998. In my opinion, that evidence points to Wilkie's awareness that the board had not been informed of the true position and he took no steps to disclose to the board information of which he had been aware since

late 1997.

Scrivens testified that during the June 1998 annual audit he had discussed with Wilkie and Mainprize the adequacy of FAI's reserves. Both confirmed the reserves were adequate. 167 Wilkie agreed he never alerted the auditors to the fact that a shortfall in respect of the Big Six and ALAS had been identified. He had met the auditors in August 1998 to discuss reinsurance contracts but he did not think it necessary to raise the shortfatl with them. 168

In an email dated 20 March 1998 Lindsay Self, a reinsurance manager employed by GCRA, told other GCRA employees of a telephone conversation he had had with Mainprize and Wilkie. In the email Self refers to 'they' (that is , Wilkie and Mainprize) confirming that additional reserves would be taken up by F AI over three years, totalling $65 million. 169 On 24 March 1998 Mainprize sent a letter to Self in relation to the Big Six and ALAS accounts, confirming that the 'expected run-off losses' for those accounts over the next three years were $25 million, $20 million and $20 million. 170 Self's email and Mainprize's Jetter are discussed in Section 14.2.2. On the basis of the email and the letter, I conclude that Mainprize was aware that FAI was substantially under-reserved and that either he or Wilkie (with the other's knowledge) advised Self of the matter and that the under-reserving would only be addressed over a three-year period.

I make no finding that Adler was aware that the published accounts of F AI as at 30 June 1998 were false by reason of under-provisioning, even though Wilkie had told him in March 1998 of $50- 60 million under-reserving problem. I make no finding because Adler testified that he understood the shortfall in F AI' s reserves had

134 Th e impacl of the FAJ acquisition

been accommodated by the reinsurance contracts entered into with GCRA. 171 It would still be incumbent on management to book the shortfall and then deal with it by recoveries under the reinsurance arrangements. There is no evidence to suggest that Adler was aware that this course had not been followed.

Similarly, in the case of Peiris the evidence does not permit a finding that he understood that F AI was significantly under-reserved at 30 June 1998 . Consequently, I make no finding that Peiris was aware that the published accounts ofFAl at 30 June 1998 were misleading.

The position in relation to Mainprize is different. I have already found that he was aware ofthe $50-60 million under-reserving. He was also aware in March 1998- as the letter he signed and which is addressed to Self attests-that FAI's intention was to take up the additional reserves in a staged manner. If he continued to be ofthat

mind, the existence of the reinsurance contracts would not affect his state of knowledge in the manner described for Adler. It is not clear on the evidence whether Mainprize understood through to the time when he signed the accounts as at 30 June 1998 that the identified shortfall in reserves was to be taken up in stages, rather than immediately, as required by Australian accounting standards. On that

basis, I make no finding that Mainprize was aware that FAI's published accounts were false by reason of under-provisioning.

14.1.11 The level of under-reserving after 30 June 1998

A summary of solvency reports filed by F AI with the Insurance and Superannuation Commission (and subsequently APRA) for June 1997, December 1997, June 1998 and December 1998 disclosed that FAI's solvency for Insurance Act purposes was marginal. 172 If F AI' s provisions were understated by even a modest amount, other things being equal it would have breached the minimum solvency requirements. The evidence also establishes that APRA considered F AI was under-provisioned by at

least $40 million from at least 30 June 1997. 173 In the light of those factors , a significant shortfall in reserves was likely to have a significant impact on the FAI's solvency for the purposes of the Insurance Act, and this should have heightened the importance given to estimating outstanding claims liabilities as accurately as possible.

In the months following 30 June 1998 Wilkie's reports to the board continued to make no reference to under-reserving in F AI. 174

On 26 November 1998 Spratt sent to Wilkie an email in which he stated,

.. . My object is to give you a brief summary of the numbers involved and illustrate that we have the capacity to give our new owners a detailed briefing on these issues straight away. This will be complete by 1st December when I have a regular review meeting with you. I propose to bring Peter Moran to that as he has day to day handling of several of these

issues ... 175

The failure ofH!H Insurance 135

On the same day he prepared a memo to Wilkie. 176 In the memo he concluded that in the international professional indemnity portfolio there was under-reserving of $75 million, plus or minus $20 million. Moran testified that he and Spratt met with Wilkie on about 26 November 1998 and discussed the memo. 177 Wilkie confirmed that he had a meeting with Spratt and Moran on about 1 December 1998. He said he read the email from Spratt but was not of the view there was a shortfall in the range of $75 million, as concluded by Spratt. 178 He said there was considerable doubt about the whole range of classes referred to in Spratt's report, and he doubted some of the recommended reserves, although he agreed that Spratt had more information about these claims than he did. 179 Wilkie said he took no steps to draw the board's attention to the $75 million shortfall because any increase in reserves had no impact on the profit and loss since it was offset by reinsurance recoverables. 180

Spratt said he sent his email of 26 November 1998 to Wilkie because he suspected HIH was not aware of the magnitude of the reserving shortfalls and that it was highly probable that for the purpose of its takeover bid HIH was relying on F AI's accounts as at 30 June 1998, which he believed to be misleading. 181 In other words, he was trying to persuade Wilkie to give HIH full and frank disclosure of the under­ reserving that was the subject of Moran's investigation. That proposal was rejected by Wilkie, with the consequence that HIH proceeded to acquire F AI without any forewarning of the significant under-reserving within F AI. At all material times, and especially following HIH' s announcement of the takeover ofF AI, the board ofF AI had a duty of continuous disclosure, which heightened the need for material information to be disclosed to the board so that the directors could decide what disclosure, if any, should be made to the market. Wilkie's failure to inform the board of the under-reserving deprived the board of the opportunity to consider that matter.

14.1.12 The level of under-reserving in FAI

Estimates of the proper level at which the reserves in F AI should have been recorded in its accounts can confuse because they are done at different times and for different purposes. This section is concerned with estimates of the level of outstanding claims liabilities, at different reporting dates up to and including 31 December 1998, based on case reserves reported at the time.

In evidence, the expert the Commission retained disclosed that there . were significant shortfalls in case estimates in respect of certain lines of account written by F AI at various report dates. In the case of the Big Six, ALAS, MDU and Crane accounts under-reserving in case estimates were as follows

182 :

• For the reporting date 30 June 1997 the FAI case reserve was $42 326 000 and the recommended case reserve was $150 476 000, producing a shortfall of $108 150 000. This figure was mistyped in WITS.0095.001 at 013.

136 The impact of the FA! acquisition

• For the reporting date 31 December 1997 the F AI case reserve was $33 546 000 and the recommended case reserve was $132 837 000, producing a shortfall of $99 291 000.

• For the reporting date 30 June 1998 the FAI case reserve was $33 518 646 and the recommended case reserve was $131 826 859, producing a shortfall of $98 308 213 .

• For the reporting date 31 December 1998 the F AI case reserve was $31 296 03 7 and the recommended case reserve was $115 548 000, producing a shortfall of $84 252 963.

It should be noted that the shortfall figure for 31 December 1997 does not include a shortfall in case estimates for other accounts totalling $24.78 million, giving a total shortfall of $124.071 million. The estimates are also undiscounted and gross of any reinsurance. Further, with the exception of figures notified in the ALAS bordereaux, which included an adverse development factor, IBNR is not included. 183

In an attempt to ascertain the extent to which IBNR would need to be added to the shortfalls, a report was prepared by Estelle Pearson. She calculated the outstanding claims liabilities for the Big Six and MDU as at 30 June 1997 , 30 June 1998 and 31 December 1998 had PWC been provided with the level of case reserves just shown. 184 Pearson's report was not disputed by any party, and there is no reason to

doubt the methodology she used in her report or her conclusions.

She indicative assessment suggested that higher case reserves would have led to an increase in the outstanding claims liability of $53 to $76 million at 30 June 1997, of $45 to $64 million at 30 June 1998, and of $30 to $39 million at 31 December 1998 . If those figures are added to the shortfall figures just listed we arrive at the following reserve shortfalls: at 30 June 1997, $161 to $184 million; at 30 June 1998, $143 to $162 million; and at 31 December 1998, $114 to $123 million.

Pearson also considered the impact of the higher case reserves on the IBNR assessment in respect ofMDU at each balance date. 185

Pearson concluded that the higher case reserves for MDU would have led to an increase in the outstanding claims liability of approximately $25 million at 30 June 1997, approximately $28 million at 30 June 1998, and approximately $32 million at 31 December 1998.186 If one adds these figures to the calculations

listed before the shortfalls are as follows:

• $186- 209 million at 30 June 1997

• $171 - 190 million at 30 June 1998

• $146-155 million at 31 December 1998.

PWC's Latham said that, had PWC been provided with genuine case estimates, their valuation would probably have been at the higher end of the range reported by Pearson, which he considered reasonable. 187

The failure of HJH Insurance 137

The effect of this evidence is that when HIH announced its offer to acquire F AI the latter was under-reserved in respect of the Big Six, ALAS, MDU and Crane to the extent of almost $200 million. Had a reserve shortfall of this magnitude been taken up in the 30 June 1998 accounts ofFAl, it would have had a very serious impact on reported profit. It could also have placed F AI General in breach of the solvency requirements of the Insurance Act and raised questions as to FAI's solvency.

14.1.13 The Commission's processes

Most of the parties against whom findings are made in this chapter complained about lack of procedural fairness. The answer to those complaints lies in the general comments made in Section 1.5 of this report. There is nothing to add here.

Kamha submitted that the matters put against him were outside the Commission's terms of reference because nothing he did caused the collapse of HIH. It is correct that the individual events did not themselves 'cause' the collapse of HIH, but in my view there is a connection to the core instruction that brings them within the terms of reference. The acquisition of FAI was a (not the) reason for the failure of HIH. To appreciate the causal significance of the F AI acquisition, to the collapse, it is necessary to understand the financial position of F AI. That, in tum, requires an appreciation of the outgoing claims provision in F AI' s financial statements at various points leading up to 30 June 1998. The events discussed in this section are germane to such an appreciation. Questions of degree certainly intrude. But they are matters of judgment. It is proper to regard the events in question as coming within the terms of reference.

14.1.14 Conclusion

In their closing submissions on this subject counsel assisting said,

The reserving practices within F AI over a long period of time were unsatisfactory giving rise to significant under-reserving from at least 30 June 1997 and possibly before that period. The causes of the under­ reserving were multi faceted. They included poor claims management but also deliberate manipulation of claims estimates by management for the purpose of improving reported profit as at various report dates. Further,

once the extent of the under-reserving was understood by management, which seems to have occurred in the last quarter of 1997, steps were taken to conceal the under-reserving from the Board, the auditors, the external actuaries and APRA amongst others. That deception continued until shortly after the acquisition ofF AI by HIH in the circumstances referred to above. 188

I adopt that description. Failure squarely to address endemic provisioning problems was a serious deficiency on the part of management ofF AI. It had a direct impact on the financial health of the company and on the way in which its financial condition was reported. As is discussed in Chapter 17, HIH had similar problems, and the under-provisioning problems and practices it inherited from F AI interacted with and compounded those it was already experiencing . The result was catastrophic.

138 Th e impact of the FAJ acquisition

14.2 The negotiation of the GCRA contract

I have already found that in late November or early December 1997 Wilkie was approached by Karnha and provided with a document summarising a reserving shortfall in the international professional indemnity portfolio of approximately $112 million.

In this chapter I deal with the efforts Wilkie and others in senior management at FAI made in the period from December 1997 until the end of June 1998 to obtain reinsurance to deal with its provisioning problems. F AI management sought to utilise reinsurance to deal with under-reserving because they wanted to offset any increase in reserves on the balance sheet with a corresponding recovery under a reinsurance contract. They believed that the accounting treatment for reinsurance would, or at least might, allow them to defer expensing the premium paid to obtain the recoveries to later years. Thus it was thought that the effect of an increase in reserves could be staged over a number of years ('smoothing').

There is a problem with this approach. 'Smoothing' was arguably possible under the accounting standards only if the contract provided for a 'transfer of risk' from the reinsured to the reinsurer. But how could this be done with a known deterioration in prior years reserves? Would a reinsurer provide cover for a loss that had already crystallised? Would F AI recognise all of the past under-reserving or only some of

it? How would F AI deal with its auditors to ensure it all passed muster? The answers emerge from the matters discussed in this chapter.

In December 1997 F AI management first approached the brokers Willis Faber Dumas Limited (Willis) but negotiations for a contract were unsuccessful. They then negotiated an aggregate excess-of-loss contract (the first AXOL contract) with General Cologne Re Australia (GCRA). This contract was renegotiated in June 1998 (the second AXOL contract). In June 1998 FAI entered into a further whole-of­ account reinsurance contract with National Indemnity (Berkshire Hathaway) (the NI contract) which was brokered by Guy Carpenter. Each set of negotiations needs to be considered in turn.

14.2.1 Negotiations with Willis

Wilkie met Stuart Stow, Willis ' s Managing Director, on 5 December 1997. According to Stow, Wilkie advised him that it was quite possible that claims reserves for casualty portfolios would have to be increased, that F AI would be undertaking a full review of that portfolio and that Wilkie was seeking a reinsurance product that would enable the effect on F AI' s balance sheet to be spread over a

number of years in the future. 189 Wilkie did not deny that these matters were discussed. 190Stow relayed the effect of his discussions with Wilkie in a facsimile to Willis Faber's New York office. 19 1 In relation to the reinsurance product that Wilkie was seeking Stow stated:

Th e failure of HIH insurance 139

[The client] accepts that there is charge to be levied for accessing the balance sheet of another insurer/reinsurer but he would prefer that there is no risk transfer whatsoever.

Wilkie's evidence was that in this discussion he conveyed to Stow the need for 'minimal risk transfer' .192

On 9 December 1997 the New York office replied to Stow's facsimile. 193 They advised that without transfer of risk the desired accounting treatment would not be possible. They also proposed a 'quasi financial transaction' which might be able to be accounted for as conventional reinsurance and which provided two sections of cover. Section A provided $50 million cover for the anticipated adverse development in reserves that was to be fully funded by the client's premium payments. Section B provided for $20 million cover which was to have a retention that meant it was unlikely to be invoked but which might nonetheless create risk transfer. This structure of a reinsurance contract and variations upon it were floated by and with F AI on a number of subsequent occasions.

On 9 and 12 December 1998 Stow met with Wilkie and most likely FAI's reinsurance manager, Stephen Burroughs. 194 Following the meeting he sent a further facsimile to the New York office. 195 In that facsimile Stow stated:

Using figures as per your facsimile, ideally the company would like to recognise the potential for $50 million worth of reserves to be transferred by virtue of a reinsurance contract where they pay a premium of say $40 million over a ten year period which should, after investment income, fund the $50 million of transferred alibility [sic]. The reserves strengthening could be stage managed over a two year period say realising

$35 million in year one and $15 million in year two to avoid the likelihood of external awareness that the contract was specifically geared to pick up the full liability.

If it is necessary to effect a degree of risk transfer, would it possible for said risk transfer to be effected by the general insurance company and have a from the holding company to ease the way behind the

scene? 1 6 [emphasis added]

The references to 'stage managed reserves', 'external awareness' and '[easing] the way behind the scenes' are interesting. Similar concepts recur in later events.

Despite Wilkie's denial 197 and Burroughs assertion that he could not recall 198, I accept that the matters referred to in the above extract were probably discussed with Wilkie and possibly Burroughs. I consider it unlikely that Stow, who was primarily a conduit of information, would have initiated such matters himself.

On 14 January 1998 Burroughs met with Stow and provided him with some financial information concerning F AI. 199 They discussed the portfolios needing protection. Burroughs provided Stow with some information about under-reserving

in MIPI and ALAS.

140 The impact of the FA! acquisition

On 9 February 1998 Stow wrote to Burroughs' 200 setting out his understanding of FAI's 'exposure' (being a reserve shortfall of $67.5 million) and the outline of a proposal for a reinsurance deal. Stow referred to a staging of the increase in reserves over three years, compared with the previous two, using a 'finite risk policy'.201 He described its structure which was in terms similar to that previously proposed. It would include a section to 'alleviate any risk transfer concerns from any outside commentary' and 'designed to attach in excess of any possible claim penetration' and 'never meant to actually pay a claim'.

Burroughs replied in a letter dated 19 February 1998? 02 He confirmed Stow's understanding that F AI was proposing to increase reserves in 'somewhat of a staged manner'. Burroughs letter also disclosed in the latest figures a reserve shortfall in F AI's PI portfolios. On the same day that Burroughs wrote his letter FAI released its

half year results for the six months ended 30 December 1997.203 Those results did not take into account the problems with reserving identified in Burroughs' letter.

Stow conveyed Burroughs' s letter of 19 February 1998 to Clark Hontz describing it as 'authority to go ahead' 204 and commence discussions with reinsurers in relation to the transaction. Burroughs instructed Stow that the deal had to be placed with reinsurers with whom F AI did not have an existing relationship? 05 Thereafter negotiations were conducted with various reinsurers. Only the negotiations with Hannover Re developed to the point where there was any real possibility of a deal.

On 17 March 1998 Henning Ludolphs from Hannover Re faxed a draft slip to Tom Cochrane of Willis which was copied to Stow. 206 The draft slip provided for three sections of cover. This slip was discussed at a meeting between Wilkie, Burroughs and Jurgen Graeber from FAI along with Cochrane and Tim Thomas from Willis at Heathrow airport on 25 March 1998. Cochrane's note of that meeting207 records

Wilkie as having suggested again that the identified under-reserving not be brought to account immediately and instead be 'spread' across five years and only recognised in the profit and loss account as claims were paid. Wilkie denied suggesting this?08

Although there were further negotlatwns concerning the Hannover Re contract being brokered by Willis they appeared to have been exhausted by approximately mid April 1998.

14.2.2 Negotiations with GCRA to 24 April 1998

Early in 1998 F AI commuted a whole-of-account excess-of-loss contract which it had with GCRA and which pre-dated 1998. In the course of arranging the commutation Wilkie advised GCR that F AI was talking to reinsurers about some sort of financial reinsurance quote and he would advise GCRA of their plans in a couple ofweeks.

Several officers or employees of GCR and GCRA were involved in the negotiations which followed. For GCRA the personnel involved were Geoffrey Barnum, its managing director, Christopher Byatt, chief financial officer, Self, general manager

Th e failure of HIH Insurance 141

treaty and Andrew Smith, a treaty manager. Representatives of the 'alternative solutions division' of GCR who were involved from time to time included Tore Ellingsen, Milan Vukelic, John Houldsworth and John Byrne.

On 18 March 1998 a meeting was held between Wilkie, Self, Vukelic and Ellingsen. The meeting was followed by a lunch where they were joined by Mainprize?09 According to Self, during that meeting and possibly the lunch, Wilkie expressed an interest in a product that would protect FAI in relation to MIPI 'in case' the reserves were insufficient. When pressed on the extent of their 'potential inadequacy', he stated that it was 'in the order of $65- 80million'.210 There is no evidence that Mainprize was present when Wilkie discussed these matters.

On 20 March 1998 Self sent an email to a group of GCR and GCRA staff-Q 11 reporting on a conversation he had with Wilkie and Mainprize. According to the email, amongst other matters, they stated that they wanted 'the agreement' confirmed by the morning of Wednesday 26 March 1998. They confirmed that 'the reserves will be taken up by them over three years and [were] likely to be ' $25 million in year 1, $20 million in year 2 and $20 million in year 3- a total of $65 million. The email also said that there were only five people at FAI-who would know of the transaction: Adler, Mainprize, Wilkie, Burroughs and Peiris. Self confirmed that the persons at GCR and GCRA who would know of the transaction included Vukelic, Ellingsen, Self, Smith, Houldsworth, Barnum and Byatt. Wilkie agreed that these matters were discussed although he denied that any staging of an

increase in reserves was proposed by him. He said it could have been mentioned by Mainprize. 212 Mainprize denied making any such suggestion. 213 For the reasons set out below I have difficulty with Wilkie's and Mainprize's denials.

As previously mentioned, on 24 March 1998 Mainprize sent a letter214 to Self in relation to 'Big 6 and ALAS' confirming that the 'expected run-off losses' for those accounts over the next three years ( 1998/99 to 2000/01) was $25 million, $20 million and $20 million '. Smith relayed this letter to Ellingsen. 215

I have already concluded that Wilkie was aware that F AI was substantially under reserved. I am satisfied that Mainprize was also aware that F AI was substantially under reserved. Further, that one of them, in the presence of the other, advised GCRA that F AI proposed only to increase reserves over time rather than bring the deficiency to account immediately. In Wilkie's case, Selfs email is consistent with the proposals recorded by Stow from Willis following his meeting with Wilkie in December 1997 and with Cochrane's note of the Heathrow meeting concerning Hannover Re described above.

In the case of Mainprize, there is Self s email and the letter of 24 March 1998 which he signed. Mainprize pointed to the fact that the letter of 24 March 1998 appears to have been drafted within the CPID section of F AI and that the phrase 'run-off losses' does not necessarily convey a statement of under-reserving. Rather, it could be a reference to cash flows. He contrasted it with Burroughs's letter of

19 February 1998 to Willis which explicitly refers to under reserving. In effect Mainprize contended that CPID was not keeping him 'in the loop'. Whatever Wilkie

142 Th e impact of the FA! acquis ition

may have conveyed to Self on 20 March 1998 in relation to the reserve increase Mainprize did not understand it as referring to under reserving which would only be brought to account over three years.

I accept that the letter of 24 March 1998 can be read in the manner suggested by Mainprize if the phrase 'expected run off losses' means expected payments. But such payments are not 'losses' if provision has been made for them. If provision had been made for $65 million of future payments (which it had not) there would be little point in F AI seeking another $65 million of cover. The evidence reveals that Mainprize learnt that F AI was paying at least $55 million in premium for $65 million in cover. I have difficulty understanding why F AI should pay such a premium if it had already made provision for the 'run off losses' . In addition, Mainprize was aware that F AI was seeking protection over and above existing reserves and the ultimate amount of cover obtained was $65 million. If he read the

letter as a reference to cash flow it must have seemed an unusual coincidence that cash flow out of existing reserves would be $65 million and the cover obtained under the contract above existing reserves was also $65 million. In fact existing reserves were approximately $17.5 million.

On 25 March 1998 Ellingsen sent, as an attachment to an email, a draft slip to Barnum, Self, Smith, Houldsworth and Vukelic. 216 He stated that it would not be apparent 'at first glance' how the slip worked, but that it would be cashless and that a key factor was the offset clause. The draft slip proposed that there be three sections of cover: one for year 2000 risk, one for professional indemnity and one for non-recoverable reinsurance. The overall aggregate limit was to be $65 million. The premium (described as 'basic premium') was payable over five years and totalled

$55 million. The entirety of the premium was to be held on deposit by the reinsured on behalf of the reinsurer until 1 July 2002. The reinsurer was not to be obliged to make cash payments before 1 July 2002 .

Smith stated that, around this time, Ellingsen and Houldsworth explained that the $10 million 'gap' apparent on the face of the contract was to be made up by other profitable (for GCRA) reinsurance business to be written at the same time. 217 He stated that it was envisaged at this stage218 that this other business was to be risk bearing. I think this explains the significance of the offset clause. Without it GCRA would have been exposed to a known loss of $10 million on the contract. It was

because of the other business that the structure of the transaction was not 'apparent at first glance'.

On 26 March 1998 Self and Smith met Mainprize. According to an email Smith sent soon after the meeting to Ellingsen219 he 'explained the bare bones structure with him'. Mainprize recalled attending a meeting where Self discussed a proposal for a cashless transaction. 220 I think it likely that Smith discussed the slip with Mainprize

(which provided for $55 million in premium and $65 million in recoveries) and advised him that there would be 'other business ' .

Later that day Ellingsen and Houldsworth met Wilkie and Burroughs in Cologne. Wilkie agreed that the draft aggregate excess of loss contract was discussed and

The failure ofHIH Insurance 143

conceded that it was possible that Ellingsen mentioned writing 'other business' associated with the AXOL contract. 221 I consider it likely that Ellingsen, having conceived a structure for the transaction which involved F AI agreeing to cover the 'gap', would have conveyed that to Wilkie.

A video conference was held between at least Smith, Barnum, Vukelic and Ellingsen around 2 April1998.222 During the video conference someone said that without being satisfied on the adequacy ofF AI' s reserves, the matter could not be taken further. 223 I infer from this and the fact that GCRA had sought advice from external .solicitors concerning the enforceability of the offset clause224 that there was a concern that if F AI' s reserves were inadequate there was a risk of F AI' s insolvency. In such an event GCRA might be exposed to having to pay claims but be unable to receive the premiums.

On 8 April 1998 Burroughs delivered to Self a document entitled 'Spread Loss Reinsurance Cover' with an accompanying spreadsheee25 , both prepared by Burroughs226 , together with a note prepared by GeoffTrahair. 227 Trahair's note identified a reserving shortfall for professional indemnity for $91.9 million. Burroughs referred to the shortfall being 'recognised' over three years. The spreadsheet supplied by Burroughs mapped the effect of the policy and the increase of reserves (staged over a number of years) to F AI' s balance sheet. Burroughs included a proposal for a policy. It was similar to those suggested to him by Willis, namely a section of cover designed to take up adverse development but which was subsumed in a policy with a number of other sections which might have conveyed the appearance of risk transfer but which were not intended to pay claims. In an email sent on 9 April 1998 Burroughs advised Wilkie that this information had been provided to Self.

On 16 April 1998 Smith sent to Burroughs the latest draft of the slip under cover of a facsimile stating that it 'shows the bare bones structure as promised'.228 His fax stated that the main points on the slip at that stage were the 'transaction method and the sudden death/cancellation/offset and insolvency clauses'. Further, it was up to

Smith and Burroughs to add the detail on things such as the underwriting and premium level. The slip was very similar to what had been shown to F AI in late March 1998. On the same day Smith sent Burroughs another fax stating that, before any final proposal was made, GCRA would like to review underwriting and claims files.

From 20 to 23 April1998 Smith and initially Self, then Ellingsen229 conducted that review at the offices ofF AI. They were provided with information and assisted by Trahair amongst others. 230 Ultimately, they produced a report.231 It stated that the object of the exercise was to determine whether F AI would be able to meet its obligations to GCRA under the proposed transaction. 232 I have already noted GCRA' s conclusions about deficiencies in F AI provisioning practices. By considering information that GCR had obtained directly from the relevant brokers overseas the report concluded that F AI had a shortfall in the relevant portfolios of about $110.78 million. The FAI internal studies had come up with a reserve

144 The impact of the FAJ acquisition

shortfall totalling $91.9 million. 233 The GCRA report concluded that even with a shortfall of $110.78 million FAI had a conservative net asset surplus of $73 million. 234

According to Self, after the due diligence he attended a meeting with Ellingsen, Barnum, Byatt and Smith235 in which Ellingsen advised that the alternative solutions division would not proceed with the transaction unless there was an 'arrangement of $12.5 million additional premium that ... would not be affected by claims'. Self stated that he queried Ellingsen on the fact that there would be no claims attached (as it meant that there would be no risk attached to that business) but was told 'don't worry about it'. 236

GCRA submitted that Selfs references to 'premium' being claims free in this meeting, and the meeting with Wilkie and Mainprize which followed, was somehow different to business or contracts being claims free. I have difficulty with this submission. The supposed distinction was not explained and I was unable to see any foundation for it in the evidence. Otherwise I infer that, although the due diligence

reported F AI as having a net asset surplus of just over $73 million, Houldsworth and Ellingsen were concerned about FAI's financial strength and exposure to claims generally. They insisted on a transaction in which there was no risk of any loss to GCRA, including any risk that might flow from the insolvency ofFAL

Self told the Commission that he then attended a meeting with Ellingsen, Wilkie and Mainprize at FAI's offices sometime between 23 and 25 Aprill998.237 According to Self, Ellingsen explained to Wilkie and Mainprize238 that, as part of the transaction, GCRA would require payment of $12.5 million as premium under other policies in respect of which there would no claims. 239 Self stated that Wilkie and Mainprize said they would need to consider Ellingsen's proposal and that he (Self) later heard that they had accepted. 240 Wilkie recalled attending a meeting around that time at which Ellingsen made reference to or 'introduced' the 'concept of additional premium through other insurances' but he could not recall whether he said this business was to be 'free from loss recoveries' .241 He stated that it was his

understanding that this additional business was to be 'risk bearing and also with a low likelihood or prospect of claims that was made' and that he and Mainprize were both 'surprised at this particular request being made' ?42 Mainprize's recollection was consistent with Wilkie's.243 He denied that there was any mention of any

business being 'claims free'. 244

I accept Selfs evidence that, at this meeting, Ellingsen introduced the need for the other business be 'claims free'. In my view the evidence points to the concept of additional business to make up that gap having already been introduced by this stage. It would have been obvious that GCRA would want it to contain minimal

risk. Selfs evidence that at this meeting Ellingsen outlined the need for there to be 'claims free business' is consistent with Wilkie's lack of complaint when he was asked by Self some days later to sign a letter to the effect that there would be 'no claims' under the six 1 May 1998 contracts. 245

The failure of HIH Insurance 145

14.2.3 Finalising the slip and finding the 'other business'

On 27 April 1998 Burroughs sent a facsimile to Smith246 headed 'Supplementary Classes Under Spread Loss Cover' in which he listed 14 types of cover and suggested a meeting to 'structure these on a whiteboard' for the following morning. Smith divided them into two groups. He isolated six contracts which were ultimately to be the cover provided for by the 'other business' (that is, the six 1 May 1998 slips) and identified another eight types of cover which were to be included in

Section 3 of the revised draft of the AXOL policy.

On 28 April 1998 various versions of the AXOL slip passed between Smith, Houldsworth and Ellingsen. The revised draft of the slip that emerged is identical to that which was ultimately signed. Smith refined the description of the cover provided by Section 2 of the contract, that is, the cover for MIPI and ALAS, so that it allowed recovery by FAI where payments ' and the provision for outstanding claims set by F AI' exceeded the retention levels. It allowed F AI to choose between continuing to under reserve and only deal with the claims as paid or raise the reserves in a staged manner over time.

Between 27 and 30 April 1998 Smith and Burroughs had a number of meetings to discuss the terms of the ' other business' which was to be written in connection with the AXOL contract (and which became the six 1 May 1998 slips). 247 Self attended at least one of those meetings?48 Smith and Burroughs embarked upon those negotiations knowing that the amount of premium to make up the 'gap' (and GCRA' s fee) was $12.5 million. They discussed the types of cover and attachment points. They ' fine tuned' the premium payable under the contracts so that it equalled $12.5 million of premium and was payable in two instalments of $6.25 million on

1 May 1998 and 1 July 1998 respectively?49

On 28 April 1998 Burroughs sent an email to Wilkie and Mainprize250 updating them on the negotiations with GCRA. Wilkie responded to Burroughs and copied his response and the initial email to Peiris? 51

At some point in this period Smith prepared and signed the six 1 May 1998 contracts.252 On their face they provided for the different types of cover agreed with Burroughs and in total provided for payment of $12.5 million premium in two equal instalments of $6.25 million on 1 May 1998 and 1 July 1998. 253 It is not clear when these slips were provided to F AI but on 6 May 1998 Burroughs signed a letter addressed to Smith accepting the 'cover' provided for in the slips. 254 Given that they were an indispensable part of the overall transaction I consider it likely they were provided to FAI prior to the signing of the first AXOL contract on 6 May 1998.255

14.2.4 The creation of the 'side letter'

One of the more contentious documents examined during the inquiry was the so called 1 May 1998 side letter. 25 6 The letter was addressed to the managing director of GCRA and had provision for it to be signed by Wilkie as 'Chief Operating Officer'. The text of the letter was as follows:

146 Th e impact of the FA! acquisition

Despite the contractual intention of the reinsurance contracts detailed below [i.e. the six 1 May 1998 contracts], unless mutually agreed by both parties, F AI will not seek reinsurance recoveries [under those contracts] .

There were two versions of the side letter. One was signed by Wilkie and one signed by both Wilkie and Mainprize.257 Underneath the version with Mainprize's signature was the handwritten description 'Finance Director'.

Exactly how the letter came to be prepared and signed is not entirely clear from the evidence. I do not need to resolve all of the inconsistencies. Byatt, Self and Barnum were all aware around this time of the proposal for a side letter. They conceded they knew that a signed letter had been received from F AI prior to the execution of the first AXOL contract on 6 May 1998. I also find that Smith had the same knowledge. He was the principal contact in Australia for the alternative solutions division. He conceded that prior to the completion of the negotiations with Burroughs concerning the other business he knew that F AI had agreed that no claims were to be made under those polices25 8 and there was a proposal for a side letter. 259 He received an

email from his colleagues overseas on or about 1 May 1998260 referring to the side letter as the 'plan'. I think this is a reference to the planned method of achieving the no claims status of the six 1 May 1998 contracts.

A draft of the letter appears to have been prepared by either Smith or Self but Burroughs added the description of the six contracts and the words 'unless mutually agreed by either party' .261 Burroughs sent the text of the letter to Wilkie on 1 May 1998.262 At some point Wilkie signed the letter and it was sent to GCRA.263

Wilkie said that he was then contacted by someone from GCRA, whom he believed was Self, who requested that someone from the 'main board' sign the letter. 264 He took it to Mainprize and explained to him that ' ... GCR would like an extra signature on this letter' and Mainprize signed.Z 65 Mainprize told the Commission that Wilkie brought the letter to his office and stated:

They have asked for this letter to be signed. There is nothing in it. Don't worry about it. Just sign it. 266

Thus Wilkie and Burroughs were both aware of the side letter and its terms. I address Mainprize's position below.

14.2.5 6 May 1998

On 6 May 1998 Self and Smith attended at the offices ofF AI. They took with them two copies of the AXOL contract. Smith and Self say they met Wilkie and Mainprize and that either Wilkie or Mainprize requested that the contract be backdated to 16 March 1998 to which they agreed. They said that Wilkie and Mainprize then signed the contract followed by Smith and Self?67 Smith dated the

agreement. 268 The result was that the last page of the contract recited that it was 'Signed in Sydney this 161h day of March 1998' .Z 69

But the agreement was not signed on behalf ofF AI by Mainprize. It was signed by Wilkie and by the group secretary Robert Baulderstone. When this was brought to

The failure of HIH Insurance 147

Smith's attention he conceded27 0 that he may have been incorrect about Mainprize being present at all when the document was executed. Self conceded it was not Mainprize' s signature but still insisted that Mainprize was present. 271 He could not give any explanation for the presence of Baulderstone's signature. 272 Baulderstone recalled that the document was presented to him for signature by Wilkie on a date some time after 16 March 1998, in circumstances where Wilkie had already signed

it and it had already been signed on behalf ofGCRA and dated 16 March 1998.273

Wilkie said that he and Mainprize were present when Self and Smith attended with a contract for signature on 6 May 1998 and that Mainprize requested that it be backdated to the time when discussions started.274 He thought the date of 16 March 1998 was 'probably a guess on the part of Self and Smith'.275 Otherwise he said he did not discuss the backdating with Mainprize. 276 Wilkie could not recall whether Baulderstone was present when he, Smith or Self signed the contract or whether he took it to Baulderstone's office for signature.277 Wilkie had no explanation as to why Mainprize did not sign (assuming he was present).278 Mainprize denied being present when it was signed.279 It appears that on 6 May 1998 he left Australia on a business trip. 280

I accept that Mainprize was not present when the first AXOL contract was signed. Had he been present it is likely he would have signed the document. I accept that neither Smith nor Self initiated the backdating proposal. It must have been Wilkie who did so.

Self also told the Commission that after the meeting of 6 May 1998 he was telephoned by Wilkie who requested that he confirm that if the first AXOL contract ' did not perform to the regulator's requirements, could it be unwound or voided or whatever' ?81 Wilkie told the Commission that he said to Self 'if the main aggregate excess of loss contract was, for whatever reason, not to operate .. . then the cover under the other supporting business was to be suspended, all premiums to be retumed. 282 Self said that together with Smith283 he drafted a letter which Self signed and sent on that day stating:

Dear Daniel,

Following our discussions in your offices and by phone, we hereby agree that should the performance of the aggregate excess of loss reinsurance contract, made between our companies, be prohibited or rendered inoperable in consequence of any law or regulation which is in force in Australia then we will suspend the cover provided under those contracts set in your letter dated I May 1998 [i .e. the six I May 1998 sl ips] and return in full any premiums already paid there under less any claims paid or due for payment. 284 [emphasis added]

In its submission GCRA placed considerable significance on the terms of this letter. I have addressed it below.

148 The impact of the FAI acquisition

14.2.6 The second AXOL contract: negotiations

On 28 May 1998 GCRA invoiced F AI for the two instalments of premium due under each of the six 1 May 1998 contracts?85 The first instalment of $6 .25 million was paid on 29 May 1998286 and the second instalment was paid on 9 July 1998. 287

On 23 June 1998 Wilkie met Ellingsen and Self. Wilkie requested that GCRA provide further reinsurance cover for $40 million.288

Smith prepared a redraft of the existing slip. He emailed it to Byrne, Houldsworth and Ellingsen.289 He added two additional sections, one for general and products liability written prior to 1 June 1998 (Section 4) and the other for professional indemnity business written prior to 1 June 1998 (Section 5) with an additional $20 million of cover ($ 1 0 million per section). Under this draft the total cover in aggregate was $85 million with premium of $75 million. As with the first AXOL contract all of the premiums would be held on deposit until 1 January 2003 and

none of the claims would be paid before 1 January 2003 with the claim payments to be offset against premiums held on deposit.

The next version of the draft was sent by Byrne from Dublin under cover of an email received in Australia on 25 June 1998?90 This email and its response are of particular significance. Byrne stated as follows:

Andrew,

Have read the slip and discussed it with Tore and JBH. I am not sure that it is achieving the client's objectives because the premiums payable over the life of the contract are A$75m for A$85m limit. This means that the max Income Statement relief you can get is A$1 Om (for this year and over the life of the contract). I thought that the client was now looking for

A$30m this year.

For example 1998 premium is A$7m. If F AI put up a loss reserves of A$3 7m this year under this contract, it might look for Income Statement relief of A$30m. However JBH and I think that auditors would look at the contract and say that for the remainder of the contract, you are legally

bound to pay premiums of A$68m (75-7) for the next 4 years but the unextinguished limit is only A$48m (85-37). The result being that you have a known future loss on this contract of A$20m at 30 June 1998. Auditors would insist on either accruing for A$20m premium or booking an A$20m loss. Client would have told auditors that it needs to increase

reserves by A$37m but would only get A$IOm credit in the Income Statement. If JBH and I are wrong, Chris, please let us know. If we are right, here are some suggestions (not discussed with JBH), but clearly I am not as close to this and am not on top ofF AI objectives.

1. The gap between the aggregate premium and aggregate limit needs to increase to A$30m, which could be done as follows:

2. Ask them to pay the 1998 premium of A$7m on 117/98.

3. Looking at the A$ forward interest rate yield curve A$7m would yield 5.85 per cent pa from 117/98 to 1/1 /2003. We would therefore earn interest of A$2 .04m on the A$7m for the 4.5 years

The failure ofHIH In surance 149

that we hold it. So with this, we increase the limits in Sections 4 and 5 from A$1 Om to A$11 m per section.

4. We introduce a Section 6, a cat layer with a A$ 18m limit, which will not pay until all reinsurance has been recovered. Clearly, we do not expect to pay this layer but they need to persuade their auditors that they will pay it. Auditors might be prepared to accept this on the basis that they are under-reserved elsewhere.

5. Do we trust them (and our u/w skills) for the cat layer via a side letter or do we have contract with them which has emerging losses of A$ 18m??

6. If 4 does not then we need to provide a A$ 18m limit that will pay but we must place a contract with them to recover this A$ 18m (Newcap!). I do not think it would be good enough to rely on preferred premium terms going forward, we need to cover the liquidation risk.

7. We then change the overall aggregate limits from A$65m and A$20m to A$65m and A$40m respectively.

I think this should enable them to get A$30m Income Statement relief this year but clearly Chris will need to assess these suggestions both from the accounts and tax perspective of GCRA and Australian GAAP for F AI. As for the next A$20m, via a letter of intent, I'm afraid 1 have no brainwaves.

Byatt explained that Byrne's reference to 'Income Statement Relief was to 'profit'. 29 I

Byatt responded in an email sent on 25 June 1998292 sent to Self and Ellingsen. He enclosed a further draft of the contract. 293 Byatt understood the point being made by Byrne concerning the limiting effect on FAI's accounting treatment of the gap between maximum recoveries and premiums payable under the contract. 294 However, he proposed reducing the maximum recoveries available under the contract to $87 million and removing some of the premiums payable ($13 million) to create the gap referred to by Byrne.Z 95 He proposed making up the $13 million gap under various other deals including a 'claims free contract' _2 96

Self prepared a further draft. 297 Byrne sent a response to Selfs draft. 298 That was copied to Smith, Ellingsen, Houldsworth, Vukelic, Barnum and Byatt. Byrne was concerned that Selfs draft only had an apparent gap between maximum recoveries and premiums of $9.75 million which, based on the views he had previously expressed, could only lead to F AI obtaining 'income statement relief of $9.75 miilion. He attached two drafts, one described as 'draft 4' which had a gap of $9 .75 million premium between maximum recoveries and premiums and the other described as 'draft 5' which had a gap of $27 .75 mi11ion.Z 99 Byrne stated that the decision as to which of the two contracts would be written came 'down to FAI's accounting objectives'.

On 26 June 1998 Byatt sent an email resolving Byrne's dilemma in favour of the version with a gap of $10 million (in fact $9.75 million). 300 His email suggests this

150 Th e impact of the FA/ acquisition

followed a discussion with Mainprize in which Mainprize said he was 'happy with version 4'. Neither Byatt nor Mainprize could recall any such conversation.301

During the afternoon of 26 June 1998 Smith and Self attended the offices of F AI and met Wilkie and Mainprize.302 They took with them two copies of the second AXOL contract. 303 They also took a letter signed by both of them on behalf of GCRA addressed to Wilkie indicating GCRA's willingness to involve its 'overseas group' in seeking to obtain the additional $20 million cover that Wilkie had requested on 23 June 1998. 304 The contracts were executed in each others' presence with one copy and the letter handed over to Mainprize and Wilkie. Self and Smith received a letter dated 26 June 1998305 (the '26 June 1998 side letter') signed by Mainprize and Wilkie which stated as follows:

Despite the contractual intention of Section 6 in this Aggregate Excess of loss Reinsurance Contract, unless mutually agreed by both parties, F AI Insurance Group will not seek reinsurance recoveries under this section of this reinsurance contract.

Wilkie's evidence was that he was asked by Selfto provide this letter. 306

As a result of the various changes described above the second AXOL contract dated 26 June 1998 provided for six sections of cover. The first three sections were identical to the three sections in the first AXOL contract and like that contract they had an aggregate limit of $65 million. The second AXOL contract also contained an additional three sections of cover for nominated portfolios. Each of sections 4 and 5

had a limit of $11 million and section 6 had a limit of $18 million for each and every loss occurrence (in excess of $25 million). There was a total limit for sections 4, 5 and 6 of $22 million, making an aggregate limit of $87 million for the entire contract.

The total premium payable under the second AXOL contract was $75 million for sections 1 to 5. Unlike the original contract, $7 million of premium was payable by 30 June 1998. 307 The remaining $68 million was to be held 'on deposit' by FAI until 1 January 2003. This reflected the cashless premium method adopted in relation to the first contract. The 'premium' for Section 6 was $2.25 million payable in five

instalments of$450 000 each over four and half years commencing 30 June 1998. 308 As with the first AXOL contract, this contract provided that GCRA was to hold any recoveries payable to F AI on deposit and was not obliged to make cash payments until 1 January 2003. 309

14.2.7 The substance of the GCRA transaction and risk transfer

By 6 May 1998 the first AXOL contract was in existence. It had the following features.

• It provided for maxtmum recoveries of $65 million against premmms of $55 million.

The failure ofHIH Insurance 151

• It was cashless in the sense that FAI would hold its premium payments 'on deposit' for GCRA until 1 July 2003 and GCRA would not be obliged to make payments for recoveries until that date.

• It contained an offset clause allowing recoveries to be set off against premiums due under the contract and due under 'any other agreements' .

By that time there were three other relevant aspects of the arrangements between F AI and GCRA. First, the six 1 May 1998 contracts under which F AI was required to pay $12.5 million premium in cash in two equal instalments on 1 May 1998 and 1 July 1998. Second, the letter dated 6 May 1998 from GCRA confirming that if there was any regulatory matter which affected the performance of the first AXOL contract, then the premiums payable under the 1 May 1998 contracts would be repaid less claims paid or due for payment. Third, the 1 May 1998 side letter, the tenor of which was that unless mutually agreed, F AI would not make claims under the six 1 May 1998 contracts.

The combined effect of these agreements at the time of their inception was that, in effect, F AI paid $12.5 million in cash to fund the gap between the maximum recoveries of $65 million and premiums of $55 million payable under the first AXOL contract and GCRA's fee. As the first AXOL contract was 'cashless' in the sense discussed above and contained the offset clause, it followed that, subject to any possibility of GCRA having to pay claims under the 1 May 1998 contracts, there was otherwise no possibility of GCRA ever incurring any loss under the transaction. This is because the total premiums receivable exceeded the recoveries payable. Any credit risk was overcome by the upfront payment of $12.5 million

cash under the six 1 May 1998 contracts, combined with the offset clause.

By 26 June 1998 the first AXOL contract had been replaced by the second AXOL contract. The substance of the revised transaction was one in which GCRA agreed to pay claims on 1 January 2003 totalling $87 million. However, $68 million of that amount was to be funded by F AI through the premiums held on deposit and applied as an offset. For the remaining $19 million, FAI was obliged to pay $19.95 million by 1 July 1998 and a further $450 000 per year for four years-a total of

$21.7 5 million.

It was common ground that a contract could not be regarded as reinsurance capable of accounting treatment as such unless it contained 'transfer of risk'. But the degree of risk transfer and the means of assessing it were contentious matters. I will return to consider that issue later.

It was submitted on behalf of GCR and GCRA that, in this respect, there was a difference between the concept of risk transfer and the probability or possibility that the reinsurer may Jose under the transattion. Although several examples were cited to support the distinction I found the point difficult to grasp. To my mind there is no relevant difference in this context provided the assessment of the possible loss to the

reinsurer is undertaken from the commencement of the agreement and not with the benefit of hindsight.

152 Th e impact of the FAI acquisition

It was also submitted on behalf of GCRA that each agreement operated according to its tenor and had to be considered according to its terms. They could not be considered 'a sham'. According to this argument the first or second AXOL contract, containing on its face an apparent transfer of risk or possibility of loss to GCRA, was always a genuine contract of reinsurance irrespective of the other six contracts and the side letter. In other words, even if the six I May contracts were found to be deficient, the integrity ofthe AXOL contracts would remain intact.

Again, I take a different view. There is ample evidence to support the view that the various agreements and letters formed part of the one larger transaction. It was not seriously contended that GCRA would have offered either AXOL contract without receiving $12.5 million in cash under the six 1 May 1998 contracts and without the protection of the side letter. It is not unusual to have separate documents evidencing

an overall transaction. The fact that there are separate documents does not of itself lead to a conclusion that any of them is a 'sham' in the legal sense. But it may, and in this case does, raise the question why the transaction was documented in the way it was. This is a matter to which I will return.

The significance of the various agreements forming part of one larger transaction relates to the assessment of risk transfer. An experienced auditor engaged by the Commission, Greg Couttas, pointed out the accounting standards required an assessment of the substance or economic effect of a transaction (not just a contract) and concluded as follows in relation to the assessment of risk transfer:

Determining whether a reinsurance contract indemnifies a ceding enterprise against loss or liability relating to insurance risk requires a complete understanding of all contracts or agreements with related insurers. Although an individual contract may appear to indemnify the

ceding enterprise, the risk assumed by the reinsurer through one reinsurance contract had been offset by other contracts or agreements. A contract does not meet the conditions for reinsurance accounting if features of the reinsurance contract or other contracts or agreements

directly or indirectly compensate the reinsurer or the related reinsurers for loss. That compensation may take many forms, and an understanding of the substance of the contracts or agreements was required to determine whether the ceding enterprise has been indemnified against loss or liability

relating to insurance risk. 310

I accept the correctness of this approach. On my reading of Barnum's evidence he d b f . "l . 311 appeare to eo a s1m1 ar view.

GCRA submitted that a logical extension of this approach would require the entire contractual history as between F AI and GCRA to be examined to see whether, on an analysis of all contracts entered into over time, the reinsurer was at risk of loss. I do not think this is correct. Contracts that are entirely independent and separate fall to be assessed on that basis. The same does not apply to contracts that are linked. I can best explain what I mean by reference to a related submission advanced on behalf of GCRA.

The failure ofHJH Insurance 153

This involves the concept of 'good business' and 'bad business'. It is not an uncommon practice for reinsurers to offer risky reinsurance to a client in the expectation of receiving more profitable reinsurance in the future . It was contended that each such contract entered into has the question of risk transfer assessed by reference to its terms only. To my mind whether that is correct depends upon the precise characterisation of the expectation or right to 'good business'. If it is no more than a hope or expectation of a continuing commercial relationship I can accept that the 'good business' may not affect the assessment of risk transfer of the

' bad business' . But that is not what happened here. The related business was entered into at the same time as and because of the first AXOL contract. The contracts were expressly linked by the offset clauses and the 6 May 1998 letter. They were part of one transaction.

Against that background I have to consider the effect of the 1 May 1998 side letter and the letter of 6 May 1998.

It would be artificial to attempt an explanation of the legal effect of the 1 May 1998 side letter in relation to the transaction as a whole. I doubt that either party would have approached a court to seek a ruling in relation to its legal effect. I consider the letter means what it says. No claims would be made under the six 1 May 1998 contracts 'unless mutually agreed by both parties'. These words were superfluous because the parties could always have agreed to alter their arrangements. The significant point is that without those words F AI bound itself (commercially if not legally) not to make claims. I recognise that, in accordance with the plain meaning of the words used, GCRA could have agreed to the making of claims. But the question which arises is: under what circumstances?

Smith told the Commission that his understanding was that if a claim arose under the six slips F AI 'would then have to come back to the table and renegotiate some other business so that, over time, that deficit was made up'. 312 In another part of his evidence he put it in a slightly different way: 'we would then have to look at another way of making up that shortfall over time with them, based on an ongoing relationship with F AI' . 313 I have no difficulty in accepting this explanation. But the effect of the side Jetter, so understood, is to convert an obligation to pay a claim arising under the six 1 May 1998 contracts into a discretion to do so. It was a discretion that GCRA would only exercise (according to Smith) if there were further deals negotiated at the time which 'made them good'. All of that was in the :future. Unless and until that occurred the 1 May 1998 side letter removed any risk of loss to GCRA. It eliminated any residual risk transfer that remained from the combined effect of the first or second AXOL contracts and the six 1 May 1998 contracts.

The 26 June 1998 side letter removed the potential for there to be claims under section 6 of the second AXOL contract. But it had a much more limited effect on the transaction between F AI and GCRA than the 1 May 1998 side letter. It did not significantly affect risk transfer under the revised contract.3 14 On the other hand it operated to disguise the fact that the 'premium' payable for section 6

154 The impacr of the FAI acquisition

($450 000 per annum) was little more than a fee for renegotiating the original deal. 315

GCRA placed great store on the last proviso to the letter of 6 May 1998 ('less any claims paid or due for payment'), and submitted that it was 'telling evidence that the parties expected the six slips to operate according to their tenor, the side letter of 1 May 1998 notwithstanding'. At first glance the argument has attraction. But on closer analysis I think the proviso is consistent with the full operation (according to

its tenor) of the 1 May 1998 side letter.

Self said he and Smith drafted the letter but Self did not refer to it as evidence that he contemplated that claims would be made. Smith and Byatt said they had no recollection of having seen the letter. Barnum said he did not see it until some time after the event. 316 In other words, no one from GCRA indicated that the purpose or one of the purposes, of the letter was to permit claims to be made or that it envisaged this to be the case. Nor did Wilkie (who requested that the letter be provided) say that this was its purpose.317

In any event it is compatible with the view that I have taken of the 1 May 1998 letter. If claims had been made under the six 1 May 1998 contracts and accepted by GCRA in the exercise of discretion, then GCRA would have done so in respect of a risk for which FAI had paid the premium of $12.5 million. If (due to the triggering of the 6 May 1998 letter) the whole transaction was to be unwound and GCRA was required to return the premium, then it would follow as a matter of commercial logic that F AI should not retain the benefits it had received or was to receive by way of claims paid in the exercise of discretion.

14.2 .8 Relevant knowledge of individuals

I turn now to the understanding of each of the individuals in relation to the GCRA transaction and why it was documented in the way it was. Barnum, Byatt, Self, Wilkie, Burroughs, Smith and Mainprize were all aware on or by 6 May 1998 of the AXOL contract, the six 1 May 1998 contracts, the 1 May 1998 side letters and the connection between the arrangements.318 They knew that the $12.5 million premium

made up the 'gap' between premiums and maximum recoveries.

Both Self 19 and Barnum told the Commission that they were aware of the absence of risk transfer under the transaction. This question and answer appears in Barnum's evidence:

Q. It was your assumption always in 1998, wasn ' t it, that the

arrangements between GCRA and F AI involved the absence of risk to GCR?

A. Essentially, yes. 320

In light of Selfs evidence and the email of 1 May 1998, I infer that each of Ellingsen and Houldsworth were of the same view. Indeed it was their intention that the other business would be claims free. I have already referred to Smith's

Th e failure ofHJH Insurance 155

understanding above. In my view the effect of his understanding was that if F AI insisted on payment of a claim under the six 1 May 1998 contracts, GCRA would insist on a further transaction to recoup that payment. Smith was not prepared to concede that it followed from that view that there was no transfer of risk or that it even affected risk transfer. But I think it can be inferred from what he said that he was aware that there was no risk transfer.

Byatt told the Commission he suspected that the effect of the letter was to remove any risk transfer under the six 1 May 1998 contracts and the overall transaction between FAI and GCRA. 321 But he proffered this explanation:

From an accounting perspective, the purpose of the side letter was to enable there to be certainty that claims could not be made and therefore $2.5 million of the premiums payable under the six slips could be booked as a fee. 322

He did not explain the difference between an 'accounting perspective' and an 'undetwriting perspective'. To my mind, in this context there is no difference. The accounting standards require that accounting treatments reflect the substance and economic effect of a transaction. The side letter could only assist in obtaining Byatt' s desired accounting treatment if it meant that the six 1 May 1998 contracts would be claims free. That is also the undetwriting result. If it were othetwise then GCRA's accounts would be false. Not surprisingly GCRA did not press such a submission on me.

Burroughs conceded the 1 May 1998 side letter had an effect on risk transfer323 but pointed to his inclusion of the words 'unless mutually agreed'. He said these words were inserted to add some uncertainty to the request that had been made by GCRA. The request had been conveyed to him by Wilkie in these terms: 'they want it that no claims will be made under the six treaties' . 324 Be that as it may, I consider that Burroughs knew the letter affected risk transfer and strongly suspected that it removed it altogether. Like Smith, he put it no higher than that if claims arose there would be a chance to sit down with GCR and negotiate further prospective reinsurance to balance the situation. 325

Wilkie agreed that if F AI could not seek recoveries under the six 1 May 1998 contracts without GCRA's prior agreement- that is, if the letter meant what it said- then GCRA was not exposed to any risk under those contracts or the transactions as a whole. 326 But Wilkie said that he believed the letter did not operate according to its terms. He said that when he requested the letter Self indicated that the prime reason for it was to allow GCRA to recognise income in either 1997 or

1998 more rapidly than would othetwise be the case. Self had added that GCRA was also concerned that there was 'potential for claims to be paid more quickly or more rapidly than would othetwise have been the case' .327 I have difficulty with this explanation. There was no other evidence of anyone from GCRA ever communicating such a reason to Wilkie. The letter says nothing about timing of claims.

156 Th e impact of the FAJ acquisition

Wilkie also maintained that he did not regard the side letter as effective. This was because he thought there was a 'commercial understanding' between F AI and GCRA to the effect that if F AI wanted to make claims for the 'sake of a small amount of money' GCRA would have to pay them. 328 If there was such an understanding it does not appear to have been articulated or discussed at the time by any of the participants, including Wilkie. It would lead to a situation where there was a side understanding contradicting a side letter which in tum contradicted the

six contracts which were in tum part of a larger transaction involving the excess of loss contract. The commercial logic of such a contractual arrangement is difficult to comprehend.

Wilkie also suggested that he believed the side letter was not 'effective as a legal document' and on one occasion referred to it as a 'letter of comfort' .329 I do not agree and I do not think he held that view at the time either. Wilkie was not a qualified lawyer and never sought legal advice on the operation of the letter. The side letter is an extraordinary document and I do not believe that a senior executive

like Wilkie would form a view of its legal effect contrary to its plain meaning without seeking advice. I think Wilkie was aware when he signed the I May 1998 side letter that it removed risk transfer under the transaction between F AI and GCRA.

Mainprize told the Commission that he did not read the 1 May 1998 side letter when Wilkie brought it to him for his signature. 330 This is a difficult proposition to accept. The letter is short and the language is not difficult to follow. It begins with a rather stark phrase: 'Despite the contractual intention ... '. It is, I think, unlikely that it would have escaped the eye of someone asked to sign it. Mainprize was aware that the letter was being sought by GCRA331 and that it was of some importance to

them. 332 He was also aware that there was 'other business' associated with the arrangements under which $10 million premium and A$2.5 million fee was payable.333 Mainprize had time to sign the letter and add the words 'Finance Director' under his signature. 334 I think it is likely that he read the letter.

In his submissions Mainprize says that there was no reason for him to draw any connection between the side letter and the 'other business' which was related to the AXOL contract. Otherwise he notes that there was no evidence that anyone (including Wilkie) had discussed its terms with him. He points to his evidence that Wilkie told him it was not of significance. I have considered carefully Mainprize's evidence and the submissions of his legal representatives. But I cannot escape the conclusion that he knew what it was he was signing. Mainprize was a director and

the chief financial officer of the F AI group. Although Wilkie had been primarily involved in negotiating the transaction, Mainprize had taken part at critical times. He spoke to Self on 20 March 1998; he attended a briefing on 26 March 1998 with Self and Smith; he attended a meeting with Ellingsen, Self and Wilkie on 24 April1998 and he was copied in on Burroughs' email of 28 April 1998. I do not think he was kept 'out of the loop' on the deal. On the contrary, these matters would

have brought home to him its significance. Against that background he was presented with a letter that is both brief and extraordinary in its terms. He signed it.

The failure ofHIH Insurance 157

I think he read the letter and understood that it was putting into effect what Self says that Ellingsen conveyed during their meeting on 24 April 1998, namely that the 'other business' would be claims free.

The positions of each of Wilkie, Mainprize, Barnum, Ellingsen, Byatt, Smith and Self are the same in relation to the second AXOL contract. Each of them was aware of the nature of the revised deal. The May transaction did not transfer risk to GCRA. The situation was the same under the arrangements as revised in June.

Inevitably the transaction involved FAI paying out more in premiums than it could get in recoveries. Without risk transfer it could not properly confer on F AI the accounting benefits that it was seeking to obtain. No one gave evidence of any other commercial benefit or advantage that the transaction would confer on F AI. Pearson estimated that as at the date of entry into the contract it had a negative financial value (to FAI) of $6.9 million. 335 In this sense the transaction was both without risk to GCRA and financially detrimental to F AI. On this basis it is open to draw the inference that each of the persons concerned would have been aware that this was the case.

Although there is no evidence that Burroughs was involved in the June negotiations, the findings that I have made in relation to the May arrangements carry over. Otherwise, I do not consider his position to be relevantly different to the others.

For the sake of completeness I should mention the position of Peiris and Adler. Peiris told the Commission that he was aware that in April 1998 there were discussions or negotiations between F AI and GCRA over a reinsurance contract. But he denied that he was ever aware that there was a series of other reinsurance contracts that were ultimately connected to or supporting the reinsurance contract that was eventually negotiated. Peiris received a copy of an email from Burroughs to Wilkie dated 28 April1998 which referred to the other business and suggested a connection with the AXOL slip .

336 But without being familiar with the background

I am not sure that a reader of that email would understand the close connection between either the AXOL contract and 'the other business'. I make no finding against Peiris.

Adler denied any knowledge of the six 1 May 1998 contracts, any connection between them and any AXOL contract with GCRA, the back dating of the first GCRA contract to 16 March 1998 or the existence or content of any of the side letters. I consider the evidence does not sustain a finding that Adler was aware of those matters. Wilkie told Adler about the AXOL and that GCRA wanted

'additional supporting business' for a further $12.5 million, but there is no evidence he was told anything further.

14.2.9 Purpose in documenting the transaction

The next question is why the transaction was structured in the way it was. Each of Burroughs, Self, Smith, Barnum, Wilkie and Mainprize understood that F AI could only book recoveries in excess of premiums under the AXOL contract if it was

158 The impact of the FAI acquisition

accounted for as reinsurance. This required there to be risk transfer. 337 Each of them denied having any improper motive or purpose in documenting the transaction338 in the way it was documented or in executing such documentation. Counsel assisting submitted that each of them and Ellingsen and Houldsworth understood that the transaction was documented to enable F AI to present an AXOL contract to its auditors (and to APRA if required), as one showing the appearance of a transfer of

risk. This arose by virtue of the apparent gap between the amount of premiums and the maximum recoveries payable, with the other elements of the transaction which removed the transfer of risk hidden. By their actions each of them was submitted to have facilitated the documentation of a transaction with that object in mind.

I cannot overlook the fact that each of the persons I have named had knowledge or awareness of these things:

• all of the constituent elements of the transaction

• the accounting significance of the need for risk transfer

• the object of the exercise of finding ' other business' (which was then sought to be made 'claims free') was to make up the gap between recoveries and premiums apparent on the face of the AXOL slip.

This, it seems to me , provides a basis from which an inference of the type suggested by counsel assisting could be made. In arriving at that view I have paid careful attention to what the witnesses said in evidence and the submissions put on their behalf. But one thing resonates with me . No one advanced a plausible explanation as to why the cover provided for in the six 1 May contracts was not incorporated

into either AXOL contract. 339

It was submitted to me on behalf of GCR and GCRA that I should not infer that any of their officers had an intention to conceal, or to assist F AI in concealing, information from auditors. Counsel assisting pointed to the evidence that steps were taken to ensure the side letter was disclosed to certain directors ofF AI or at 'board

level'. Byatt said he wanted the letter signed by 'at least a board member' and he specifically requested that Adler sign?40 But Selfs email of 20 March 1998 341

referred to under-reserving at F AI and to the fact that the only people at F AI who were to know about it were Adler, Wilkie, Mainprize, Burroughs and Peiris. This class involved two directors who already knew of the transaction. Similarly, Vukelic's email to Self referred to having the side letter signed 'by eg Rodney and Tim'. 342 In any event it says nothing about disclosure to the auditors. I think it likely

that the disclosure contemplated by Byatt was to the limited class of people referred to in Self s email.

GCRA further submitted that the structure of the documentation was, in effect, 'simple history' and there was no attempt to hide the connection between the two sets of contracts. I presume ' simply history' means the way things emerged as a matter of chronology. But the structure of the agreement involving two sets of business was fixed almost from the outset. 343

The failure of HIH Insurance 159

It was only after the due diligence was completed that GCR and GCRA insisted the 'other business' be claims free. The l May 1998 side letter became the means of ensuring that occurred. The 'other business' was conceived so that the proper extent of risk transfer was not apparent on the face of the first or second AXOL slip. The

1 May 1998 side letter was conceived so that the removal of risk transfer was not apparent on the face of either the AXOL slip or the six 1 May 1998 contracts. I consider that the documentation was structured for that purpose and that each of Wilkie, Mainprize, Burroughs (in the case of the first AXOL contract), Ellingsen, Houldsworth, Barnum, Byatt, Self and Smith understood that and by their actions sought to facilitate it.

This conclusion as it affects Wilkie, Smith and Self is strengthened when a search is made for an explanation why the first AXOL contract was backdated to 16 March 1998. Although Wilkie said he did not initiate the proposal he thought it was ' satisfactory' as an 'agreement in principle' had been reached in March 1998. But I could not find in the evidence a basis for the assertion that an agreement in principle had been reached in March 1998. No other witness advanced a plausible reason. I am left to infer that the reason was to introduce a temporal gap so that a person reviewing the AXOL contract would be less likely to draw a connection between that contract and the six I May 1998 contracts.

Self asserted that he and Smith agreed to the backdating of the AXOL contract 'because the XOL inception date was I January 1998' and they had been delayed by another reinsurer. 344 Smith stated that he and Self agreed to the backdating because 'the date on which we had originally met was approximately 16 March'. 345 But to preserve the integrity of the inception date it would not matter whether the date of

signing was 16 March 1998 or 6 May 1998. As I have said, I have difficulty understanding the significance of the date of the original meetings in the context of a signing date.

By the time of the second AXOL contract there was no longer a need for it to be backdated. It bore the date 26 June 1998 and the six contracts bore the date 'I May 1998 '. There was no obvious connection between them at all.

14.2.10 GCRA's understanding of FAI's accounting objectives

Counsel assisting also submitted that the exchange of emails between Byrne, Byatt and others in June 1998 (which I have already described) was significant. It was submitted that to reveal that GCR, GCRA and their officers structured and documented a transaction designed to enable F AI to reveal that GCR, GCRA and their officers involved themselves in structuring and documenting a transaction designed to enable F AI to book a profit and not disclose the absence of risk transfer. In their submissions GCRA and the relevant officers denied that there was any such involvement or that the emails contain any such suggestion.

The emails refer to the amount of 'Income Statement relief (profit) that FAI was expecting to book for year ending 30 June 1998 under the GCRA contract, namely

160 Th e impact of the FA! acquisition

$30 million. In fact F AI booked a profit of $28.824 million. 346 The emails also speculate about FAI's management dealing with its auditors. The question is whether they contemplated the auditors being misled? Byatt was questioned extensively on these matters. 347 When it was put to him that if the auditors were shown the AXOL, the six 1 May contracts and the 6 May side letter they would not have allowed FAI to account for the profit he answered: 'it sounds correct'. 348 By this he agreed the auditors would have concluded that there was no risk transfer. He was further questioned as follows:

Q. You also knew that if they were acting properly, F AI should disclose the side contracts and the letter to their auditors?

A. Yes.

Q. Again, because that's what you did with your auditors?

A. Yes.

Q. I don't want to labour the point, but if you could just, if possible, either agree or disagree with the proposition: that is , you suspected the only way that F AI could achieve what they wanted to achieve, that is booking a significant profit on those contracts

for the year ended 30 June 1998, was by not showing to their auditors the six contracts and side letter?

A. I'm not aware of them being able to do it any other way.

Q. You were prepared to assist F AI to achieve its objectives of booking a profit for the year ended 30 June 1998 by suggesting a method for structuring a contract which had on its face a significant gap between premiums and maximum recoveries?

A. Yes. 349

The debate in the emails passing between officers of GCR and GCRA in June 1998 concerned the means by which the apparent gap between premiums and recoveries on the face of the AXOL slip could be increased and then recouped by other means. If the 'other means' were disclosed to the auditors and were found to negate risk transfer, the perceived advantage flowing from the 'gap' would disappear. I think Byatt must have appreciated F AI would be likely to keep the absence of risk transfer

under the transaction from its auditors in order to book a profit under the contract for the year ending 30 June 1998.

I consider that a similar understanding was held by at least Ellingsen. After the announcement of the takeover of F AI by HIH in September 1998 Ellingsen sent Vukelic an email350 in which he contemplated HIH wanting to reverse the transaction between F AI and GCR, no doubt because the economics of it were poor

from FAI's perspective. He concluded that HIH would not do that because the

The failure of HIH Insurance 161

impact upon FAI's accounts would be 'too volatile'. I take this as indicating that he understood that F AI had booked significant profits for year ending 30 June 1998 based on that contract which would have to be reversed if HIH 'cancelled' the contract.

Barnum received copies of the exchange of emails in June 1998 between Byrne, Byatt and Self. He said that he expected F AI would show its auditors the AXOL contract for the purpose of having them accept it as reinsurance351 and that in doing so it was likely that they would book claims recoveries in excess of premiums. 35 2 Yet he conceded as follows:

Q. Well given that you understood that essentially the arrangements were ones which involved no risk to GCRA, must you not have understood that F AI could not properly account for the transaction as one of reinsurance?

A. Yes. 35 3

Q. Do you agree that in August 1998 you understood that the result of this transaction your company had entered into put F AI in a position where it could pass off as reinsurance something which, in substance, was not reinsurance?

A. Yes.

Q. And you were prepared to allow the transaction to proceed notwithstanding that fact; do you agree?

A. On the face of it, yes .

354

I think it follows that Barnum was aware that F AI could only achieve its accounting objective of booking a profit under the transaction if the auditors were unaware of the elements of the transaction that removed risk transfer. I think it likely that he appreciated they would take that course.

Although he was a recipient of Byrne and Byatt's emails, Smith denied that he understood FAI's objective was to use the contract to book a profit of approximately $30 million for the year ending 30 June 1998. He stated that he understood FAI's objectives were to obtain the 'accounting treatment [of a re-insurance contract] to build up the reserves over time'.355 I do not think these objectives are materially different. Smith's idea of using the contract to 'build up reserves' was merely a description of recognising a liability for the relevant amount in FAI's balance sheet356 and using the contract to raise a corresponding recovery, which would exceed the premium expensed. If the auditors were to accept the accounting treatment, a profit would be recorded under the contract in that year. On this basis it would be open to conclude that Smith knew of FAI's intended use of the AXOL contract to book a profit and that it was likely they would not disclose to the auditors the existence of either the six 1 May 1998 contracts and the side letter.

162 Th e impact oft he FAJ acquisition

Self conceded that he knew the transaction as renegotiated in June 1998 did not, as with the original transaction, involve any risk of loss to GCRA. It does not seem that Self had the same depth of knowledge of FAI's proposed accounting treatment as that of the other GCR and GCRA personnel referred to above. But on the

knowledge that he did have I think it is open to conclude that he appreciated that F AI was unlikely to disclose to its auditors all matters necessary to assess risk transfer.

14.3 The negotiation of the Nl contract

Since about 1996 F AI had been in discussion with the reinsurance brokers Guy Carpenter about whole of account reinsurance protection. As at 1 June 1998 no active negotiations with any reinsurer were on foot.

Sometime around 2 June 1998, Geoff Bromley, a senior executive with Guy Carpenter received a telephone call in his London office from Adler. 35 7 Bromley reported the substance of the conversation in an email to John Tuckfield, the joint managing director of Guy Carpenter's Australian subsidiary. The email stated:

Spoke to Rodney at 5.45am your time this morning ... He needs a stop loss cover and wants it fast!

He needs A$20million of cover for each of 1997/8 and 199819 to try and ensure his earnings meet expectations.

He notes that it is the 1990 and prior years that have caused the slight problem and losses from those years have certainly affected the results for 1997/8 and are expected to also potentially cause a problem in 1998/9.

Rodney would like to finance these recoveries over the next 6 or 7 years.

We need to get this in place by the 301h June and ideally we need to do this with an Australian Based entity to avoid tax .. . [emphasis added] 358

Adler said that he did not mention seeking reinsurance to obtain any 'betterment in results' but to provide additional protection for F AI for several years so as to insulate the company against any unexpected losses that may arise?59 Bromley said he understood Adler to be seeking a product designed to provide the flexibility of enabling F AI to book a loss recovery in the event that underwriting losses for the

1997- 98 underwriting years were unexpectedly and unacceptably high. He understood Adler wanted to book a large recovery in both the 1997-98 and 1998-99 years360 and that he intended to finance these recoveries over the next six or seven years.

14.3.1 Further improving the results

I think it is safe to accept Bromley's evidence as to what Adler told him. It is consistent with his contemporaneous email. Both Bromley and Tuckfield referred to this as F AI' s purpose in various communications with reinsurers. The objective of

The failure of HIH Insurance 163

seeking to book an immediate profit is relevant to an assessment of the legitimacy of the steps that were taken. It is also important to consider how the objective was achieved.

From the time of Adler's request until 23 June 1998 Bromley and Tuckfield sought to broker a reinsurance contract between F AI and Swiss Re. It seems that the deal being considered involved 'parallel contracts ' . Ultimately no such cover was taken out. It is not necessary for me to consider the negotiations with Swiss Re other than to note that during the course of those negotiations Wilkie disclosed to Tuckfield

that F AI was substantially under-reserved in its international professional indemnity portfolios. I have addressed this in Section 14.1 above.

By the evening of 22 June 1998 the negotiations with Swiss Re had come to an end. It seems that the 'deal breakers' were Swiss Re's need to have the deal signed off by the Insurance and Superannuation Commission and a concern on the part of Swiss Re of further potential shortfalls in FAI's reserves which would lead to a need for

'further finite or structured relief for 98/99'. In an email to Tuckfield reporting on this, Bromley told him he had contacted Ajit Jain of Berkshire Hathaway (NI) and that Jain agreed to look at a 'quick fix' and was prepared to consider it promptly. Bromley requested Tuckfield provide him with a placement slip that would be acceptable to F AI together with an outline of what they would be prepared to accept by way of parallel contract, the cash flows they would agree to and the margins they would be prepared to pay. Tuckfield replied by facsimile on 23 June 1998.361 He enclosed a draft contract he called 'placement slip'.362 The slip was on Guy Carpenter letterhead and bore the reference code 'FAI/389/98'.

14.3.2 Negotiations with National Indemnity

On 23 June 1998 Bromley sent a fax to Jain. 363 He requested assistance from Berkshire Hathaway and provided a general description of F AI and Wilkie's position. Along with the facsimile Bromley sent a number of documents including the draft placement slip364 sent to him by Tuckfield. The fax stated inter alia:

164

4) .. . Daniel is of the view that certain elements of the Commercial Division have been under-reserved and indeed have been operating unprofitably for some time.

5) FAI have a I st July and 30th June financial year.

6)

They require an effective improvement in their 1997/98 results of A$40 000 000 which is little under US$25 000 000 at current rates of exchange. This is the net betterment being sought (i.e. net of the premium payable).

F AI would then like to repayment of this amount over

the next 4 years or so.

The Proposal calls for the Reinsurer to write an Aggregate Excess Cover for F AI on the 1997/98 underwriting year.

Th e impact of the FAI acquisition

The attachment point for the cover would be such that it will be quite some years before paid losses would reach the attachment figure. Indeed, F AI would be prepared to warrant by side letter that they will not make any cash recovery from that contract until an agreed final commutation date .

7) F AI would then write a parallel cover or covers for the Reinsurer on a portfolio of Reinsurer's choice, whereby F AI would incur losses over the next 3 or 4 years. Again it is intended that F AI assume incurred losses on these 'parallel' covers and these

incurred losses assumed by F AI, would offset the liabilities under the other contract.

8) To ensure there is no credit risk, F AI would set up the necessary Trust Fund in the USA to secure any losses by F AI under the parallel contract (s).

9) There would be a full commutation at the end of 2002 which would result in the Reinsurer being left with an appropriate margin for the use of his Balance Sheet.

As you can see, the contemplated structure is very simple. F AI get a A$40 000 000 benefit for their 1997/98 P&L account. They pay a premium for this.

FAI then write appropriate parallel contract(s) for the Reinsurer and secure their obligations under those contracts by setting up an appropriate Trust Account in the USA.

A full and final commutation will take place at say the end of 2002.

You already have certain financial information relating to the F AI.

Attached is a Draft Slip of the contract needed by FAI to effect their required betterment by the 30 June 1998. I would note we could have an appropriate review clause thereby giving us a little more time to choose the appropriate parallel contracts that FAI would write for National

Indemnity.

Additionally I have enclosed additional portfolio triangulations to demonstrate that there is no chance of claims becoming payable under this aggregate cover prior to the commutative date . As stated, in any event F AI would be happy to also encapsulate this in a side letter.

The enclosed placement slip reflected these requirements. A parallel slip was not enclosed.

On 24 June 1998 Jain returned the draft placement slip to Bromley with a number of handwritten amendments.365 He rejected the use of parallel contracts. Instead he proposed a policy providing cover of $50 million per underwriting year for five years subject to an overall limit of $100 million. There was an additional clause providing that total losses in excess of $50 million would only be recoverable in the

event of an earthquake event causing a total market loss greater than $20 billion. Premiums were $11 million per year in advance of each underwriting year totalling $55 million. Upon commutation premiums plus earnings but less a fee of $2 million

Th e f ailure of HI H Insurance 165

would be returned to the reinsured. This proposal was similar to that suggested by Willis in December 1997- namely a policy with two sections of cover, one fully funded by premiums payable over a number of years with the second layer only triggered upon the occurrence of a very remote event.

On 24 June 1998 Bromley advised Tuckfield that he had a response from Nl 'although on a different basis from that we were after'. 366 He described the cover that was being offered and noted that the cash flows and overall net charge under the contract would be problematic for F AI. In relation to the parallel cover Bromley said Jain thought it too transparent and problematic if discovered by the regulator.

On 25 June 1998 Bromley reported to Tuckfield that he had requested some concessions from Berkshire Hathaway.367 On 26 June 1998 Tuckfield sent an email to Bromley requesting a number of changes on behalf ofF Ae68 He requested that cover commence for the 1997-98 year and be ' appropriately backdated' and that there be a reduction in the first year's premium with the difference to be made up by increases in subsequent years. Both Bromley and Tuckfield said that the request for a reduction in the premium in the first year was to maximise the potential benefit, that is, i.e. the gap between recoveries and premiums, in the first year of the contract. 369

Tuckfield's evidence was that the instructions to backdate the cover were gi ven to him in a meeting at FAI attended by himself, Wilkie and possibly Mainprize.370 He said they had discussed the facsimile which Bromley had sent to Jain on 23 June 1998371 and Wilkie had requested that the contract be backdated.372 He recorded these instructions on a copy of the slip sent with Bromley's fax .373 Tuckfield said that he did not ask why Wilkie wanted the cover backdated because he saw no problem with the clients' request as it did not impact on the cover already agreed to and recorded in the placement slip. 374 Tuckfield said that he thought one reason for the request was to ensure the cover was for the whole 1997- 98 year which by that time had partially expired. 375 Wilkie denied making any such request stating that his only instruction was for 'cover' to be backdated to 1 July 1997

376 as

opposed to the backdating of the policy or cover note. Mainprize denied giving an instruction for the policy to be backdated.377

As a result of further discussions between Jain and Bromley, two alternative draft slips were produced by Tuckfield and he sent them to Wilkie and Bromley. 378 In preparing them he took a number of steps with the slip which I regard as significant. Consistent with the suggestion that they had to be 'appropriately backdated' both slips were dated 4 July 1997. Previously the draft slips had not been dated. Thereafter all drafts of the slips bore that date. The slips now appeared on paper headed 'Carpenter Bowring .3 79 and referred to a proposed 'Intermediary clause' which named 'Carpenter Bowring Australia Pty Ltd' as the intermediary.

380 All the

previous drafts were either on blank paper381 or on paper bearing the letterhead 'Guy Carpenter' 382 and referred to the intermediary as 'Guy Carpenter Ltd' . 383 The exception was for the first one sent by Tuckfield to Adler on 7 June 1998

384 which

was sent under cover of a facsimile 385 referring to it as the contract 'we have been

166 The impacl of/he FAJ acquisilion

discussing over the last couple of years'. Thereafter every version of the slip referred to the intermediary as 'Carpenter Bowring' .386 The significance of this is that in December 1997 the Australia branch of the broking house changed its name from Carpenter Bowring to Guy Carpenter.387 Finally, a reference code was added to the slips being 'FAI/389/97' .388 There is no suggestion in the evidence that the reference to '1997' was a clerical error. It was used in all the subsequent drafts of the placement slips. 389 As noted previously, on 23 June 1998 Tuckfield had sent

Bromley a draft slip bearing the code 'FAI/389/98'.390

On 26 June 1998 Bromley sent an emaie91 to Tuckfield and requested further changes. He also reported that NI had provided an audit and review clause for inclusion in the slip. 392 Bromley told Tuckfield that he was unhappy with the clause and had re-drafted it and sent it to NI for approval. The clause proposed by NI required F AI to provide a copy of the term sheet (placement slip )393 to their

auditors and obtain a letter of non-objection from the regulators before 31 July 1998. 394 The clause also conferred on the reinsurer a right to request copies of the non-objection letters.

In the draft prepared by Bromley the clause removed any obligation on the reinsured to provide a copy of the term sheet to the regulator. Rather, it placed an obligation on the reinsured to warrant that they would provide the term sheet to the auditors, provide a copy of the auditor's letter of non-objection to the reinsurer upon request and further warrant to provide details of the contract (as opposed to a copy of the contract) to the regulator. 395

On 28 June 1998 Jain responded by facsimile to Bromley's request for changes to the disclosure and review clause.396 He told Bromley that his draft disclosure and review clause was acceptable if it included a further condition that F AI provide a letter of non-objection from their auditors and a representation of disclosure of the terms to the ISC by the end of July 1998. If this was not done NI would return the

premium and the cover would be null and void. Subject to this, Jain told Bromley the other changes requested in his 26 June 1998 fax 397 had been agreed to. A revised slip was attached to the fax. 398 Bromley advised Tuckfield of this by an email sent on the same day. 399

Bromley responded to Jain's facsimile of 28 June 1998 in a fax bearing the date 29 June 1998 although it has markings bearing the date 30 June 1998. 400 Bromley told Jain that F AI wanted two further modifications before the deal could proceed. First he advised Jain that the 'CPO ofF AI' required more risk in the contract for its approval by the auditors. Bromley told Jain that F AI would be prepared to pay an additional premium for this. Second, he also told Jain that F AI would not accept the clause in its current form. Bromley proposed a clause by which F AI would warrant

that by no later than 31 August 1998 it would provide 'full details' of the reinsurance to the ISC 'as required' by the Insurance Act and to their 'external auditors for review' and would further warrant that if any objections were raised they would immediately advise the reinsured.

The failure of HJH In surance 167

Bromley testified that he was unsure whether the reference to 'CFO' in the email was to Wilkie or Mainprize. His stated that he could recall Mainprize insisting there be a reduction in the earthquake trigger from a loss of $20 billion to $2 billion.401 Tuckfield also recalled a conversation with Mainprize in which Mainprize requested a reduction in the quantum of the 'trigger point' .402 Mainprize, however, denied having ever discussed the earthquake trigger clause with Bromley. He said he had no recollection of having any involvement in the negotiations for the contract. 403 He conceded he had a general understanding of the terms of the contract through discussions with Wilkie at the time. 404 But he said that he could not recall being aware that it contained a specific earthquake trigger clause. 405 Wilkie said he was on holidays from 28 June 1998 and thought Mainprize was then involved in the negotiation of the contract. 406

The revised version of the disclosure and review clause first prepared by Bromley removed NI's right to seek a letter of non-objection from the regulator or the auditors. It also removed the statement warranting that the term sheet had been provided to the regulator. Bromley's further amendments to the disclosure and review clause407 only required F AI to provide details of the contract to the ISC in accordance with the provisions of the Insurance Act. To comply with that clause F AI would only need to provide details of the contract to the ISC in accordance with their usual procedures for reinsurance. In relation to disclosure to the external auditors, Bromley's amendment required F AI to warrant that full details of the contract would be provided for review by 31 August 1998. The cumulative effect of these variations was to remove any obligation on the part ofF AI to provide advice of third party verification to NI. It left NI with only F AI' s assurance it would be done. There is nothing improper in this, but it takes on significance in relation to the explanations given for the preparation of a backdated cover note. I will return to the backdating a little later.

A further marked up version of the slip408 was sent to Bromley by NI on 29 June 1998. In a covering facsimile 409 NI made changes so that the limit for the

1997- 98 year was to be $50 million, and the subsequent years $15 million and the earthquake trigger would be $5 billion loss causing event. NI asked Bromley to draft the wording on changes to the limit of cover.

On 29 June 1998 Bromley forwarded NI's facsimile to Tuckfield with some handwritten comments. 410 Bromley told him that NI had agreed in principle to the

'initial' $50 million limit applying over more than one year. The effect of this was to enable F AI to use the limit as they need to 'eg. $20 million for 1997-98, $20 million for 98/99 = $10 million for 99/00 plus $15 000 000 for each of 1997- 98 onwards if the EQ occurs'. Bromley reported to Tuckfield that he had discussed the revised disclosure and review clause with Berkshire and would call him to discuss this matter. Tuckfield could not recall whether those discussions occurred.411

On 30 June 1998 Bromley emailed Tuckfield to report that he was awaiting written confirmation on discussions he had had earlier in the day and was expecting a revised version. 412 He told Tuckfield that he thought the disclosure and review

168 Th e impact of the FA! acquisition

clause would be accepted and that the only matter outstanding was the attachment point of the cover. Bromley then asked Tuckfield to follow this up with Trahair. Later that day Bromley provided Tuckfield with a copy of the placement slip that he had forwarded toNI for signing. 413 He noted that further changes had been made to the placement slip in that the attachment point on the cover had been reduced from

80 per cent to 70 per cent to ensure sufficient recoveries for the first year. 414 Wilkie agreed he gave that instruction. 415

On the evening of 30 June 1998 Bromley received the final placement slip signed by Stirling on behalf of NI. 416 The slip was later countersigned by Wilkie and Mainprize on behalf of FAI.417 The evidence as to the sequence of events surrounding the time, date and place that the slip was signed by Wilkie and Mainprize was contradictory.418 Whatever the sequence of events the date of acceptance on the slip was 30 June 1998. This date was inserted by Stirling when he

signed the slip.419 The signed slip reflected the final version of the disclosure and review clause agreed by the parties. The clause in its final form reflected the terms proposed by Bromley in his facsimile of 29 June 1998, subject to a requirement that F AI immediately notify NI of any objections from the ISC or their external auditors.

14.3.3 The Nl contract-its substance

The placement slip encompassed the various issues that had been negotiated between the parties throughout June. I will summarise them.

First, it had two sections of cover. The first was for $50 million over the life of the contract and the second, subject to a remote trigger (being the earthquake event), enabling further recoveries up to $30 million. The latter section was intended to create risk transfer sufficient for it to qualify as reinsurance. Recoveries could be made if the loss ratio exceeded 70 per cent of net premium income. Recoveries were restricted to an overall maximum of account recoverable $50 million 'per underwriting year of account' and subject to an overall aggregate of $80 million.

Second, the payment of recoveries were to be 'finally established' at 36 months from the inception of the relevant underwriting year of account. 420 The result was that any recoveries raised for the financial year 1997-98 could not be paid earlier than 30 June 2000. Recoveries for the financial year 1998-99 would not be paid earlier than 30 June 2001.

Third, the premium clause provided for the payment of cash premiums of $55 million over the life of the contract with $5.5 million payable for 1997-98 and payments of $12.3 7 5 million thereafter. The premium clause further provided that the premium was 'payable in advance on each ofthe five underwriting years'.421

Fourth, there was an offset clause which enabled each party to offset balances due to the other against any sums which are or may become due from the other party. 422 It further provided:

The failure of HIH Insurance 169

Reinsurer to have the specific right to offset any loss payments due hereunder against future premium payments which will become due to reinsurer, without discounting the amount of those premiums. [emphasis added]

This clause had particular significance. It was possible that as at 1 July 2000, assuming no earthquake event had occurred, F AI could seek a cash recovery from NI of the full amount of the cover being $50 million. However by that time they would have only paid $44.625 million in premiums and fees exposing NI to a credit

risk between that time and 1 July 2001, when the last premium payment of $I 2.375 million was due. The offset clause avoided this result by enabling NI to offset the future premium payment from F AI against any present obligation it had to pay recovenes.

Fifth, there was a commutation clause entitling F AI to commute at any time prior to 6 July 2001. 423 On commutation the reinsurer was to return the 'Experience Account Balance', which was effectively the premiums paid and the investment return on those premiums less A$2 million and A$600 000 for each underwriting year 1998-99, 1999-00, 2000-01 and 2001-02 which was referred to as the 'reinsurer's margin for risk transfer'.

Sixth, the document contained the disclosure and the review clause424 negotiated between Bromley and agreed to by Jain as described above.

Seventh, the placement slip included matters consistent with Tuckfield's comment to Bromley that the slip would need to be 'appropriately backdated' .425 It bore the date 4 July 1997426, it had an internal reference code 'FAI/389/97'A' and it had an intermediary clause referring to the intermediary as 'Carpenter Bowring Australia Pty Ltd'.427 On the other hand, when he signed the document Stirling from NI added the words 'Dated June 30, 1998'.428

14.3.4 Benefit to FAI?

The evidence demonstrates that the NI contract was negotiated and finalised with the primary purpose of enabling F AI to record a profit on the transaction in 1997/98 of between $20 million and $40 million so as to boost its results and possibly book a further profit in the financial year 1998-99. This was to be done by raising a recovery under the first section of the policy that exceeded the premium payable by the specified amount of desired profit. This was despite the fact that FA( would eventually fund that cover through subsequent years' premium. Wilkie agreed that this was his motive for securing the contract, being what he said Adler requested of him.429 Bromley and Tuckfield told the Commission that that was also their understanding and they continually referred to it in emails reporting on their d. . . h FAI 430 ISCUSSIOnS Wlt management. '

This intention is reflected in the terms of the placement slip and cover note. The trigger point of 70 per cent of net premium was chosen so that it was certain to be called upon. The calculation of recoveries in the contract was defined to be

170 The impact of th e FA/ acquisition

10 per cent of 'net earned premium431 , which was stated in the contract as being estimated at $700 million.432 If this was evenly distributed through out the 1997-98 underwriting year of account that would lead to a maximum recovery of approximately $35 million for the financial year ending 30 June 1998 (one half of

10 per cent of $700 million) if the loss ratio was 80 per cent or higher. According to Tuckfield the loss ratio advised to him by Trahair on 30 June 1998 was likely to lead to $20 million recovery. In the end result, it was calculated to be

$35.758 million. 433 Further, the premium payment arrangement was altered to increase the amount payable in the financial year 1997-98 which had the effect of increasing the net amount booked under the contract in that period.434

Ultimately a benefit of $29.078 million was booked. However, overall the contract was highly unlikely to be of commercial benefit to F AI. Pearson estimated that the chance of the earthquake event occurring over the life of the contract was in the range 1.8 to 3.3 per cent.435 If the earthquake did not occur the contract had a maximum value to F AI of negative $6.7 million. 436 F AI had not sought earthquake cover. It was suggested by NI.

Leaving aside the question of backdating, counsel assisting submitted that, to the knowledge of Wilkie, Bromley and Tuckfield, F AI through Wilkie sought and obtained the contract for the sole purpose of inflating its results at 30 June 1998 (and possibly 30 June 1999). It did so by entering into a contract which was overwhelmingly likely to be loss making and which did not confer any genuine benefit upon F AI. It was submitted that it was entered in circumstances that were

not in the interests of F AI and that this would have been known by all of Wilkie, Tuckfield and Bromley and thus Guy Carpenter.

A number of submissions was put in response to this, principally by Guy Carpenter, Bromley and Tuckfield. These included that the making of such a finding would breach the rules of natural justice because at no stage during the inquiry were they put on notice that a submission in those terms might be put or that it was a subject

matter being considered. It was also submitted that it would be wholly inappropriate to make such a finding where the NI policy involved no question of risk transfer being disguised (unlike the GCRA transaction), and it was otherwise just a policy of financial reinsurance with the benefits of financial reinsurance being widely accepted.

I do not accept the natural justice objection. However I do not propose to make a finding of the type adverted to. It cannot be the case that entering into a contract of financial reinsurance will, of itself, involve a breach of the duty owed by a director to the company. But the reverse is also true. It depends on the circumstances of the negotiations and the documentation. In this case I do not think it can be said that the NI policy was to Wilkie's knowledge worthless to FAI unless it can be said that he

knew it could not be accounted for as reinsurance. He was requested by Adler to obtain a contract and he did. The contract contained some risk transfer, albeit not much. There is no evidence from which I could infer that he knew that the desired accounting treatment was incapable of being obtained or only capable of being

The failure ofHIH Insurance 171

obtained if the auditors were misled. It is quite different from the GCRA transaction where, as I have found, there was an attempt to hide the absence of risk transfer and an apprehension that if the full transaction was disclosed then no beneficial accounting treatment could be obtained.

In my view the same applies to Bromley and Tuckfield although more would need to be established for them to be said to be knowingly involved in any breach of duty by Wilkie. As I understand the effect of Bromley and Tuckfield's evidence, it was that they were asked by their client to obtain a contract with the maximum possible risk transfer that a reinsurer will accept for cover that was obtained. They did so. They may have had doubts whether F AI could account for it in the way they wanted but they could not have been certain of the outcome. To my mind there is nothing improper or inappropriate in a broker (effectively) saying to his client in these circumstances 'here is a contract with some risk transfer, I have doubts you can account for it as you want but it is a matter for you'. Save for the backdating of the documents there was no contemplation that F AI could only achieve what it wanted if it deceived its auditors and there was no facilitation of that.

14.3.5 Tuckfield prepares a backdated cover note

On 1 July 1998 Tuckfield sent an email to Bromley. It stated:

Two things I'd like amended.

Can the slip say 19997/98 in arears [sic] others in advance [sic]. Don 't want anyone asking why the premium wasn't paid when due.

Is the last sentence of the Experience Account section really necessary? And I note the acceptance is dated 30/6/98. 437 [emphasis added]

Bromley replied:

As we discussed, I am happy that you change the cover note to reflect what we need and also the date of the Cover Note if this is possible within our system.438

The premium clause in the placement slip provided for payment of all the premiums in advance yet it was dated 4 July 1997. Cover commenced from the beginning of the 1997- 98 underwriting year, namely 1 July 1997. By 1 July 1998 it was therefore impossible for that payment to be made in accordance with the terms of the slip. Anyone who queried when payment was made might suspect that the contract was

not entered into in July 1997.

After he received Bromley's response Tuckfield set about preparing a cover note. 439 The cover note was prepared in the form of a letter to F AI. It was addressed to F Al at '77 Pacific Highway, North Sydney'. This was the address of FAI's insurance division only until the end of 1997.440 It began 'In accordance with your instructions we have arranged cover as follows'. Like the placement slip, the cover note bore the date 4 July 1997, contained a 1997 reference number being 'FAI/389/97'A' and named Carpenter Bowring as the intermediary. Unlike the placement slip, the cover

172 Th e impact of the FA! acquisition

note did not contain any reference to its being executed on 30 June 1998. Tuckfield also made the changes referred to in his email to Bromley, namely he altered the premium clause so that it referred to the first premium payment being payable in arrears and removed the last sentence of the definition of 'Experience Account Balance' .

44 1 In addition, he prepared the cover note on 'Carpenter Bowring' letterhead and signed off as a ' Director' of 'Carpenter Bowring Australia Pty Ltd' , even though at that date he was not a director of a company that bore that name. 442 Finally he altered the warranty clause so that the words 'an aggregate of' before

'AUD 50 000 000' were deleted. 443 There are other changes but they are of no significance.

During this inquiry there was significant debate as to what this cover note conveyed and why Tuckfield undertook these steps. The debate was complicated because the preparation of backdated cover appears not to have affected the auditors' approval of the proposed accounting of the contract, although I consider it was intended to.

Whatever else the backdated cover note sought to suggest, to me it clearly conveys that the document itself was created on or about 4 July 1997. No other conclusion is reasonably open given that it bears that date and every other aspect of the document is consistent with the document being created on or around that time. To that extent

the document is false. What is not as clear is the consequence of a cover note bearing that date. It could mean, for example, that there was a final contract entered into between FAI and NI on or about 4 July 1997 and that the cover note was issued on that date to record that fact. It could also mean that on or about 4 July 1997 an agreement in principle was reached which was sufficiently certain for the broker to issue a cover note but the final slip was still to be negotiated, Neither of these possible meanings is consistent with the true position, namely, that the negotiation

of the contract with NI was commenced and completed in June 1998.

In the absence of any satisfactory explanation, in my view it would be open to conclude that by the various steps he took Tuckfield intended that the misleading impression I have identified was conveyed by his cover note. Counsel assisting submitted that I should so find. They contended that Tuckfield prepared the cover note in the form he did with the intention of arming F AI management with a document which they could seek to convince the auditors that a contract was entered

into in early July 1997 such that the cover was wholly prospective and at the inception of the contract the level of recovery was uncertain. Counsel assisting pointed to the terms of the email of 1 July 1998 that passed between Tuckfield and Scrivens and in particular to the statement 'Don't want anyone asking why the premium wasn't paid when due'. Counsel assisting submitted that this must have

been a reference to the auditors of F AI who upon receipt of a cover note dated 4 July 1997 providing for premium payable in advance, might query why that premium was not paid during the 1997- 98 year.

Both Guy Carpenter and Tuckfield disputed that Tuckfield had any intention to mislead anyone when he prepared the cover note. They point to Tuckfield's explanation that he was merely putting into effect FAI's instructions. 444

I have

The failure of HIH Insurance 173

referred above to the evidence concerning the meeting between Tuckfield, Wilkie and Mainprize in which the instruction to backdate was said to have been given. It is not clear on the evidence whether the instruction was to 'backdate cover' or backdate the contract. Either way, it raises a question as to what use Tuckfield thought would be made of the document he prepared.

I think the email of 1 July 1998 is revealing. Tuckfield suggested that the reference to 'anyone' in his email was to the 'people' in NI and PAl's accounts departments because they might otherwise be confused by the slip providing that premiums were payable in advance when the first year's premium was not paid. 445 I find another

interpretation more appealing. The cover note was not sent to NI. No one there would be disabused of any confusion that may have arisen. F AI management already had the placement slip which indicated it was signed on 30 June 1998. That was the reason the premium was not paid earlier. I cannot understand why anyone, including anyone in the accounts department at F AI, would be confused about that. I think Tuckfield had in mind that the auditors ofF AI might ask questions about the timing of the contract if they realised that premiums were supposed to have been paid in advance but had not been paid during the 1997-98 year.

Tuckfield said he drew comfort from the terms of the 'disclosure and review' clause in the NI policy in that it required F AI to provide full details of this reinsurance to the ISC and their auditors. 446 He asserted that he did not anticipate the possibility of any 'sharp practice' in the quality of the information which FAI would provide to its auditors or the ISC. 447 But Tuckfield was well aware of the negotiation of that clause. 448 F AI had sought to water down the various disclosure clauses in the contracts being negotiated. It eventually obtained a clause that did not require to produce to NI any document emanating from a third party (for example the ISC or their auditors) setting out its satisfaction with the policy. I do not see how Tuckfield could draw any comfort from the final form of the clause on this issue.

Guy Carpenter and Tuckfield also contended that evidence given by Scrivens, the relevant audit partner at Andersen, concerning what Tuckfield told him in August 1998 in relation to the timing of the NI contract was inconsistent with the intention attributed by counsel assisting to Tuckfield at the time of the preparation of the cover note. The statement attributed by Scrivens to Tuckfield in August 1998 was as follows:

Q.

A.

174

Just restricting yourself to this particular subject, what is it that you recall Mr Tuckfield said to you on that occasion about that subject?

I asked Mr Tuckfield why this contract was covered only by a cover note and why there wasn't a contract in place. He said that it was in the process of fjnalisation with National Indemnity, but negotiations had been going on much earlier in the year, particularly in determining using the actuaries and assessing the premiums under the contract.

449

Th e impact of the FA! acquisition

Tuckfield denied making any such statement to Scrivens. Even if he did, I do not think that the making of such a statement in August 1998 is inconsistent with him creating a misleading cover note. The existence of a cover note created on or about 4 July 1997 still leaves a number of possibilities as to when an agreement between the parties was finalised and then when that final agreement was documented. But it

is false nevertheless.

There was also debate as to whether Tuckfield, at the time he created the cover note could have anticipated that its apparent date of creation would be, or could be, material to the auditors' assessment of its accounting treatment. The lengths that he went to in preparing it lead me to infer that he must have thought that.

In any event it is sufficient for me to say that the fact of creation of a cover note on 4 July 1997 could be considered material. At the very least it conveys that parties had developed the proposal to such a point at that date that a cover note could be issued. Such a consequence could be seen as relevant to an assessment of the extent

of risk transfer. With a contract being negotiated in early July 1997 the assessment of the likelihood that an earthquake event would occur would be undertaken over a longer period than if the contract was negotiated and entered into in late June 1998. More significantly, the fact that a reinsurance contract was negotiated in June 1998

and signed on 30 June 1998 and that a substantial recovery would be raised for the financial year ending on that day was more likely to arouse scrutiny than a contract giving rise to a similar recovery negotiated in early July 1997. The former might suggest a contract contrived to obtain a particular accounting result because it related to a known deterioration of past results, as the NI contract was. The latter

might be taken as suggestive of a contract entered into to protect against a possible deterioration in future results.

I find that in preparing the cover note in the way he did Tuckfield knew it was misleading. I also find that he contemplated that F AI management might use it to mislead their auditors to accept it is a cover note created on 4 July 1997.

14.3.6 Bromley and the backdating of the cover note

Counsel assisting also submitted that Bromley, in approving the backdating of the cover note, shared Tuckfield's knowledge that it might be misleading and his understanding as to what F AI might do with it. I do not accept that submission and make no such finding. The submission rested solely upon Bromley's reply to Tuckfield' s email of 1 July 1998. To my mind this is an insufficient basis to make such a finding. I note that there was no evidence that Bromley was aware of the various steps Tuckfield took in preparing the cover note.

Th e failure of HIH Insurance 175

14.4

14.4.1

The preparation of the APRA returns and FAI's accounts for the year ending 30 June 1998

The APRA returns

In July 1998 Roger Colomb prepared FAI's annual reinsurance returns for submission to APRA. Each page of the returns and the various annexures were signed by Wilkie. 45° Colomb then submitted them to APRA. 451 They included a form 403, 'statement of reinsurance-proportional/non proportional treaties', which provided various information on treaties in a table format with each of the reported reinsurance treaties identified by a code. Included in the table was information concerning four of the six 1 May 1998 contracts452, the NI contract453 and the second AXOL contract. 454 An attachment to the form 403 gave further information in relation to the separate sections of cover provided under the second AXOL. 455 Copies of the treaties were not provided nor were they required to be.

Also provided to APRA was a form 401, 'statement of reinsurance guideline compliance of proposed arrangement' which posed various questions in relation to the responding insurers' reinsurance programme overall. 456 It included the answer 'no' to the following question:

Is any treaty contract subject to a special clause which acts to limit reinsurers liability. If yes, attach a copy of each clause and identify the treaty.

Among other contracts it was referring to four of the six 1 May 1998 contracts which were subject to the 1 May 1998 side letter and the second AXOL contract which was affected by the 26 June 1998 side letter. Wilkie was asked about the effect of the two side letters on the negative response. Wilkie reiterated his understanding that he did not regard the arrangements as necessarily limiting the reinsurers' liability457 and that it was his expectation that the returns had been prepared properly. 458 I do not accept this answer. I have already addressed Wilkie's understanding of the meaning and effect of the side letter. Wilkie claimed that he did not read the return. 459 Given the significance and proximity of the negotiations for the reinsurance contracts I consider it likely he read the forms quite closely. He did not have any reason to believe that Colomb was aware of them. He knew Colomb was not involved in the negotiation of the GCRA contract. I think Wilkie signed his name to FAI's APRA form 401 in July 1998 knowing the answer to the above question was misleading.

14.4.2 The accounts-overview

The Commission did not have an extensive documentary trail which enabled the precise history ofFAl's accounting treatment for the GCRA and NI contracts for the year ending 30 June 1998 to be reconstructed. In general terms, however, that process involved F AI determining the quantum of recoveries and premiums that would be booked for the year ending 30 June 1998. F AI proposed the accounting

176 The impact of the FA/ acquisition

treatment to Andersen and provided Andersen with such documentary material as it considered appropriate. Meetings were held with Andersen in which F AI argued in favour of its accounting treatment before presenting the proposed accounts to the audit committee and the board ofFAl for approval.

14.4.3 Determining the recoveries and premiums to be booked

With the second AXOL contract, F AI took up the full extent of the cover provided by each of sections 4 and 5 ($11 million each).460 These sections were only included at the end of June 1998. It seems likely that F AI specifically sought those sections of cover in the knowledge that there had been a significant deterioration in that business for that year. Section 2 of the second AXOL contract provided cover for adverse development in MIPI and ALAS. It allowed FAI 'flexibility' in that FAI could choose to raise recoveries for paid claims that exceeded the retention level or

raise recoveries for both paid claims and any increase in reserves they chose to make. The latter course involved the risk that, if the amount of the reserves increase was too large, suspicion might be raised that there had been under-reserving in the past. This would heighten the risk for F AI if the auditors did not accept the proposed accounting treatment. 461 By 22 June 1998 FAI had decided to play it safe. F AI staff were advised that payments in 1998 equated to recoveries. 462 Recoveries of $9.574 million were raised in respect of Section 2 of the second AXOL contract

being payments since 1 January 1998.463 FAI had sought the AXOL contracts to enable it to stage manage an increase in reserves over time. This is not what F AI did as at 30 June 1998 .

As I have previously said the total premium due as at 30 June 1998 was $7.45 million. For reasons that were never explained the amount of premiums actually booked by FAI as at 30 June 1998 was $2.75 million.

Trahair calculated the recovery under the NI contract as at 30 June 1998 for the underwriting year of account 1997-98 to be $34.578 million.464 The premium due for the NI contract was $5.5 million. This amount was booked.465

Thus for year ending 30 June 1998 FAI booked a profit of $28.824 million under the second AXOL contract and $29.078 million under the NI contract. The result of booking profits in those amounts for year ending 30 June 1998 was that it guaranteed F AI would incur substantial losses over the remaining life of each contract even if maximum recoveries were obtained.

14.4. 4 Deal ings with Andersen

Some time in late July 1998, Daniel Vanderkemp (an audit senior for Andersen) was handed a draft summary of the underwriting accounts by Marc Gross, a senior accountant with F AI.466 The summary disclosed an unusually high level of reinsurance recoveries. Vanderkemp asked Gross for copies of the contracts and

related calculations referable to those recoveries. Within a few days he was provided with copies of the second AXOL contract, the NI cover note, and a memo from

Th e failure of HIH Insurance 177

Trahair calculating the recoveries under the NI contract467 and spreadsheets supporting Trahair' s calculations.468 He was not provided with or informed of the existence of the NI placement slip, the six 1 May 1998 contracts or any of the side letters. I do not suggest that Gross was aware of their existence.

After considering these documents, Vanderkemp and Anita Lee (an audit manager for Andersen) had a number of meetings with Peiris and Gross to discuss the contracts and FAI's proposed accounting treatment. 469 During a meeting on 13 August 1998 Peiris and Gross insisted that the contracts were true reinsurance and thatFAI's proposed accounting treatment was in accordance with AASB 1023 . They told Vanderkemp that, because reinsurance recoveries were uneven, F AI was unable and could not be expected to calculate future recoveries with any certainty in the current year and therefore could not evenly match premium against reinsurance recoveries. 470

To document the accounting issues arising out of the two contracts Vanderkemp prepared a number of memos.471 Vanderkemp interpreted the NI contract as allowing $80 million of cover over the life of the contract, with the secondary trigger only operating as a limit on recoveries in any one underwriting year. Vanderkemp noted that that both contracts appeared to be outside the norm given the relatively high amount of premiums being paid for the relatively low amount of cover being provided. In relation to the NI contract, he considered that given the trigger was a loss ratio of only 70 per cent it seemed likely that the total amount of

recoveries (which he thought was $80 million) was likely to be claimed. He also concluded that this was the case for the second AXOL contract.472

Vanderkemp noted the proposed accounting treatment for each contract would result in there being a known loss for the remaining years. On the best case for F AI, the available recoveries would be exceeded by premiums still to be paid and the effect of the accounting treatment was to increase income into the current year at the expense of later years. Vanderkemp compared F AI' s proposed accounting treatment with two alternatives, one involving the expensing of premiums in the same proportion as recoveries were booked, and the other expensing the premium equally over the five years of each agreement. Each of these approaches produced a substantially lower profit on the contracts in the year 1997-98.473 Vanderkemp also calculated the internal rate of return on the contracts. He concluded that both contracts were more akin to deposit transactions than reinsurance474 and considered that they did not qualifY as reinsurance under the US standards.

In mid-August 1998 Vanderkemp discussed his concerns with Alan Davies, the 'relationship partner' for the 1998 F AI audit, and then on 17 August 1998, with Scrivens. 475 Vanderkemp showed Scrivens the relevant documents including the GCRA contract, the NI cover note,. the spreadsheets, his memoranda and AASB 1023. 4 76 After these discussions Scrivens realised he had two matters to determine: whether the contracts involved risk transfer and, if they did, what was

h

. . 477

t e appropnate accountmg treatment.

178 The impact of the FAI acquisition

Scrivens discussed the matter further with Vanderkemp. Vanderkemp's analysis of the NI contract quantified the net present value of the cash flows under the contract (assuming recoveries of $80 million) as 15 per cent of premiums. Scriven's felt this exceeded any achievable interest rate. He therefore concluded that the NI contract

involved a material transfer of risk. 478 Scrivens also noted that Vanderkemp's calculation of an internal rate of return under the second AXOL contract was based on what he considered to be an incorrect assumption, namely that F AI would pay the premiums rather than hold the premiums and thus have the opportunity to earn

interest on the monies. 479 Having regard to the gap of approximately $10 million between maximum recoveries and premiums payable he considered that it also involved a transfer of risk. 480 Scrivens also discussed the contracts with colleagues in London. 48 1

In their meeting of 13 August 1998 Peiris became aware ofVanderkemp's concerns about F AI' s proposed accounting treatment for the two contracts. 482 F AI management began to document their position on the contracts and the accounting treatment. I infer that was for the purpose of ensuring they all adopted a consistent position with the auditors.

On 19 August 1998 Trahair sent a memo to Wilkie, which was copied to Mainprize and Peiris.4 83 Trahair reminded the recipients of the nature of the cover provided by the two contracts and the recoveries and premiums expensed. He discussed what he understood were the issues that had arisen with the auditors in relation to the NI contract, including whether the recoveries should be discounted. 484 He stated that

Andersen's approach was an attempt 'to retrospectively adjust the terms of the NI deal in light of the facts that have emerged'. F AI was to repeat this argument on a number of occasions.

Meetings were arranged with Andersen on 21 August 1998. In anticipation of the meetings, briefing notes were prepared by Trahair and Peiris in the form of questions and answers concerning the NI and GCRA contracts.485 On 20 or 21 August 1998 the briefing notes were circulated to Wilkie and Mainprize. In

relation to the NI contract the briefing notes stated:

Q. When was the contract taken out?

A. We've had discussions over the last several years with

Carpenter Bowring - John Tuckfield and Geoff Bromley about the sort of deal we would like to put in place. They've been able to help us set up a deal for 5 years commencing in 1997/98 with National Indemnity. The cover note is dated 4 July 1997,

although discussions took place during the year to fine tune the details, as it would with any normal reinsurance treaty.

Q. Is it reinsurance?

A. Yes. The deal is prospectively based for 5 years . ..

The way the contract works is that over the period we effectively pre-fund a claim, according to the pattern of risk. It may have happened that we went 2 or 3 years into the contract period before

Th e failure of HJH Insurance 179

making a claim. In that event we would have paid, say $18-$30 million and then may have needed to call upon the contract to fund say 5% or perhaps $3 5 millio n.

In the event we've had to call upon the contract in the first year. We can't, however, go back and reset the conditions of the contract in light of what has actually happened.486

These statements do not accord with what happened. The first answer and the reference to the deal being 'prospectively based for five years' suggest that an agreement was reached at the beginning of the financial year, namely about 4 July 1997, subject to fine tuning during the year. The discussions with NI did not

begin until 23 June 1998. The briefing notes also suggest that the claims experience was entirely undeveloped at the time that contract was entered into. This explains the comment that the contract may have run two or three years before claims were made. But when the contract was negotiated it was understood by all the parties that there would be a large recovery made for the financial year ending 30 June 1998.

The briefing notes referred to the second AXOL contract as being a 'multi-year deal covering a range of retroactive and prospective covers relating to existing PI/PL portfolios' with the 'retroactive component [seeking] to protect FAI against further adverse development in domestic PI and PL and specifically in two overseas PI programs'. The briefing notes then include a section headed 'Accounting' which was added by Peiris.487 Peiris argued that the premium expense being recognised by FAI was 'in line with the pattern of reinsurance services', that 'recognition of the premium expense on any other basis would not give effect to the original intention of the parties to the contract', and that to require F AI to adopt a different method of recognising premium expense 'would be to reinterpret the intention of the parties by reference to subsequent events'. The difficulty with this is that it was understood by the persons who had negotiated the contract what the pattern of recoveries would be in its first year.

On 21 August 1998 Scrivens met Wilkie and Peiris, and then met Peiris and Mainprize.488 Scrivens said that in both meetings he was assured there was effective transfer of risk under each contract. He said that Wilkie told him that they entered into the contracts to protect FAI's. 'bottom line from erratic claims development' so as to give FAI some certainty in its ability to 'underwrite risks going forward' .489 I accept that this was said. Scrivens also said that Peiris told him that F AI' s proposed accounting treatment was in accordance with AASB 1023 and that there was no relevant accounting standard which required F AI to recognise future premium costs in another reporting period. 490 During neither of these meetings was Scrivens advised of the existence of the 1 May 1998 side letter, the six

1 May 1998 contracts, the connection 'between the AXOL contracts and the six 1 May 1998 contracts, the 26 June side letter or the period over which the NI contract was negotiated. 49 1

180 Th e impact of the FAJ acquisition

About 25 August 1998 Scrivens had a discussion with Mainprize and Adler in which he advised them that he remained unconvinced that FAI's proposed accounting treatment was 'accepted industry practice'. Adler suggested that they would set up a meeting with the relevant underwriters. 492 It was arranged for the following day.

In anticipation of that meeting Peiris circulated another email to Adler, Mainprize, Wilkie and Trahair setting out management' s position. 493 The email summarised the terms of the NI and GCRA contracts. In relation to the NI contract, Peiris recorded that the limit of recovery was $80 million in total and made no reference to the earthquake trigger. Under the heading 'Business Rationale', Peiris asserted that FAI entered into the NI contract to protect its 'total underwriting result from any adverse development', which involved it negotiating a loss ratio which would 'give [FAI]

this certainty' and that the reinsurance recovery recognised 'is merely a function of the underlying loss ratio'. Again, this statement was not an accurate reflection ofthe true position. Reinsurance recoveries were determined first and the loss ratio was set with a view to obtaining that recovery, so as to achieve a specific 'net betterment' as desired by F AI.

With the GCRA AXOL contract, Peiris asserted that its business rationale was that 'F AI wanted to protect itself against further adverse development/new notified claims'. The reality was that the contract was initially sought to protect against already notified but not recognised claims. Peiris also summarised what he

understood Andersen's position was, namely that FAI could not claim in year one of the contracts 'a greater net benefit than the total net benefit under the two contracts' and that they were seeking an adjustment down to reflect that. 494 In response to this, Peiris reiterated the position previously outlined, namely Andersen's approach

'effectively reinterprets the intention of the contracts by reference to the size of recoveries made in year 1 '. He added that his summary should be read in conjunction with the briefing paper produced 'at last Friday's meeting'. This was a reference to the questions and answers to which I have already referred.

14.4.5 Meeting on 26 August 1998

On the morning of Wednesday 26 August 1998 there was a meeting or meetings involving Scrivens, Wilkie, Trahair, Peiris, Mainprize, Barnum and Smith of GCRA and Tuckfield.495 Each witness gave a different version of what occurred at that meeting and there was some dispute as to whether Barnum and Tuckfield were present at the meeting at the same time. Most of the witnesses had difficulty

recalling who was present and their recollections were generally vague.

Scrivens gave evidence that Barnum and Tuckfield were present at the same time during the meeting. He recalled that there was a number of matters discussed with them which were of considerable importance to his ultimate decision to approve the accounting treatment that F AI proposed for each contract. The substance of these

matters included:

The failure of HIH Insurance 181

• an assertion that Tuckfield told him that the NI contract had not been finalised but was in its final stages of preparation.496 Further, Tuckfield told him the negotiation of the NI contract 'had been going on much earlier in the year, particularly in determining and using the actuaries and assessing the premiums

under the contract'49 7

• Scrivens said that he advised Barnum and Tuckfield that F AI was proposing to raise $30 million in recoveries against each of the contracts and that 'this is in excess of the total reinsured amount net of reinsurance premium payable under the contract in each case ' 498

• Barnum said GCRA had made a considerable effort to structure the contract to 'ensure the premium pattern reflects our expectation of how claims would arise under the reinsurance contract' 499

• that in effect Barnum and Tuckfield confirmed that the relevant reinsurer would account for it in a similar way-namely, raise a loss in one year that exceeded the net possible loss under the transaction500

• that Barnum assured him this 'is absolutely normal accounting treatment in the reinsurance industry'. 501

These assertions were disputed by both Tuckfield and Barnum and the evidence of the other witnesses was contradictory and vague. I do not propose to summarise the evidence. Having reviewed all of the evidence and considered the various submissions I am not prepared to make any adverse findings against Tuckfield, Barnum or Smith based upon what was alleged to have been said or not said during this meeting. The only matter about which I am confident is that during this meeting no one told Scrivens of the existence of the 1 May 1998 side letter or the six

1 May 1998 contracts. However, in the case of Barnum and Smith, I am not satisfied that they were conscious of the need to disclose those matters to Scrivens. They were called to the meeting on short notice. They told me that they thought the meeting was to discuss accounting for reinsurance generally, as opposed to the AXOL contract in particular.

I have not overlooked the documents referred to me by counsel assrstmg and counsel for Andersen as corroborating Scrivens's evidence. Perhaps the most contentious of these was a memo prepared by Scrivens sometime between 27 August 1998 and 27 September 1998502 recording his decision to accept ·FAI's proposed accounting treatment. While I accept that Scrivens came away from the meeting with certain impressions, many of which are referred to in the memorandum, I am unable to find that anything said by Barnum, Smith or Tuckfield created those impressions. The memo is inaccurate in some respects. There are too many uncertainties for me to afford it much weight.

I was also referred to Gooley's handwritten file note which records a conversation between himself and Scrivens on 26 August 1998. It states:

182 Th e impact of the F AI acquisition

F AI-M. Scrivens

Re insurance broker and reinsurer

(Say it is normal to ale as F AI wants to do)

(Says that .reinsurers will ale on similar basis) 503

[emphasis added]

This is not sufficient to resolve my doubts as to what was said at the meeting. It is equally consistent with Scrivens interpreting what was said as confirmation of these matters rather than faithfully relaying them.

Counsel assisting took me to a position paper prepared by either Peiris and Trahair or just Peiris 504 and provided to Andersen on 27 August 1998.505 It contained the following statements which were submitted as being corroborative of much of what Scrivens said he was told on 26 August 1998:

The issue of the proper accounting treatments for these transactions was considered and it was agreed that industry practice would apply.

In subsequent discussions with the relevant broker/insurer, it was indicated that:

reinsurers account for premiums received over the period of cover and record a claims expense as and when they receive a notification from the re-insured p.e. a mirror image of the expected treatment by the reinsured/0

The treatment adopted by FAI recognises that FAI has transferred both underwriting risk and timing risk under the contracts. That is, FAI's accounting treatment allows the calculation of reinsurance recoveries in any particular year to exceed maximum possible ' benefit' that could flow from the contracts, on the basis that that the contracts cater for both timing risk and underwriting risk. This position has been clarified by our reinsurers and brokers as beinfo the treatment they would adopt in accounting for th ese transactions. 07 [emphasis added]

In the circumstances I have to treat with great caution the accuracy of any document emanating out of F AI during this period. I do not regard this document as a sufficient basis for what would be a very serious finding against any of Barnum, Smith or Tuckfield.

14.4.6 The pos ition paper

It is also necessary to note certain other statements in the position paper sent to Scrivens after the meeting on 26 August 1998. The section on 'Business Rationale ' for the GCRA and NI contracts included the following:

The GCR contract was entered into to protect F AI against further adverse developments/new notified claims arising from certain classes of business written in prior years ...

The failure of HIH Insurance 183

The National Indemnity contract was entered into to protect the Group's underwriting result as a whole ...

The treaties were concluded after extensive negotiations and due diligence investigations by the relevant broker/reinsurer and their internal actuaries. 508

The NI contract was put in place to book profits for the year ending 30 June 1998. The GCRA contracts were entered into to offset known losses from already notified claims. Neither of them sought to protect against any potential further adverse development. At no stage did Guy Carpenter or NI undertake a 'due diligence' or have the matter considered by an 'internal actuary'.

Under the heading 'Accounting', the position paper made similar points to those made in Peiris' email of25 August 1998 when summarising FAI's position and that of Andersen. The section entitled 'Difference in Position' includes this statement:

The Arthur Andersen view [i.e. that FAI should not be asked to book a benefit profit in year one that exceed the net benefit available over the life of the contract(s)] only deals with total underwriting risk and have sought to reinterpret the intention of the contracts by reference to the subsequent experience of the F AI Group. 509 [emphasis added]

I have already expressed my reservations about that proposition.

14.4.7 Scrivens's conclusion and dealings with the audit committee

Following the meeting on 26 August 1998, Scrivens had a number of discussions with colleagues within Andersen in Australia and overseas. He also discussed the matter with a former partner, Justin Gardener. Scrivens decided to approve the proposed accounting treatment and recorded his reasons for doing so in the

memorandum I have noted above. I will consider Scriven's deliberations in more detail shortly.

The year end results for F AI were due to be discussed at a meeting of the audit committee of the F AI board on 3 September 1998. Shortly before the meeting the committee was provided with Andersen's presentation relating to the audit of the draft results prepared by management.

5 10 In a section of the report entitled 'Year

End Matters', the audit committee was advised of the existence of two five-year reinsurance contracts and the proposed accounting treatment for those contracts. They were advised that the maximum benefit obtainable under those contracts was $34.5 million and that F AI was proposing to book an overall profit of

$58.15 million for year ending 30 June 1998. The presentation also stated that:

The differential of ($58.15 m less $34.5 m) $23 .65 m represents a shortfall which will result in a guaranteed ac;counting loss in the next 4 years .

Accounting treatment is acceptable practice within the industry. 511

On 3 September 1998 Scrivens attended the meeting of the audit committee of FAI. 512 Various witnesses gave different accounts as to what was said during that

184 The impact of the FA! acquisition

meeting. As I do not propose to resolve all the discrepancies I will only refer to parts of the evidence that describe generally the positions adopted by the various participants.

Scrivens said that in his presentation to the audit committee he specifically commented on the GCRA and NI contracts. He brought to the committee's attention the fact that the net benefit booked under the contract for the year ending 30 June 1998 exceeded the maximum benefit obtainable under both contracts of $34.5 million. This wrongly assumed an $80 million recovery available under the NI contract. He pointed out that the difference would be a shortfall which would

result in an accounting loss for the next four years. 513 He mentioned to the meeting his discussions with the reinsurers and his conclusion that there was effective risk transfer. He also provided a breakdown of the premium and recoveries booked compared to the maximum benefits payable under the contracts. Scrivens said he advised the meeting that there would be a $23.65 million shortfall which would

inevitably result in an accounting loss over the next four years. 5 14

Scrivens said that Mainprize advised the audit committee that there had been meetings with the reinsurers, including GCRA, and 'they have confirmed very clearly that this is the way they account for this type of contract' . He said that Peiris confirmed that the contracts transferred reinsurance risk and timing risk. 515

According to Scrivens, Hill commented that the approach appeared to be ' strange accounting' but added 'but if that's the way the industry deals with it, so be it' .51 6 Scrivens said he advised the committee of the need to ensure that the disclosure of the reinsurance arrangements in the accounts was appropriate and that Hill indicated that Mainprize and Peiris would need to 'propose the appropriate notes'. 517

Hill disputed that Scrivens said anything to the meeting concerning ' effective risk transfer of risk'. Nor could he recall Scrivens explaining the breakdown of the future shortfall of $23.65 million or Mainprize referring to meetings with reinsurers. He recalled that he and Harris expressed concern with the contents of the note to the

draft accounts. He said they were not taken to the details of the reinsurance contracts but instead relied upon Andersen's expertise in relation to both the appropriate accounting treatment and for advice as to whether the accounts (including the notes) complied with the relevant accounting standards.518 Hill denied saying anything to the effect that what was proposed 'looks like strange accounting', but accepted that

there was some concern raised as to the appropriateness of the proposed accounting treatment. He said he accepted the treatment recommended by both management and the external auditors. 519

Hill also said that he wanted the notes to the accounts amended so that they reflected 'the fact that the reinsurance contracts of the company included long term contracts'. He said he was satisfied that the final form of the note advised a reader of the accounts accordingly. 520

Harris' recollection was that the discussion at the audit committee meeting in relation to the reinsurance contracts was solely concerned with their accounting

Th e failure ofHIH insurance 185

treatment. 521 He said he understood the effect of the accounting treatment was that there would be a guaranteed future loss 522 with the consequence that he found the 'whole series of numbers [in the audit presentation] somewhat surprising'. 523 He also recalled Hill making a statement to the effect that it was not the accounting treatment that one would expect and that there should be a clear statement in the accounts that the 'contracts were of a long term nature' . 524 He said it was agreed that management and auditors would together frame an appropriate note. 525

Peiris did not give any evidence to the Commission which would assist in resolving differences in the evidence given by Hill and Scrivens. He did state that the reference in the minutes to the need for a note to be expanded to 'highlight the practice associated with these contract's was a reference to the guaranteed future accounting loss. 526 However he could not explain why it was that the note he assisted in preparing did not address that practice but instead referred to the long­ term nature of the reinsurance contracts. 527 Peiris said that he was aware that Andersen's calculation of the known future loss as being $23.65 million was predicated on at least one or two earthquakes of the prescribed magnitude. 528 Although he conceded it would have been important for the audit committee and the

board to have been made aware of that matter when considering the proposed accounting treatment, neither he nor anyone else advised them. 529

One point that was drawn specifically to the attention of the non-executive members of the audit meeting was the fact that the proposed accounting treatment would lead to a 'guaranteed accounting loss' of $23.65 million in future years. No other accounting question or question of risk transfer was raised for discussion. No alternative accounting treatment was put forward to the audit committee. Nor was any perceived ambiguity in the relevant accounting standard and its consequence for the accounts raised with the audit committee.530 I am satisfied that Hill raised a query or made some comment to the effect that the proposed accounting treatment was unusual and that it was agreed that there would be a clarifying note to the accounts which was meant to highlight the long-term nature of the reinsurance contracts. Although the evidence is unclear as to whether the guaranteed future Joss was also meant to be highlighted in the note, the fact is it did not do so.

The minutes of the meeting record that the audit committee agreed to place the draft 1998 financial statements before the full board 'for their consideration and adoption'. 531 The minutes contain the following entry in relation to Scrivens's presentation in relation to reinsurance:

186

The year end matters raised in the Audit Report were then discussed. The treatment of the reinsurance contracts was reviewed and it was agreed that the accounting policies note should be expanded to highlight the practice associated with these contracts.532

Th e impact of the FAI acquisition

14.4.8 Board adoption

The preliminary final report for the year ending 30 June 1998 was tabled at a meeting of the board of FAI on 9 September 1998.533 The board resolved to adopt the accounts and authorised Adler and Mainprize to sign.534 It also resolved to appoint Landerer and Adler to finalise the letter that would accompany the financi al statements and determine if a final dividend should be paid.535 The minutes make no

reference to the reinsurance contracts or their accounting treatment.

On 9 September 1998 Baulderstone sent to the Australian Stock Exchange a letter signed by Adler announcing the results 536 and a copy of the final accounts537 signed by Adler and Mainprize538 and by Scrivens. 539 Adler's letter advised the exchange that FAI had 'incurred' an operating profit before tax of $8.6 million but that, after deducting abnormal items, the loss was $4.7 million. 540 The accounts included some $31.574 million of recoveries under the GCRA AXOL contract.54 1 Consistent with my earlier findings I am satisfied that Mainprize signed off on the accounts when he knew that they contained a false entry being those recoveries.542

The accounts contained the following note in relation to the accounting treatment of recoveries under the reinsurance contracts:

Premium ceded to reinsurers is recognised as an expense in accordance with the pattern of reinsurance service received. Accordingly, a portion of outwards reinsurance premium is prepaid and deducted from the provision for unearned premium. The economic entity has entered into a number of

long term reinsurance contracts under which the premiums expense in future periods will be in accordance with the pattern of

reinsurance services received. 43

Under the heading 'Outstanding Claims and Reinsurance Recoveries' the following note appeared:

Reinsurance and other recoveries on paid claims, reported claims not yet paid and claims incurred but not reported are recognised as revenue. Recoveries receivable are assessed in a manner similar to the assessment of outstanding claims. Reinsurance recoveries on long term reinsurance

contracts are recognised on the same basis. 544

Peiris said that he drafted these notes and they were reviewed by Mainprize.545 Scrivens told the Commission that he observed those notes prior to signing the accounts. He considered that they adequately disclosed the treatment of the NI and GCRA contracts and did not misstate the financial position ofF AI.

546

Three matters arise out ofF AI' s management successful efforts to book in excess of $57 million of profit under the GCRA AXOL and NI contracts for year ending 30 June 1998. First, how should they have been accounted for? Second, is it correct, as they assert, that Andersen were misled both by management and by others? Third, was Andersen's audit approval of the accounting treatment, including its

deliberations with the audit committee, appropriate?

The failure of HIH Insurance 187

14.4.9 The appropriate actuarial and accounting treatment

The Commission received reports from Pearson concerning risk transfer under the GCRA and NI contracts. 547 Reports were also received from Couttas addressing the appropriate accounting treatment of those contracts in light of Pearson's conclusions

548 and Andersen's audit approval of FAI and HIH's accounting treatment for those contracts for the financial years 1988, 1999 and 2000. 549

The starting point for this discussion is the accounting standards. AASB 1 023 is specifically directed to 'Financial Reporting of General Insurance Activities ' . Section 7 is headed 'Reinsurance premiums'. Clauses 7.1, 7.1.1 and 7.1.5 of AASB 1023 respectively provide:

7.1 Premium ceded to reinsurers must be recognised by the direct insurer as outward reinsurance premium expense in accordance with the pattern of reinsurance service. Retrocessions must be recognised by the ceding reinsurers expenses in accordance with a pattern of reinsurance service.

7 .1.1 It is common for insurers to reinsure a portion of the risks that they accept ...

7.1.5 Outwards reinsurance premium expense will need to be recognised in the profit and loss account in accordance with the pattern of reinsurance service. For proportional reinsurance, this would normally be consistent with the pattern of risk of the underlying direct insurance policies. For non-proportional reinsurance, the expense would normally be recognised over the period of the re-insurance arrangement.

I make three observations concerning this part of the standard. First Couttas stressed that AASB 1023 does not stand alone. In particular he emphasised the requirements of AASB 1001 which require that any accounting treatment reflect the substance of the transaction where substance and form differ. A consequence of this was the conclusion he reached concerning the method of assessing risk transfer under the GCRA AXOL contract described above. I have already accepted that approach as correct.

Second, AASB 1023 does not expressly address the question of the quantum of risk transfer necessary for a particular contract to be considered 'reinsurance' for the purposes of the standards. It merely refers to a 'portion of the risk' being reinsured or risk being 'ceded'. Pearson, in consultation with Couttas550, applied the test in the US accounting standards. This involves an assessment whether 'significant

insurance risk' passed and whether there is a 'reasonable possibility that the reinsurer may realise a significant loss'. 551 In applying that test Pearson was guided by a US industry 'rule of thumb' being the so called ' ten/ten rule,' namely, whether there is a 10 per cent chance that the insurer will lose more than 10 per cent. 552 Counsel for Andersen criticised both P!!arson and Couttas for this approach. They submitted that the accounting standards do not refer to such a requirement and that the test adopted is too subjective. It is implicit in those submissions that any risk transfer was sufficient provi ded that it was not insignificant.

188 The impact of the FAI acquisition

This may be a valid criticism of the accounting standard. But I do not believe it undermines the evidence of Pearson or Couttas. I note, for example, that Vanderkemp adopted a similar approach on this issue. 553 No other witness said that the approach adopted by Pearson and Couttas was inappropriate. Other witnesses all qualified the phrase 'transfer of risk' by an adjective such as 'significant',

'sufficient' or ' material' .554

One of the many deficiencies in AASB 1023 is that it is silent as to the quantum of risk transfer. Practitioners in the field interpret the standard by applying their training and experience. I am satisfied that it was accepted practice for actuaries and auditors considering contracts of reinsurance for them to be accounted for in accordance with AASB 1023 only ifthey contained 'sufficient' transfer of risk. This test is admittedly uncertain but it required more than a minimal risk transfer.

Third, a question was raised as to interpretation of clause 7 .1.5 insofar as it purports to deal with the expensing of premiums for non-proportional reinsurance. In their submissions counsel for Andersen contended there is no ambiguity in clause 7 .1.5. The clause provides that for non-proportional reinsurance 'the expense would

normally be recognised over the period of the reinsurance arrangement'. The result is that premium is expensed over the period for which benefits may properly be regarded as remaining available under the arrangement. ' So long as there is life in the contract, the premium may be expensed over that life. ' 555

If this is correct it would seem to follow ' recognising' premium expense 'over the period of the reinsurance arrangement' would mean expensing that premium on a straight line basis. But clause 7 .1.5 does not say that. In my view the clause is ambiguous as to the manner of the recognition of the premium expenses. The reader

is driven back to the first sentence of clause 7.1.5 . It requires recognition 'in accordance with the pattern of reinsurance service'. What is not identified in the case of non-proportional reinsurance is the ' reinsurance service'. It could either be the availability of protection or the expected recoveries or both, depending on the circumstances.

The ambiguity is exemplified by the approach adopted by Vanderkemp when he first considered the method of expensing premiums under the GCRA and NI contracts. He prepared two spreadsheets, one which involved the expensing of premiums in the same proportion as recoveries and the other expensing on an equal

straight line basis. In my view each of these methods can be seen as recognising the premium 'over the period of the reinsurance arrangement'. I note that Scrivens shared this view as to the ambiguity of the standard on this point. 556 I consider this further in Section 16.3.5 in relation to the Swiss Recontract.

The failure ofHIH Insurance 189

14.4.10 The GCRA AXOL contract-risk transfer and accounting

Pearson concluded that, considered independently of the side letters and the six I May I998 contracts, the AXOL contract did not involve the assumption of sufficient insurance risk by GCRA to conclude that it was genuine reinsurance. 557 This is of significance because Andersen were presented only with the AXOL contract. Pearson's analysis involved removing the $68 million of recoveries that FAI was funding throughs its premiums held 'on deposit'. She concluded that, in exchange for its premium payments to GCRA of $9.25 million, FAI was entitled to a maximum possible recovery of $I9 million payable no earlier than

I January 2003. Taking into account the timing difference between the payment of the premiums by F AI and the receipt of the maximum possible recovery, Pearson calculated558 that the maximum value of the contract to F AI at its commencement date of26 June I998 was $5.6 million. 559

Andersen contended that this analysis ignored the 'commercial substance' of the contract. In Andersen's view treating the contract as merely involving a maximum recovery of $87 million ignored the different limits of indemnity provide for in sections I to 6. 560 I do not accept this. Pearson's analysis of the substance of the transaction is entirely consistent with its terms, particular the offset clauses which would allow the premiums held on deposit to be offset against the equivalent amount of recoveries. The premium payments were not referable to the individual sections of cover but were payable for Sections I to 5 combined.

Consistent with his emphasis on AASB I 00 I, Couttas accepted Pearson's analysis of the AXOL contract. The first $68 million of recoveries was to be funded by F AI through the premiums it was to hold. Accordingly, the substance of the transaction was that for the balance of the possible recoveries, namely $I9 million, F AI was paying $9.25 million. Couttas concluded that no recoveries should have been taken up under the GCRA contract until it was more probable than not that the recoveries would exceed the $68 million which was being funded by the premiums held on deposit by F AI (and su bsequently HIH). 561 In his view no recoveries should have been accounted for in the year ending 30 June I998 as the expected recoveries had not then exceeded $68 million. Only $9.25 million of reinsurance premium should have been expensed. The result was that F AI' s profit for the year ending 30 June I998 was overstated by $29.749 million. 562 I accept this approach and conclusion. It follows from my acceptance of Pearson's analysis.

Pearson also considered whether there was risk transfer under the GCRA contract on the assumption that the premium payments due under the six I May I998 contracts in reality were a further fee for that contract and no risk was transferred under those contracts. 563 This was by way of taking into account the side letter and six I May I998 contracts. She conclw;led that, based on those assumptions, there was no transfer of risk under the transaction between F AI and GCRA. This was because there was a total premium payable from F AI to GCRA (over and above the premium deposit of $68 million) of $21.75 million compared with a maximum recovery of $I9 million. 564 Pearson determined the value of the GCRA contract at

190 The impact of the FA! acquisition

the date of its commencement (26 June 1998) to be negative $6.9 million. 565 On this basis, Couttas considered that the transaction should have been accounted for as a deposit transaction by attributing a negative value on its inception of minus $6.893 million. The result was that FAI's profit before tax for the year 30 June 1998 was overstated by $35.6575 million. 566 I accept these conclusions.

14.4.11 The Nl contract-risk transfer and accounting

Pearson analysed the NI contract against two scenarios. The first involved an assumption that the earthquake trigger did not occur, but otherwise there was a best possible outcome for F AI, namely the maximum recovery (of $50 million) available at the earliest date, namely 1 July 2000. Under the second scenario the earthquake trigger occurred with F AI making a claim for the maximum $50 million payable on

1 July 2000 and for the remaining $30 million payable on I July 2001. 567

On the first scenario Pearson concluded that the maximum value of the NI contract to F AI as at 4 July 1997 was negative $6.7 million. 568 Pearson also noted that at each balance date the commutation value of the contract (that is, the balance of the experience account) would exceed the value of the contract to F AI. On this basis it was likely that the contract would be commuted some time prior to 6 July 2001. 569

On the second scenario, Pearson determined F AI would have to make premium payments of $4.4 million-that is, the premiums payable which were not deposited into the experience account570 - for a maximum possible recovery of around $26 million payable no earlier than 1 July 2001. This would be contingent on an earthquake of sufficient loss magnitude occurring in Australia during the five years ending in 2001-02 and on there having been an adverse experience in excess of $50 million and up to $80 million under the terms of the contract. 571 She agreed with certain KPMG actuarial estimates which estimated the probability of an earthquake event of that magnitude occurring

2.2 per cent and 4.1 per cent. Pearson concluded that the probability of NI being called upon to make any payment under the contract was therefore less than 4.1 per cent. 572 Pearson concluded that NI had not 'assumed significant insurance risk' under the contract because 'the probability of a significant variation in the amount and timing of payments by the reinsurer is remote (less than 4.1 per cent)' and 'there is only a very small probability (less than 4.1 per cent probability)' that NI would realise any loss under the contract'. 573 If the period for the earthquake

event was four years rather than five years, the probability would reduce to 3.3 per cent. 574

On the basis of Pearson's analysis, Couttas concluded that the NI contract should not have been accounted for in accordance with the requirements of AASB 1023. 575 Even if it were, unless an earthquake event sufficient to trigger the extra cover had occurred, the contract could only be analysed as one involving the payment of

$55 million premium for $50 million cover. Thus, for year ending 30 June 1998 Couttas would have reversed the recorded profit of $29.078 million. If it were accounted for as a deposit arrangement Couttas would have booked a small

The failure ofHIH Insurance 191

investment with the overall result that F AI' s results were overstated by $31.078 million. 576 Alternatively, if it were to be accounted for as reinsurance, PAl's profits for the year ending 30 June 1998 were overstated by

$34.078 million. 577 I accept both Pearson's and Couttas's evidence on the NI contract and its proper accounting treatment.

14.4.12 Was Andersen misled in relation to the GCRA contract?

It was contended by both counsel assisting and counsel for Andersen that Andersen were misled in relation to both the NI and GCRA contracts by a number of statements and omissions by FAI management, assisted by representatives from GCRA and Guy Carpenter.

I have already concluded that the AXOL contract was designed to facilitate the misleading of Andersen by the withholding of the various matters which affected or removed risk transfer, namely the six 1 May 1998 contracts and the 1 May 1998 side letter. That which was contemplated was put into practice. Andersen was never advised of those matters, the connection between the six 1 May 1998 contracts and the second AXOL contract or the 26 June 1998 side letter. There was ample opportunity for such disclosure to occur throughout July and August 1998 including:

• the meeting between Wilkie, Peiris and Scrivens on 21 August 1998

• the meeting between Mainprize, Peiris and Scrivens on 21 August 1998

• the meeting involving Scrivens, Wilkie, Mainprize, Barnum, Smith and others on 26 August 1998 and

• the meeting of the audit committee on 3 September 1998 at which Mainprize was present.

I have not made a finding against Smith or Barnum in relation to their attendance at the meeting on 26 August 1998. In the case of Mainprize and Wilkie I am satisfied that their failure to inform Andersen of the above matters was deliberate. They were both qualified accountants and I consider they would have been aware of the relevance of those documents and matters to the performance of Scrivens's duty as auditor.

I am satisfied that disclosure of those matters to Scrivens would ultimately have led him to conclude that there was no transfer of risk under the GCRA AXOL contract. 578 The side letter of 1 May 1998 is an extraordinary document and it is difficult to see how knowledge of it could have done other than cause Scrivens to reach a different conclusion. It is likely that in the event full disclosure was made, Scrivens would have insisted that the transa..,tion be accounted for as a deposit arrangement in the manner suggested by Couttas. 579 If that occurred the reported operatingprofit before tax would have been $35.675 million lower. 580

192 Th e impact of the FAI acquisition

14.4.13 Was Andersen misled in relation to the Nl contract?

I have already concluded that the cover note prepared by Tuckfield falsely conveyed that it was created in July 1997. I have also referred above to a number of statements concerning the NI contract that did not reflect the true position and that were made by F AI management either in exchanges between themselves in preparation for meetings with Andersen or direct to Andersen. They included statements that the NI contract:

• was prospectively based for five years

• had been entered into to protect F AI' s bottom line from erratic claims development and to protect the underwriting results as a whole

• as accounted for, transferred underwriting risk and timing risk.

I mention also the statement in the position paper that Andersen's objection to the accounting treatment of the contract sought to ' ... reinterpret the intention of the contracts by reference to the subsequent experience of the group'. 581

The true position was that the contract was entered into to obtain a specific accounting result, at a time when there was no uncertainty as to the claims for the first year. Further, that the large recovery that occurred in the first year was 'planned', in the sense that the contract was designed to obtain it.

Were all or any of these statements made by the authors in a deliberate attempt to give a false impression? I have already dealt with Tuckfield's conduct in preparing the cover note. I do not find that he participated in the misleading of Andersen in July and August 1998. I think Wilkie's statement to Scrivens on 21 August 1998 concerning the reason for the entering into of the NI contract was false. The contracts had little to do with protecting against any future volatility. Otherwise there is an unsatisfactory disconnection in the evidence between those who prepared the documentation for submission to the auditors and those who were aware of the true facts. For example, Wilkie accepted that the position paper was 'incomplete' in that it made no reference to why he was asked by Adler to obtain the NI contract.

But there is no evidence he saw it at the time. 582 Peiris was the author of many of the statements, yet the evidence that he knew they were false is not sufficiently compelling for me to make any adverse findings against him.

The next question is whether the statements had any effect on Scrivens's conclusion that the NI contract was genuine reinsurance and could be accounted for as contended by FAI management. For a variety of reasons and despite what I have found concerning what Tuckfield knew when he prepared the cover note, Scrivens

in fact believed that 'agreement in principle' had been reached around the beginning of the 1998 fiscal year. 583 He thought, however, that the cover note itself came into exist sometime between January and June 1998. 584 With both Scrivens and Vanderkemp the most significant factor which led to their acceptance of the NI contract as genuine reinsurance was the belief that each of them held that the

The failure of HIH Insurance 193

total recoveries available under the contract, irrespective of any earthquake, were $80 million and not $50 million. 585

In the result, I do not think Scrivens's analysis of the NI contract was affected by his understanding of when 'agreement in principle' was reached or any of the other statements to which I have referred. In his evidence Scrivens never asserted that he was so affected. 58 6

14.4.14 · Andersen's audit approach

In reviewing the course of Andersen's audit ofFAl's proposed accounting treatment for the GCRA and NI contracts, two points should be borne in mind. First, based on the findings I have made, F AI management, senior executives from GCRA and a senior executive from Guy Carpenter engaged, to various degrees, in conduct which misled Andersen as to either the true circumstances surrounding the negotiation of or the terms of the reinsurance contracts or both. It was not suggested to me that any auditor acting reasonably should have suspected that conduct of this sort was being carried out. To the contrary, Andersen was placed in an invidious position by the conduct which was directed at misleading them.

Second, there was a number of 'risk factors ' which warranted additional scrutiny being given to the reinsurance contracts and their impact on the accounts for the year end 30 June 1998. Andersen was aware that on previous occasions FAI had engaged in one-off transactions which had the effect of ameliorating a substantial loss or converting a substantial loss to a small profit. 587 Both reinsurance contracts answered that description. Andersen knew that at least the GCRA AXOL contract also had been entered into almost immediately prior to the relevant report date. 588 The entire 'risk' under each reinsurance contract was being assumed by one reinsurer. 589 This raised the possibility that they were special deals designed specifically to improve the F AI balance sheet. A further risk factor was the right to commute conferred on F AI by the NI contract. This raised the possibility that the contract would be used in its early years merely to gain a beneficial accounting treatment and then, when it had served that purpose, commuted.590 Scrivens' s understanding of the timing of the preparation of the cover note meant that it had been 'backdated' and this was also a matter he accepted set 'alarm bells ringing' .591 Scrivens agreed that the two contracts were the most significant issue that he had to deal with when undertaking the 1998 audit of FAr.s 92

14.4.15 Risk transfer

Scrivens had two tasks to undertake in relation to each contract. He had to determine whether they contained sufficient of risk to constitute reinsurance and he had to consider whether the proposed accounting treatment was appropriate.

There is no basis for any criticism of Scrivens and Vanderkemp in the way they assessed risk transfer in relation to the GCRA AXOL contract.

194 The impact of the FAJ acquisition

With the NI contract, Vanderkemp's internal rate of return analysis quantified the net present value of the cash flows under the contract as 15 per cent of the premiums which, in Scrivens's view, exceeded any achievable interest rate and in turn lead him to conclude that the contract involved a material transfer of risk. This figure was arrived at because Scrivens and Vanderkemp interpreted the contract as though

$80 million was recoverable irrespective of an earthquake but that recoveries in any one underwriting year could only exceed $50 million in the event the earthquake event did occur. 593

I do not believe this is a proper interpretation of the NI contract.

The relevant parts of the NI cover note provided as follows:

CLASS:

LIMIT: ...

Subject to the Limit and Warranty, to indemnify the Reassured for all losses of whatsoever nature arising .. .

The Sum Reinsured hereunder shall be limited to a maximum further 10% of the Reinsured's Net Premium Income for the covered Underwriting Year of Account. Subject to the Warranty and to an overall maximum account recoverable hereunder of AUD50,000,000 per

underwriting year of account and an Aggregate of AUD80,000,000.

Under no circumstances shall the Reinsurer be liable for more than AUD50,000,000 for any one underwriting year of account or AUD80,000,000 in the aggregate by reason of entering into this reinsurance contract ...

WARRANTY Losses in excess of AUD50,000,000 under this Reinsurance are recoverable only in the event of Earthquake loss in Australia of a magnitude greater than AUD5,000,000 in a given Underwriting Year Account and subject to a limit of AUD15,000,000 for such Underwriting Year. 594 [emphasis added]

Scrivens and Vanderkemp interpreted the warranty clause as only limiting the amount recoverable in a given underwriting year of account as opposed to the amount recoverable under the contract in its totality. This requires the clause to be read as if it said that 'losses in excess of A$50 000 000 in a given underwriting year of account under this Reinsurance' are recoverable in the event of a relevant earthquake loss. However, the warranty does not say that and in fact expressly operates to limit recoveries 'under this Reinsurance'. I think this is a reference to the reinsurance contract in its totality.

If Scrivens's and Vanderkemp' s construction was correct there would be a direct inconsistency between the warranty clause and that part of the clause headed 'Limit' which provides that 'under no circumstances' shall the reinsurer be liable for more than $50 000 000 for any one underwriting year of account. This part of the clause

is not expressed to be subject to the warranty clause. This is reinforced by the preceding paragraph which subjects the meaning of the phrase 'Sum Reinsured' both to the 'Warranty' and to 'an overall maximum amount recoverable hereunder of A$50 000 000 per underwriting Year of Account'. I am satisfied that the proper

The failure of HJH insurance 195

way of reading the two clauses together is to read the warranty clause as operating to limit losses in excess of $50 000 000 in aggregate under the contract. I can see how, to a person without legal training, the interpretation which appealed to Scrivens and Vanderkemp might commend itself. But it cried out for legal advice or at least discussion with F AI management. They did not take either step.

14.4.16 Approval of accounting treatment

I have already indicated my acceptance of Couttas's opmwn as to the correct accounting treatment of the GCRA contract. I should note two matters of principle he raised concerning the audit approval ofF AI' s proposed accounting treatment.

First, Couttas stated that the booking of any amount of recoveries under the AXOL contract when they had not yet reached $68 million was a departure from ' well recognised non-controversial ... accounting principle' .595 This followed from his analysis of the substance of the transaction in that F AI was self funding the first $68 million in claims. I accept this conclusion. The contract provided for the withholding of premiums 'on deposit' by FAI with provision for offset against recoveries. To my mind that made it obvious that F AI was self funding at least the first $68 million in recoveries. Unless recoveries exceed $68 million no cash would flow back to F AI.

Second, Couttas also criticised Andersen ' s acceptance of an accounting treatment which involved the recognition of a profit for year ending 30 June 1998 greater than the maximum profit that could be achieved over the term of the contracts. This criticism was based on the fact that the payment of any recoveries under the contracts was predicated on F AI paying to the reinsurers all the premiums payable. This meant that the premiums should also have been recognised as at that date. 596 Couttas referred to Statement of Accounting Concepts 4597 which requires that a liability should be recognised in the statement of financial position when and only when it is probable the future sacrifice of economic benefits will be required and the amount of the liability can be measured reliably.598 The future premium obligations ofF AI satisfied both criteria. In its response Andersen contended that the existence of a premium overhang was not prohibited by AASB 1023 and that it was anyway ambiguous. That may be so but AASB 1023 operates against a background of fundamental accounting principles and known future losses not brought to account violate those principles. I accept Couttas' s criticism.

Couttas ' s view on the second point was initially subscribed to by Scrivens but he was ultimately persuaded to accept PAl's proposed accounting treatment. Scrivens says he was persuaded to the contrary by three matters: the statements made to him by Barnum and Tuckfield; his dealings with other colleagues within Andersen; and an alleged ' :::ushion' in the Queensland CTP reserves. I will consider each in turn .

I have set out above the essence of the statements that Scrivens asserts were made to him by Barnum and Tuckfield on 26 August 1998. Although I am not prepared to make any adverse findings against Barnum, Smith or Tuckfield in relation to those

196 Th e impact of the FA! acquisition

statements, it does not follow that I do not believe Scrivens. Rather, the evidence is not sufficiently certain to make it safe for me to reach conclusions adverse to Barnum, Smith or Tuckfield. Assuming Scrivens was told those matters, what was their significance? Scrivens said he placed little weight on anything Tuckfield said599 and those matters can be ignored. I accept, from Scrivens's perspective, that had he been told by Barnum or gained the impression from things that Barnum said, that GCRA would in effect adopt mirror accounting and recognise' in the current year a loss that exceeded the maximum possible loss under the contract, then it would be a significant matter for Scrivens to consider. As to what was 'normal accounting treatment in the industry', it would have been appropriate to make

inquiries of other reinsurers or general insurance market participants rather than just relying on the statement of a counter party to one of the agreements.

Scrivens spoke to a number of other people within Andersen on this topic. He spoke to Wildig, one of Andersen's London-based insurance experts. Wildig arranged for Halley, a UK partner of Andersen, to forward to Scrivens a presentation on reinsurance which contained a detailed discussion of the appropriate accounting treatment of reinsurance contracts. 600 One of the examples in the PowerPoint presentation appeared to provide some support for Scrivens's conclusion. It set out a hypothetical example of the balance sheet of a reinsured and a reinsurer over a three year period. The contract involved a reinsurance premium of 300 units for potential

recoveries of 300 and then set out an example where the full 300 units of claims on the reinsurance policy were raised in year one. In the first year, the reinsured recognised 300 units in recoveries and expensed only 100 units in premium with the reinsurer adopting a mirror accounting treatment. 601

Scrivens said that after he reviewed that presentation he telephoned Halley who confirmed that all the recoveries could be taken up in the first year and there was no need to expense the remaining premium. The difficulty with the example is that it contemplated cash changing hands when premiums and recoveries were booked. 602 Unlike the AXOL and NI contracts, the payment of recoveries was not deferred and

not contingent upon future premiums being paid. The example involved a transfer of timing risk. It was a critical difference, although Halley's advice seems to me to be inconsistent with Australian accounting standards.

Scrivens also spoke to three Andersen Australia partners, Davies, Gooley and Gardener. In relation to Davies, Scrivens stated that he placed no weight on what Davies conveyed to him in reaching his conclusion as to whether to accept or reject F AI' s proposed accounting treatment. 603 There were some differences in the evidence given by Scrivens and Gardener as to what they discussed. I do not

propose to resolve those differences. The difficulty with Scrivens's discussions with Gooley and Gardener is that he only gave each of them a brief precis of the contracts and the proposed accounting treatment. He overlooked the crucial significance of the absence of timing risk upon the accounting treatment for each contract. By virtue of the deferred recoveries and offset clauses, payment of

recoveries was always contingent upon the payment of the entire premium over the life of the contracts.

The failure of HIH Insurance 197

The other reason given by Scrivens for accepting the proposed accounting treatment was his view that F AI had 'considerable potential cushion in its claims reserves of Queensland CTP'. His memorandum of 27 August 1998604 noted that while F AI had adopted a lower risk estimate for this reserve 'the actuaries report on the reserve points out that the company has assumed superimposed inflation in its calculation which is not evidenced by trends in the market' .605 This is not what the actuary 's

report606 stated. The authors pointed to a possible 'cushion' of $35 million in relation to the Queensland CTP reserves. But it was in a context where they had concerns about other portfolios. It only enabled the actuaries to conclude that their aggregate central estimate was reasonable. It did not provide any basis for the drawing of any 'comfort' in relation to any other figure in FAI's accounts outside the portfolios being valued in the report.

14.4.17 Failure to discount recoveries

Undiscounted figures for the recoveries under each of the GCRA and NI contracts were booked in F AI' s accounts notwithstanding that recoveries would not be paid until several years later. 607 Andersen's justification for this was Trahair's contention that, as the claims that gave rise to the recoveries were not discounted, it was

inappropriate to apply a discount to the reinsurance recoveries. 608 AASB 1 023 provides that gross claims incurred by the insurer should be discounted over the expected period that those claims will be paid out so as to arrive at the present value of expected future payments.609 It also provides that non-current reinsurance recoveries should be discounted back from the time of expected receipt. 610 I am satisfied that the booking of recoveries based on undiscounted claims was erroneous. There is no guarantee that the discount period for the underlying claims against the reinsured will coincide with the discount period for the recoveries from the reinsurer.

14.4.18 Dealings with the audit committee and the note to the accounts

I am concerned at the lack of information given to the F AI audit committee to allow them to assess the appropriateness of the proposed accounting treatment for the reinsurance contracts. They were not told the substance of the debate that had occurred between Andersen and management in the course of Scrivens coming to

his conclusions. Nor were they told about the internal debate within Andersen or Andersen' s view that there was some ambiguity or uncertainty as to the interpretation and application of the relevant accounting standard. 611 They were not presented with any alternative accounting treatment which might represent a more cautious view of the way in which these contracts should be accounted for. The risk factors listed above were not brought to the committee's attention.

The only issue that was highlighted to the committee was the 'guaranteed accounting loss' for future years. On that issue the non-executive directors queried its treatment but were effectively told by Andersen that it was in accordance with industry practice. 612 The audit committee was presented with a common view from

198 Th e impac/ of/he FAI acquisilion

auditors and management in relation to what was clearly a significant and contentious issue for the company's accounts.

It is not the role of auditors to set out for the audit committee the entire history of the audit and a point by point discussion of the way each transaction was dealt with and the way each difference of opinion with management was resolved. However, this matter was very unusual and highly significant. Scrivens acknowledged that it was the most significant issue he dealt with in the course of the 1998 audit. 613 The consequences of the adoption of the accounting treatment were significant both for F AI and for those who relied on its accounts. The company would be commencing the next financial year with a known future loss which would be realised in either that year or subsequent years. The subject matter of the transactions (reinsurance) was highly technical. The audit committee and the board had to reach a decision whether the accounting treatment proposed by management was appropriate. In my view it was a matter on which they required, and could reasonably expect, detailed

independent and reliable professional advice. I consider that these matters strongly suggested that Andersen should have explained the nuances and the ambiguities of the application of the accounting standard to the transactions, including any doubts they had about the proposed accounting treatment.

I also note that the audit committee was advised that there was a $23.65 million future guaranteed accounting loss in relation to the contracts. 614 This figure was derived from what I consider to have been an incorrect construction of the NI contract. In my view, the future guaranteed accounting loss was, subject to a

catastrophic earthquake occurring, $53 .65 million not $23.65 million. For a company whose shareholders' funds (excluding minorities) were $224 million as at 30 June 1998615 , it was a material figure.

The minutes of the audit committee meeting record that the 'accounting policies note should be expanded to highlight the practice associated 'Nith [the reinsurance] contracts'. The evidence suggests that this was a reference to the long-term nature of the contracts. I find this puzzling. The 'practice' highlighted to the committee was a mismatch between premium payments and reinsurance recoveries which would

inevitably result in a guaranteed accounting loss in future years. I have already set out the form of the note finally agreed upon and approved by Scrivens. The note was unhelpful. It did not inform the reader that a particular interpretation of what was seen to be an ambiguous or uncertain accounting standard had been adopted. It did not make clear that premium was being expensed according to 'the availability of protection under the reinsurance contract' 616 and not according to the proportion of available recoveries booked. The one matter that should have been highlighted in

the notes was the fact that F AI' s proposed accounting treatment would ensure that there would be a future loss of at least $23 .65 million (and most likely $53.65 million) in subsequent years. The note is silent on this important piece of information.

The failure ofHIH Insurance 199

14.4.19 The non-executive members of the au dit committee

The only criticism of the non-executive directors ofF AI that was submitted to me concerned the terms of the disclosure in the notes accompanying the accounts. Andersen's presentation to the audit committee highlighted that there would be 'guaranteed accounting loss' in future years. Counsel assisting submitted that the non-executive directors should have insisted upon the inclusion in the accounts of a note explaining this aspect of the accounting treatment for the contracts.

Hill and Harris sought to refute this submission. They submitted that the reference in the note to the accounts to the 'multi-year' contracts was included because that was the explanation for the accounting treatment that was given to them after they queried it. They also pointed to the fact that Scrivens was prepared to sign off on the accounts without there being any note, and that as non-executive directors they were entitled to rely on management and auditors in relation to its final wording. They further asserted that it could not be stated that the future known shortfall was to their knowledge material to the accounts and that overall counsel assisting's submission failed to have regard to the proper relationship between non-executive directors and the auditor.

I can see some merit in a number of these points. But I also accept the criticism made. In the ordinary course non-executive directors can rely on management and auditors in relation to both the accounting treatment adopted and the preparation of the accounts, including notes to the accounts. However the guaranteed shortfall was specifically brought to their attention. To my mind it was a surprising accounting result. It was obviously material to F AI, whose profit position was delicately poised. In those circumstances Hill and Harris should have scrutinised the note to ensure that the guaranteed future loss was disclosed to the market.

Having said that, no question of a possible contravention of the law arises in relation to Hill, Harris or any other non-executive director of F AI.

14.5 The takeover of FAI

HIH had an active interest in the acquisition of a strategic shareholding in F AI or a possible takeover ofFAl from at ·least early 1993 . From 1995 HIH and its advisers (in particular Colin Richardson of Hambros) kept FAI and the possibility. of an acquisition under regular review. 61 7 HIH's regular review of FAI's financial and business information was undertaken principally from publicly available information such as FAI's published accounts, Australian Stock Exchange announcements, market analyses and the like. Hambros provided advice to HIH concerning FAI's position and perfopnance regularly pursuant to its advisory retainer. Fodera understood Hambros' role in regard as including 'evaluative input into the desirability ofthe [takeover] transaction'.618

HIH' s interest in F AI arose from its perceived need to broaden the base of its short­ tail insurance lines of business so as to balance its existing and sub stantial long-tail

200 The impact of th e FA ! acquisition

lines. Cassidy also believed that the acquisition of F AI would provide HIH with a well-known general insurance brand name and access to the direct (non­ intermediated) general insurance market. Cassidy considered that the acquisition of FAI would also assist HIH to broaden its distribution channels. 619 HIH directors identified other benefits which were expected to flow to HIH from the takeover of FAI. These included the rationalisation of HIH's resources in information technology, the removal of a competitor, the raising of HIH's public profile and size, economies of scale and scope and synergies which were expected to derive from a combination ofHIH's and PAl's general insurance operations.620

14.5.1 The need for due diligence

On a number of occasions from 1995 onwards, Richardson reminded Williams and others at HIH of the need to undertake a due diligence investigation ofF AI before making any acquisition. For example, immediately following PAl's announcement of its loss of approximately $90 million for the year ending 30 June 1995, Hambros sent Williams a document described as revised 'First Stage Analysis on Project Vitamin' .621 It included a disclaimer to the effect that the information in the document concerning F AI had been obtained entirely from publicly available

sources. Hambros said that it had not undertaken any independent audit or confirmation of that information and that accordingly the document did not represent the totality of analysis and due diligence required to be undertaken by HIH before making an investment in FAI. 622

Like many of the analyses ofF AI which were undertaken by Ham bros and by others during the mid-1990s, Ham bros' 1995 'first stage analysis' articulated many reasons why a takeover of F AI should be approached with great caution. The document noted that F AI had recently reported a $90 million loss, being the fourth annual los s

reported by FAI in the preceding five years . In 1995 PAl's net tangible asset backing per share had declined from $0.85 to $0.52 and its share price of just over $0 .60 represented a multiple of only 1.2 times its net tangible asset backing per share. Immediately prior to the release of its 1995 results, shares in F AI were

trading at a discount to their reported net tangible asset backing.

Ham bros' review of the asset values recorded in F AI' s balance sheet concluded that for a number of PAl's major non-core assets 'realisable value may be below book value'. This was said to be 'especially relevant to F AI' s property assets' including the substantial St Moritz Hotel. Ham bros also noted that F AI' s solvency ratio was

low and '[F AI] will face problems if losses continue to reduce solvency further' . 623

Ham bros described F AI' s liquidity position as 'tight' and observed that:

Any offer priced above $0.50 is earnings dilutive ... although unbundled the dilution is not material until a 60 cents offer price. 624

A number of comments in Hambros' analysis drew attention to the likelihood that F AI would need to raise additional capital in the near future in order to remedy its pressing solvency issues. 625 Ham bros advised in the document that the adequacy of

The failure of HIH Insurance 201

FAI's claims provisions should be confirmed, although the assumption appears to have been made in the analysis that F AI' s provisioning was adequate. 626

Based on this and other information, Fodera took the view that a due diligence investigation of F AI should be undertaken before any acquisition could proceed. Fodera gave evidence that he conveyed this view at various times to Williams, Cassidy and Richardson and that they all agreed with him. In Fodera's view adequate due diligence on F AI would entail access at least to F AI' s auditors ' (Andersen) working papers, actuarial reports and investment valuations. 627

Richardson later drafted letters for Williams which were intended to be sent to Adler requesting the opportunity for HIH to undertake proper due diligence on F AI. 62 8 As Richardson explained to Williams, it was necessary for Adler to appreciate that HIH could not make any definitive proposal for the acquisition of F AI without the benefit of due diligence. Furthermore, if HIH was denied relevant due diligence information it would be 'forced to make conservative assumptions which will reduce value'. 629

14.5.2 Developments in 1996

Williams and Adler met in January 1996 to discuss a takeover proposal. Adler gave Williams reassurances that F AI' s claims reserves had been independently actuarially assessed by Coopers & Lybrand and Trowbridge Consulting. Williams gave evidence that he was relaxed about the adequacy ofF AI' s claims reserves when he learnt that that they had been assessed by these two independent actuarial firms. 630 However Williams was unable to recall following up with Adler to confirm whether those reviews had been concluded or their outcomes.631

Prior to the takeover the focus of HIH's attention appears to have been upon the value of F AI' s assets, not the value of its liabilities. This is apparent from the information request which Richardson drafted for Williams in February 1996. Richardson noted that the request was 'drafted with reference to Ray's instructions and stays away from "off-limit" matters such as reserves' .632 Fodera's evidence was that he did not regard FAI's reserves as being 'off-limits' for the purposes of due diligence. He maintained that actuarial reports on F AI should be obtained by HIH as part of a due diligence investigation prior to a takeover ofF AI. 633

Following the release of F AI' s half-year results to 31 December 1995 in February 1996, Richardson sent Williams a draft Jetter intended to be sent by Williams to Adler. 634 As Richardson said in his covering fax: 'the essence .. . is that without better information [HIH] cannot progress any deal on F AI'. The letter which Richardson drafted for Williams to send to Adler said that:

202

... a favourable reaction by the investment market to any transaction is vital to its success, not least as a listed instrument will form part of any consideration. In addition, [HIH's] Directors will need comfort as to the appropriateness of our due diligence enquiries on F AI before they will support any transaction and my shareholders would expect no less

The impact of the FA! acquisition

We therefore need access to more detailed information before we can sensibly move forward .. . unless we can resolve this issue of information transfer we cannot progress any further. 635

At the HIH board meeting on 22 March 1996 Richardson described due diligence as necessary before any acquisition ofF AI or a substantial holding in F AI. 636 Hambros continued to analyse F AI and its financial performance and position for HIH later in 1996. 637

14.5.3 Events in 1997

HIH 's interest in FAI continued during 1997. 638 In February 1997 Hambros gave a presentation to Winterthur Swiss Insurance Company on development opportunities for HIH in the Asian region, including the possible acquisition of one or both ofF AI and GI0.639 In March 1997 Hambros sent HIH a discussion paper on these possibilities.640 Wein (then HIH Asia chief executive) wrote to Schurpf recommending that Winterthur be proactive in taking advantage of rationalisation

opportunities in the Australian general insurance sector along the lines recommended by Hambros.641 Strategy documents were then prepared canvassing a possible takeover of FAI.642 In April1997 Hambros' analysis of the possible acquisition ofF AI observed that despite its recent reported losses:

Market commentary appears to have accepted that [F AI] is now on the recovery path and that future profits could be generated if the non insurance assets were disposed of and if an effective management team was to operate the business. [HIH] has the expertise to be able to 'tum around' this underperforming business. 643

Wein expressed similar sentiments in a document entitled 'Project Vitamin Proposed Acquisition ofFAl Insurance by HIH-Winterthur - Winterthur' which he sent to Heri and Schurpf in July 1997. Wein noted a number of PAl's weaknesses but concluded that FAI was a 'good catch' and ' an attractive buy despite some complexities'.644 Wein's paper also assumed that write-offs and additional

provisions of $94 million would be required following the takeover. 645

Wein estimated that annual synergy benefits valued at $20 million would flow from a takeover ofF AI by HIH. Despite the negatives about F AI expressed in his paper Wein recommended that a project team be assembled within HIH to begin negotiations with a view to HIH taking over F AI at a price up to $0.95 per share.

This valued F AI at $340 million. 646

HIH's active consideration of the takeover of FAI was then interrupted by the merger of Winterthur and Credit Suisse. This merger and the subsequent sell-down by public offering of Winterthur's controlling interest in HIH occupied a significant amount of HIH's management time and resources from August 1997 until the

middle of 1998 .

The failure ofHIH Insurance 203

Despite these distractions, Hambros continued to assess and to advise HIH on the F AI takeover proposal during late 1997. 647 In a letter dated 8 October 1997, Richardson commented to Williams :

[F AI] is not an asset which lends itself easily to a hostile offer due to the shareholding of Rodney Adler. Therefore we need to obtain the agreement of Rodney prior to finalising any transaction. If agreement could be reached, further work would be required prior to entertaining any binding commitments. Having said that, we need to be confident that if an attractive transaction can be negotiated HIH is prepared, in principle, to acquire [FA!].

Given this I would urge that discussions occur to ascertain the level of any interest and to obtain more reliable information. In this way greater certainty would be provided for HIH and Winterthur without the need for further analysis which may prove not only be unnecessary but also to have been founded on incorrect assumptions. 64

The potential acquisition of F AI was discussed at the HIH board meeting of 27 February 1998. 649 It did not reappear on the HIH's board's agenda until 22 September 1998. 650

Several of the HIH directors gave evidence that the possibility of HIH taking over F AI was discussed formally at board meetings and informally between the HIH directors for several years prior to 1998. In fact Williams had first commenced his discussions with Adler about a possible HIH-F AI takeover in 1989. 651 As Stitt put it:

the idea of taking over another Australian or overseas insurance company was considered in detail by the Board over a number of years, and was always subject to independent expert advice and analysis provided by Hambros. 65'2

Stitt understood that there was a wide range of possible benefits for HIH in a takeover ofF AI and that these were presented by Richardson to the board of HIH during 1997.653 Cohen gave evidence to similar effect. 654 So too did Abbott. 655

14.5.4 Events in September 1998

In February 1998 Adler told Williams that he did not want to dispose of his F AI shareholding then. 656 Nothing further happened between F AI and HIH until September 1998. 657 Events thereafter moved very quickly. On 9 September 1998 Adler couriered FAI's results for the year ending 30 June 1998 (which had been released to the Australian Stock Exchange that day) to Williams. In his covering letter to Williams Adler wrote: 'I look forward to our next discussion'. 658

Adler made contact with Williams shortly before the release ofF AI' s 30 June 1998 results in September 1998. Negotiations then began in earnest between them following the release of F AI' s results. 659 Williams asked Fodera to analyse F AI' s 1998 results with particular emphasis on the value of the company's assets and its net tangible asset backing per share.660

204 Th e impact of the FA! acquisition

Fodera's assessment was that FAI's 1998 results were quite poor and that any takeover ofF AI would be a doubtful and unpredictable enterprise.66 1 Fodera gave evidence that the main focus of HIH' s investigation of F AI at this stage was the value of its assets:

because we didn ' t have any other information on the reserving, except the knowledge that (PWC) had signed those off as at 30 June 1998 .662

Like Williams, Fodera relied upon his understanding that F AI's claims reserves were signed off by PWC.663 Fodera's evidence was that he read through FAI's 1998 accounts, including notes in respect of reinsurance, but those notes did not raise any particular issues in his mind. Fodera assumed that F AI accounted for its reinsurance

in accordance with the reinsurance services rendered. He denied that he knew anything about the GCR or National Indemnity premium overhangs at any time prior to or during the takeover. 664

14.5.5 Fodera's discussion with Davies

Fodera sought to discuss F AI' s 1998 accounts with Davies of Andersen. Davies had been F AI' s auditor for the previous eight financial years and had recently ceased to be F AI' s audit engagement partner. Davies participated in several aspects of the 1998 F AI audit as the relationship partner. 665 Fodera provided his financial analysis of F AI666 to Williams and Richardson and they discussed it together. Apart from

Fodera' s analysis and the reading of some stockbrokers' reports, press releases and similar public information about F AI, no other substantive financial information or analysis about FAI was undertaken by HIH prior to the takeover so far as Fodera was aware. 667

Fodera gave evidence that when he called Davies to see if Davies could provide any information that might help HIH in its investigation ofF AI, Davies responded that he was in a difficult position because Andersen 'represented both parties'.668 Davies told Fodera that FAI had been recently audited by Andersen and that ' a clean audit report would be issued'. Fodera's evidence was that in previous discussions Davies had commented that F AI:

had a very strong internal actuarial group. He had made the comment that the internal actuarial group had performed a detailed review of all reserves and this review was supplemented by a PWC sign off every six months. 669

Fodera passed this information on to Williams and Richardson from which he said he took a great deal of comfort.670

14.5.6 Williams on Davies

Williams also took comfort from the fact that Andersen were the auditors ofF AI and PWC were its external actuaries. 67 1 Williams had requested that Fodera to speak with Davies because:

Th e failure of HIH Insurance 205

I thought that certainly they could have an off-the-record discussion on the basis that we were looking to acquire and therefore indicate to us if there were any concerns at all through our making that- making such a bid ... [Fodera and Davies] were colleagues going back many, many years and it was for that reason that I thought it would be possible for [Davies] to be able to comment to [Fodera] , because of the long-term relationship .. . 672

Williams took the view that the relationship between Fodera and Davies was such that any information of significance could be passed on by Davies to Fodera in such a way as not to damage the relationship between FAI and Andersen. 673 Since no information was forthcoming from Davies, Williams concluded that HIH could rely upon FAI's 1998 audited accounts and that there were no matters which should concern HIH in making a takeover bid for FAI. 674 This was clearly a false basis for reliance upon FAI' s 1998 accounts. Williams ought to have satisfied himself properly whether Davies' silence meant what Williams assumed it meant. What Davies actually confirmed in the context of the HIH due diligence is considered further below. Davies himself gave evidence that any HIH director who took comfort from the fact that he served on HIH's due diligence committee in respect of the takeover was foolish to have done so. 675 His stated reason was that he had no involvement in FAI's 1998 audit. 676

14.5.7 Adler and Williams commence negotiations

Shortly after he received FAI's 1998 results Williams recommenced his takeover discussions with Adler. Those discussions covered topics such as the types of insurance business which F AI was underwriting and its key assets such as Oceanic Coal. Adler told Williams that he wanted a minimum cash offer price from HIH of $1.00 per F AI share. Williams responded that such a price was unrealistic having regard to the asset backing ofF AI' s shares and their current market price. 677 F AI' s shares had traded in the range of $0.40 to $0.50 throughout 1998 prior to the takeover.

Williams and Adler met a on bout 15 September 1998 to pursue their discussions. They discussed a mutually acceptable offer price. Williams asked Adler about the possibility of HIH obtaining access to Andersen's 1998 FAI audit working papers so that HIH could undertake a more detailed due diligence investigation of F AI. Adler declined to allow HIH to perform any due diligence on F AI and also declined to allow HIH to have any access to Andersen's audit working papers.678

Williams and Adler met on four or five occasions in September 1998 prior to the HIH board meeting on 22 September 1998 at which it was resolved to proceed with a bid for F AI. Williams gave evidence that on more than one occasion at these meetings he raised with Adler the question of HIH's access to FAI information to allow HIH to undertake a due diligence investigation ofF AI. On each such occasion Adler refused the request for due diligence. 679 Williams's evidence was that he also raised at these meetings with Adler the adequacy of FAI's claims reserves. Adler reassured him that they had been signed off by PWC.680

206 Th e impact of the FAJ acquisition

In his evidence, Williams described a process of bargaining which took place between himself and Adler concerning an offer price for F AI shares which would be acceptable to Adler. Williams said that he began with an offer of about $0.60 per share. After further bargaining, an offer price of $0.75 per share was agreed.

68 1 It appears that once that price was agreed between Williams and Adler it was not revisited by anyone at HIH, including any of the board members who met on the evening of 22 September 1998.682

14.5.8 HIH board meeting of 22 September 1998

The board meeting in question was called urgently and at short notice. Notices were given to some directors only earlier that day and so several HIH directors were not present and did not submit their apologies. 683 Cassidy, Sturesteps, Payne, Weinand Heri were overseas. None of them received any notice of the meeting. In fact only three HIH directors were present at the meeting in Sydney, namely Fodera, Head and Williams. Four other HIH directors participated by video or telephone, namely Abbott, Gorrie, Cohen and Stitt. 684

Because the notice of the meeting had been so short, at its commencement Cohen asked whether there was any legal impediment to the meeting proceeding if some of the HIH directors had not received any notice. Cohen also raised the question whether it was appropriate to deal with the matters on the agenda in the absence of those directors. Cohen was advised that there was no legal reason why the board could not proceed to consider the proposed F AI acquisition. None of the directors considered that it was inappropriate to approve the F AI acquisition in the absence of the remaining HIH directors. 685

The meeting was informed that there were a number of other interested parties who might acquire a shareholding or make a bid for F AI and that the matter was urgent. 686 Williams gave evidence that he told those present at the meeting that no due diligence had been or could be undertaken on F AI because Adler refused to allow it. He advised that Fodera had had a discussion with Davies. In Williams' s

view 'the board generally had fairly good background knowledge ofFAl' because it had 'been on the agenda for a long time ' . Williams's evidence was that the board meeting discussed the proposed acquisition ofF AI in 'great detail'. 687

Williams also said that he told those directors present on 22 September 1998 that the synergy benefits which HIH would achieve from the takeover would be within the range of $40 to $60 million per annum. 688 Williams based that estimate on his 'gut fee1'. 689 No written analysis or computation of any such synergy benefits was ever

produced by HIH, before or after the 22 September board meeting. Richardson, like W ein, had previously estimated synergy benefits of $20 million annually. The board resolved that Williams should negotiate a 'pre-bid undertaking' from Adler for acceptance of HIH's takeover offer in respect of at least 10 per cent of FAI's issued capital before the takeover commenced. Williams did not obtain the undertaking

before buying Adler's parcel and announcing the full takeover bid on 23 September 1998. 690

Th e failure of HIH In surance 207

14.5.9 Richardson

Richardson's evidence was that he was contacted by Williams possibly in the week before the HIH board meeting of 22 September. Williams raised with him again the possibility of HIH acquiring F AI and mentioned that he had had a number of discussions with Adler. Williams told Richardson that he had discussed with Adler an offer price of $1.00 per F AI share. Richardson said that both he and Fodera told Williams that $1.00 per share was not justified and that he should go back to Adler and renegotiate the price. Williams told Richardson that there would be no due diligence undertaken on FAI. Richardson's impression was that 'Williams had made up his mind to proceed'.691

14.5.10 Hambros' financial analysis

Williams asked Hambros to update its previous financial model in connection with the takeover. Ham bros qid this by reference to F AI' s published accounts for 30 June 1998 and some Macquarie Equities research analysis on FAI. 692 Hambros' financial model for Project Vitamin dated 22 September 1998 693 was the only analytical document presented at the HIH board meeting that evening. In fact only the HIH directors present in Sydney had access to that document at the meeting. Ham bros' financial model assumed that approximately $60 million of provisions would be established immediately following the acquisition in respect ofF AI asset write-downs and reserve strengthening. Hambros also estimated synergy benefits of $25 million per annum pre-tax. An offer at $0.75 per share was assessed by Hambros to be marginally dilutive of HIH's net tangible assets by approximately 5 per cent. 694

14.5.11 Cohen

Cohen's evidence was that the potential acqmsttion of F AI had been in the contemplation of the HIH board throughout 1998. He said it had been discussed on several occasions by the HIH board prior to the meeting on 22 September 1998. These discussions occurred both outside and within board meetings during that time. 695 Cohen believed that the board's understanding and consideration of the proposed FAI acquisition was 'well advanced' by the time it resolved to proceed with the acquisition on 22 September. Cohen understood from the discussion between the directors at the meeting that HIH's management 'had undertaken considerable analysis of the pricing of the proposed acquisition'.

Cohen also understood that assistance and advice had been obtained from Hambros prior to the board meeting on 22 September. Hambros had been providing advice to HIH throughout the period when HIH had considered the possible acquisition. 696 Cohen took comfort from the fact that'FAI's 1998 accounts had been audited by Andersen. 697 Cohen believed that Williams put a 'convincing case' that HIH should proceed with the takeover at the board meeting on 22 September. 698

208 The impact of the FAJ acquisition

14.5.12 Fodera

Fodera recalled Williams and Richardson leading the discussions at the board meeting of 22 September about both the proposed offer price for the F AI takeover and the consideration which would be offered, namely a mixture of cash and HIH shares. Fodera told the board about his discussion with Davies and his review of FAI's accounts for the year ending 30 June 1998.699 Fodera's evidence was that he had performed an analysis of the possible synergy benefits that might arise from the takeover. He acknowledged that his analysis was cursory because he had no detailed

knowledge of the costs which F AI incurred in its business. It would be necessary to know these costs before a proper analysis could be undertaken of the possible synergy benefits. 700

14.5.13 Abbott

Abbott recalled that the meeting of 22 September was held at short notice and that, as a consequence, he had no documentation concerning the F AI acquisition before him in Melbourne. According to Abbott, Williams spoke to the board about the advantages to be gained from HIH acquiring a stake in FAI. Williams informed the board that the bid would be at $0.75 per FAI share. The consideration- being a

mixture of cash and HIH shares- was being evaluated by Hambros. Abbott inquired whether there were any further due diligence investigations to be conducted. Williams explained that a due diligence committee would be set up. He also reassured the board that he had held extensive discussions with Adler, whom he

knew well. Abbott formed the impression from Williams's comments that there was a close relationship between Williams and Adler. Abbott also believed that Williams would be able to achieve a favourable outcome for HIH.701 Abbott said that he took into consideration that HIH and F AI were both audited by Andersen which at the time had the reputation as one of the world's leading audit firms .702 He took comfort

from the fact that Davies served as a member ofHIH's due diligence committee and that Andersen made no adverse comment about FAI's accounts for year ending 30 June 1998 for which Andersen had issued an unqualified audit report. 703 Abbott also recalled that Williams was authorised to negotiate the purchase of an initial parcel of 10 to 15 per cent ofF AI only. 704

14.5.14 Stitt

Stitt confirmed that the meeting of 22 September which he attended by telephone, was conducted in circumstances of great urgency. Williams and Richardson said that the Adler family were contemplating selling some or all of their shares in F AI and that there were other possible purchasers in the market for those shares, so 'time was of the essence'. 705

Stitt asked whether HIH could conduct a due diligence investigation of F AI. The response was that due diligence could not be conducted because it was a hostile takeover. Access to the books ofF AI had been denied, but according to Williams HIH was 'justified in proceeding'.706 Stitt accepted Williams's reassurance that he

Th e failure of HIH Insurance 209

had recently spoken to Adler who had told him that FAI's recently published accounts were accurate and that HIH could rely upon them. Stitt did not press the question of due diligence any further. 707

Stitt's evidence was that Williams advised the HIH board at the

22 September meeting that synergy benefits of about $50 million per year would arise from the takeover.

Stitt also confirmed that the resolution of the board on 22 September 1998 was that a takeover bid might be launched only after Williams had obtained a pre-bid undertaking from Adler. 708

14.5.15 Head

Head expressed his opinion that he was in favour of proceeding with the takeover before leaving the meeting prior to its end. Head was happy to rely upon FAI's 1998 audited accounts and what he assumed was FAI's compliance with its continuous disclosure obligations. He was also comforted by the fact that Andersen had audited FAI and was confident that HIH's management could manage the acquisition ofFAl because it had previously managed the acquisition of ere so well. 709

14.5.16 Gorrie

Gorrie, who participated by video from Melbourne, recalled that Williams estimated synergy benefits arising from the takeover in the range of $30 to $40 million per annum and also indicated that other strategic and market benefits for HIH would arise from the takeover. 710 Gorrie understood that no due diligence investigation would be permitted ofF AI. It would therefore be necessary for HIH to rely solely on FAI's statutory accounts 'tempered by the Board members' knowledge of FAI, in particular as to some of its non-core assets, in its analysis ofF AI' s worth'. 711

14.5.17 Financial analysis

Fodera's review of F AI' s 30 June 1998 accounts revealed F AI to be financially weak. 712 Fodera calculated that FAI had net assets of $174.5 million, equating to $0.338 per share. This compared with the market value of F AI shares on 14 September 1998 (the day on which Fodera undertook his review) of $0.44 per share. Fodera noted that F AI' s balance sheet remained highly geared, with total debt of $395.2 million supporting net assets of $337.2 million, a debt to equity ratio of

11 7 per cent.

Fodera observed that the contribution to F AI' s profits from its general and life insurance activities had declined significantly since 1997. F AI' s underwriting profit for 1997 was $29.1 million, whereas in 1998 it was only $2.4 million. F AI' s reported combined ratio had increased from 111 per cent to 112 per cent, but Fodera thought the ratio should be calculated at 116 per cent. 713 Fodera considered that F AI' s underwriting results had been arrived at by the incorrect inclusion of certain

210 Th e impact of the FA! acquisition

profits on the sale of subsidiary entities and that the true pre-abnonnal operating result from F AI's insurance activities should have been a loss of approximately $17 million. 714 Fodera recorded his view that 'The results again are quite poor' .715

Fodera's analysis was not placed before the HIH board at its meeting on 22 September. The only document placed before those directors present in Sydney was Hambros' 'Project Vitamin financial model'. 7 16 This model analysed the acquisition of F AI in a range of prices from $0.67 to $0.77 per share after noting that the F AI share price prior to the takeover was $0.52 per share.

Hambros assessed the proposed offer as 'marginally dilutive' of HIH's net tangible assets per share. At an offer price of $0.75 per share the dilution was 5 per cent per share. The earnings effect of the proposed acquisition was said by Hambros to be 'dependent on the accuracy of the forecasts used for [F AI] and the level of synergies'. Hambros concluded that the synergy benefits of the acquisition would need to be at least $25 million pre-tax annually to justify the acquisition from a financial point of view. Hambros estimated that the takeover would reduce HIH's

earnings per share by 9.5 per cent on an undiluted basis and by more than 11 per cent taking into account the dilution effect of HIH's convertible notes at an

assumed offer price of$0.75 per share.

In Ham bros' analysis the only basis upon which the takeover could be justified financially was by reference to the synergy benefits which HIH might generate by the takeover. A range of synergy benefits from $25 to $40 million per annum produced an earnings per share enhancement (on a fully diluted basis) of between 2 per cent and 14 per cent per annum. Synergy benefits worth a minimum of

$25 million per annum were required to justify the takeover from a financial point of view. Hambros' analysis assumed the accuracy of FAI's 30 June 1998 audited accounts. The financial forecasts in Hambros' model for both FAI and (curiously) HIH were based upon a stockbroker's report prepared by Macquarie Equities. The

Hambros analysis contained many untested assumptions, particularly in relation to the level of synergy benefits which might derive from the takeover.

14.5.18 Lack of due diligence

Fodera gave evidence that he did not believe that a due diligence investigation would have made any difference to HIH's level ofunderstanding ofFAl because 'it took HIH management over 18 months to discover the magnitude of the F AI under­ reserving problems'. 717 This overlooks the fact that the early revelation of any potentially significant problem in the course of a due diligence investigation might have caused HIH to reconsider the merit of proceeding with the takeover of F AI

either at $0.75 per share or at all. To the extent that any due diligence investigation of F AI might have caused the directors of HIH to reconsider the wisdom of proceeding with the takeover the investigation would have been well worthwhile.

The fact that Adler refused to allow any due diligence to be undertaken does not appear to have been given specific consideration by the board. I note that Abbott did

Th e failureofH!H Insurance 211

not consider this to be an unusual situation, given that HIH and F AI were competitors. 718 Even so, there is no evidence that the directors turned their minds to the question whether there were other means of obtaining due diligence or alternative means of gathering further relevant information about F AI. The issue seems simply to have slipped away.

14.5.19 Williams on due diligence

It is apparent from Williams's evidence that there was a paucity of information available to HIH at the time of the takeover. Williams estimated that he spent about half an hour reviewing FAT's 30 June 1998 accounts after Adler sent them to him.719 Williams did recall having observed the reinsurance note to those accounts but did not think it indicated anything untoward. 72 0 Williams then had some general discussions with Adler about various aspects ofF AI' s business and its accounts. He did not go through FAI's 1998 accounts with Adler in any detail. 721

Williams understood that the $10 million increase in F AI' s reserves shown in the Hambros Project Vitamin analysis document presented to the HIH board meeting on 22 September was based on ' sort of just a feeling .. . just to top-up the reserves' .722 Williams did not understand there to be any objective basis for any of the valuations which appeared in the Ham bros document in respect of the F AI non-core assets or the valuation of its reserves. 723 Williams did not know whether any detailed earnings forecasts for the combined HIH-F AI group were ever produced in the course of the takeover. He agreed however that it would have been an important step for HIH to take in order for it to properly assess the merits of any takeover ofF AI. 724 Williams could not explain why only the forecasts of a stockbroking analyst (Macquarie Equities) were used in the Ham bros analytical document for the expected future earnings of both F AI and of HIH itself. 725

14.5.20 Wein on due diligence

W ein did not attend the 22 September board meeting and had no notice of it. However he expected that any strategic investor in an insurance company would require that a full due diligence investigation be undertaken on the target company and he expected that a full investigation would occur in respect of the F AI takeover. 726 In fact such a full investigation was what We in understood the HIH due diligence committee would undertake in the context of the F AI takeover. It was only much later that he learnt that the due diligence undertaken was actually 'very limited'. 727

14.5.21 Adequacy of HIH board's consideration

The consideration given by the HIH directors to the proposed F AI takeover on 22 September 1998 was inadequate having regard to the size of the transaction, its importance and the risks it posed to HIH. In the circumstances which prevailed the directors of HIH should have insisted that a careful and considered assessment be

212 Th e impact of the FA! acquisition

made of the true worth and future prospects of F AI. They should have required analytical justification for the claimed synergy benefits represented by Williams. They should also have required that a proper and considered basis be shown to them for proceeding so urgently with the takeover despite Adler's refusal to allow any due diligence investigation ofF AI to occur.

Many of the analyses which had been performed since at least 1995 had indicated that there were significant financial uncertainties for HIH in taking over F AI. Accordingly, the directors of HIH in September 1998 were faced with a choice: either proceed to make a bid with all of its attendant risks or decline to do so.

Having regard to the uncertainties which attended the takeover ofF AI, the directors of HIH acted rashly by approving the takeover. They gave cursory and inadequate attention to the bid at the hastily called board meeting of 22 September 1998 , attended by only three directors in person and four directors by telephone or video

link; the remaining five directors were overseas and had no notice of the meeting. Inadequate notice of the meeting was given and an inadequate analysis of the proposal was presented at the meeting to those directors who attended.

Several of the HIH directors submitted that the F AI acquisition was a commercial decision by the board which inevitably involved commercial risk. They submitted that the fact that hindsight shows it to have been a poor commercial decision, does not necessarily evidence undesirable corporate governance on their part. I am not so much concerned with the outcome of the bid as the process of considering and approving it.

The circumstances in which the HIH directors approved the making of the takeover bid in September 1998 was a clear instance of undesirable corporate governance. This is particularly so in the case of Williams because, as the minutes of the board meeting make clear, he was instructed to obtain a pre-bid undertaking from Adler

before bidding. He failed to do so. Instead, on 23 September 1998 he caused HIH to acquire the 14 per cent shareholding in HIH which Adler offered on the market that day and immediately thereafter publicly announced the takeover bid, thereby committing HIH to proceed with it. 728

14.5.22 Project Firelight

In late 1997 Malcolm Turnbull and Russel Pillemer of Goldman Sachs Australia LLC (GSA) began discussions with Adler concerning a possible joint-venture privatisation of F AI by GSA and Adler. On 24 December 1997 Turnbull, Pillemer and Adler discussed a proposal whereby GSA in conjunction with Adler would privatise F AI Life, then F AI and finally sell F AI Life to finance the takeover of F AI. 729 Adler told Turnbull and Pillemer that 'he thought he could be a billionaire in

a few years time if he was successful in doing this', but he was concerned at the financial risk involved. Turnbull responded that the strategy of selling F AI Life (following its privatisation) in order to finance the takeover, as well as having a

Th e failure of HIH Insurance 213

joint-venture equity partner like GSA, would reduce that risk. 730 The joint-venture privatisation proposal was given the GSA codename 'Project Firelight'.

In the early months of 1998, GSA communicated with the Goldman Sachs offices in Hong Kong and New York in order to develop and refine this proposal. 73 1 Project Fi relight was presented by GSA to the principal investment area (PIA) of Goldman Sachs as 'an attractive opportunity' .

Conflicting evidence was given by Adler and Landerer about whether Adler informed Landerer about Project Firelight in 1998. It is unnecessary to resolve this conflict in the evidence for reasons which are discussed below. Turnbull had advised Adler that he should inform Landerer about the proposal and was given to believe by Adler that he had discussed Project Firelight with Landerer. 732 Adler' s proposed retainer letter addressed to Turnbull dated 5 February 1998 contained the annotation 'cc. John Landerer' at the end.733 It is not necessary for the purposes of this report to decide whether this was a proper basis for Turnbull to assume (which he did) that Landerer was fully informed about the proposal.

Whilst it would have been preferable for Turnbull to have personally and directly verified that the non-executive directors of F AI were fully aware of the nature and purport of the discussions between GSA and Adler and the work GSA performed in connection with Project Firelight, this did not happen. Nonetheless it was not unreasonable for Turnbull to rely on Adler's word that he had discussed some aspect of the proposal with Landerer and had apparently sent him his letter of 5 February 1998.

Turnbull's evidence was that he dealt extensively with Adler and Mainprize in the course of Project Firelighe34 but not with any other F AI directors and, in particular, not with any of the F AI non-executive directors. I am satisfied that none of the non­ executive directors of FAI had any knowledge prior to this Commission that Project Firelight occurred. I also accept Landerer' s evidence that he was unaware of Adler' s 5 February 1998 letter to Turnbull735 , whether or not Adler ever sent it to him.

During the course of Project Firelight, GSA gathered substantial information for the purpose of undertaking a due diligence investigation of FAI. 736 For example, the minutes of the Goldman Sachs PIA investment committee held on 31 March 1998 refer to ' the due diligence work completed to date' in connection with Project Firelight. The minutes note that the investment committee and board of directors:

recommended that the team continue with its due diligence investigation and report back its findings. 737

Both Turnbull and Pillemer gave evidence that GSA did not misuse any confidential information provided to it from F AI because GSA kept the information confidential. 738 Even so , the provision 'of F AI' s confidential information to GSA without the approval of the F AI board and without the protection of a confidentiality agreement represented a misuse of the information by or at the behest of Adler. GSA was not responsible for that misuse by Adler but it ought to have been

214 Th e impact of the FAI acquisition

1/

conscious of the fact that Adler had a potential conflict of duty and interest and that the absence of a written legal confidentiality regime created a risk of abuse.

On 15 June 1998 Adler and Turnbull met the Goldman Sachs PIA investment committee in New York. 739 Adler tabled and spoke to a written presentation on the history, performance and prospects of FAC40 After the meeting, Goldman Sachs decided to send insurance specialists to Australia to undertake further due diligence

investigations of FAI so as to decide whether Project Firelight should proceed. 74 1 Shortly thereafter, Pillemer prepared a new due diligence investigation programme and checklist for Project Firelight.742 A separate due diligence programme was prepared in respect of Oceanic Coal. 743

During the week commencing 17 August 1998 GSA undertook more detailed due diligence investigations in pursuit of Project Firelight. 744 Adler and a number of other senior managers of F AI, including Wilkie, made presentations to representatives of GSA and Goldman Sachs who had come to Sydney from New York. 745 None of the non-executive directors of FAI were aware of any of these activities. It is not clear precisely who within F AI knew the real reason why the

representatives of Goldman Sachs were visiting FAI's Sydney offices. It is clear however that a number of senior F AI executives were so aware, including Adler, Mainprize, Wilkie and Zulman.746

At the meetings which occurred in Sydney in August 1998 between the representatives of Goldman Sachs and the senior executives ofF AI, there was some discussion ofFAl's use of financial reinsurance contracts. 747 Subsequently, Renaud Haberkorn of GSA commenced working on financial models for Project Firelight. These included models based upon alternative assumptions that FAI's financial

reinsurance contracts were or were not commuted in the near future. 748 Zulman also supplied Haberkorn with 5 year forecasts for F AI' s general insurance business showing the different results which might be achieved with and without the cost of the financial reinsurance contracts. 749

The Goldman Sachs representatives who visited Sydney from New York in August 1998 were less optimistic about Project Firelight as a result of what they learnt on their trip. 750 Turnbull shared that view. Shortly afterwards, on 7 September 1998 Turnbull, Pillemer and Haberkorn sent a memorandum to Michael Pruzan and Alec

Machiels of Goldman Sachs in New York advising them that GSA had decided to abandon the Project Firelight proposal. 75 1 The salient parts of this memorandum noted that over the previous two weeks GSA had worked with F AI management further to investigate F AI. This had provided GSA with a significant amount of additional information including:

• revised 5 year forecasts by product line

• information on FAI's loan and real estate portfolios, including details concerning the appropriate write down of these assets

• information on Oceanic Coal.

The failure ofHIH Insurance 215

The GSA memorandum noted that the information provided by F AI suggested that GSA could generate 'four year IRR in the order of 25-30 per cent'. The memorandum went on to note that after due consideration GSA did not recommend that Goldman Sachs PIA proceed with the opportunity due to what were described as ' the following major reasons':

We estimate that the 'true' net assets of [FAI] (i.e. after write-downs and unwinding of reinsurance contracts) is approximately $20 million, compared with a stated book value of $220 million.

• The projections provided by management are not supported by results achieved in 1997 and 1998 and we remain unconvinced that the current management team is capable of re-engineering this business.

• We would not support this projection since we feel particularly uncomfortable with the operating returns growth assumption and the level of write down at both the Holding company level and the General Insurance level. 752

The memorandum went on to discuss a number of write-downs which would be required to assets of the holding company (principally real estate assets totalling $17 million) and the general insurance company (principally loans, real estate assets and OCAL totalling $1 04 million) and noted under the heading 'reinsurance' that:

To cover losses generated by the long tail business in 1998 , [FAI] entered into a reinsurance arrangement that provided an extra income in 1998 and would generate losses over the next 4 years. The cost of breaking such a treaty is estimated at AUD80- 100 million.753

Pillemer considered that the breaking of the treaty would be welcomed by the market as a way of cleaning up F AI General's operations. 754

The memorandum concluded with a recommendation that GSA not proceed with this ' opportunity'. GSA's time records showed that 658 hours ofwork was recorded by GSA's directors and staff on Project Firelight. Approximately 70 per cent of this work (more than 450 hours) was performed by Pillemer and Haberkorn. 755

On 12 September 1998 Turnbull and Pillemer met Adler at his home.756 At the meeting Turnbull and Pillemer informed Adler that GSA did not propose to proceed any further with Project Firelight. Pillemer gave evidence that they informed Adler that GSA's suggested write-down to the net assets of FAI were up to

$200 million. 757 The discussion then turned to a consideration of the alternatives open to Adler and the desirability of him taking steps to pursue a sale ofF AI. 758

The first time that Adler mentioned to the non-executive directors of F AI that he had had any dealings with Goldman Sachs was in a memorandum he wrote them on 14 September 1998.759 The memorandum commenced 'I thought I would give you an update report on what I have been doing' . Adler's memorandum mentioned discussions with a number of bankers concerning a possible capital raising which Adler described as 'not the main game'. Adler then reported that he had met Turnbull of GSA in the previous week (in fact it was two days earlier, at his home)

216 Th e impact of the FA! acquisition

II

and 'we have appointed Goldman Sachs to review our various options from one end of the spectrum to the other (i.e. from raising additional equity to selling outright and all the options in between)'. Adler went on to mention a series of meetings which he proposed to have with various executives of insurance companies and

other possible investors in F AI.

The F AI board had been engaged in discussions with Adler throughout 1998 on strategic planning for the future ofF AI, in the course of which Adler had presented a detailed strategic plan to the board760 , making no mention of Project Firelight. Adler did not at any time advise the non-executive directors of F AI about his discussions with GSA concerning that project. Nor did Mainprize reveal to any of the F AI non-executive directors the role that he had played in Project Firelight or the fact that it had occurred. Mainprize's evidence was that he thought GSA had

been looking at a capital-raising proposal. 761

Turnbull's evidence was that discussions which he had with Landerer and with Hill immediately prior to the announcement of HIH' s takeover bid for F AI satisfied him that Landerer and Hill knew about Project Firelight and GSA's role in it. 762 I accept that Turnbull held the view that Landerer and Hill knew about Project Firelight and GSA's role in it. But I also accept their evidence that they knew nothing whatsoever about it and that at no time in 1998 did Adler, Mainprize or any of the other senior executives ofF AI tell the non-executive directors ofF AI about Project Firelight or GSA' s role in it.

Project Firelight was but one of several things that did not come to the attention of the board during 1998. Section 14.1 analysed the evidence concerning their knowledge ofFAl's under-reserving, which was kept from the board during 1998 as well. F AI, like all listed public companies, was subject to the provisions of the Corporations Law and the Australian Stock Exchange Listing Rules which obliged

it to keep the market informed of any price sensitive information, of which it became aware. I do not suggest that Project Firelight constituted price sensitive information but the fact that no information about it or about the serious under­ reserving problems which appeared in F AI during 1998 was communicated to the non-executive directors of F AI indicates a fundamental failure of communication between the senior executives, including the executive directors Adler and Mainprize, and the non-executive directors of the F AI board. That failure of communication was at the heart ofFAl's corporate governance problems.

F AI had in place directors' and senior executives' codes of conduce63 which should have ensured that relevant information about such matters was placed before the board. F AI seemed to have no proper procedures for ensuring that those codes of conduct were observed. 764 Similarly F AI appeared to have inadequate mechanisms

for identifying and communicating price sensitive information to the market. Basically, all market releases were made by Adler. 765 FAI's systems for making market releases were inadequate. Those systems should have ensured that at least one non-executive director was involved in the process of identifying and

formulating market releases on behalf ofF AI.

The failure of HIH Insurance 217

14.5.23 The events of 23 September 1998

On the morning of 23 September 1998 a meeting took place between Adler, Turnbull, Landerer, Williams, Richardson and Mark Pistilli and John Atanaskovic of Atanaskovic Hartnell, the solicitors advising HIH in connection with the takeover. Atanaskovic understood that Turnbull was present as Adler's adviser. 766 Atanaskovic told the meeting that HIH wished to procure a pre-bid acceptance agreement with Adler, as Williams had been instructed to do at the HIH board meeting the previous evening. HIH sought a 90 per cent minimum acceptance condition for the takeover bid. Adler responded that he would only agree to the pre­ bid undertaking if there was a 50 per cent minimum acceptance condition of HIH's bid. 767 The HIH and F AI representatives debated this issue but could not resolve it. HIH insisted that it required 90 per cent minimum acceptance, which would entitle it to compulsorily acquire the remaining shareholders ofF AI.

Adler then said that he was in need of some cash and had little alternative but to sell some of his F AI shares on the market and that he proposed to let all prospective bidders for large parcels of F AI shares know that he was proposing to sell on market. 768 Early in the afternoon of 23 September 1998 Adler offered a total of 45 million F AI shares owned by the Adler family company Lader Pty Ltd on the market at $0.75 per share. HIH immediately purchased those shares. Later that afternoon Adler was informed that HIH had bought them.

Williams took the view that, despite the clear terms of the resolution passed by the HIH board on the evening of 22 September 1998 (to the effect that he should obtain a pre-bid acceptance undertaking from Adler before launching the takeover), he nevertheless had authority to proceed with the unconditional acquisition of Adler's shares for cash on the market. 769 Williams's evidence was that because he had already agreed with Adler that the price that HIH would offer in any takeover was $0.75 per share, Adler was in no doubt of Williams's willingness to offer that amount for all of Adler's 30 per cent shareholding in FAC70

HIH immediately announced its acquisition of Adler's shares (representing 14.3 per cent of the capital ofF AI) and its intention to make takeover offers for all of the remaining ordinary shares in F AI. 771 However, unlike the on-market purchase which it made from Adler for cash, HIH proposed to offer F AI shareholders alternative consideration of HIH shares or a mixture of cash and HIH shares. The share offer was one HIH share for every three F AI shares. The share and cash alternative offer was $2.25 and one HIH share for every six F AI shares. Based on HIH's closing price of$2.29 on 23 September 1998, the share offer valued each FAI share at $0.763 cents. This represented a premium of approximately 47 per cent over the closing price ofF AI shares on 22 September 1998.

14.5.24 GSA advises FAI on the takeover

On 28 September 1998 Turnbull wrote to Adler a proposed engagement letter to 'confirm the arrangements' under which GSA was engaged to act for FAI in relation to the takeover for a fee of $1.5 million. 772 Turnbull advised Adler to give no advice

218 The impact of the FA! acquisition

to F AI' s shareholders about dealing with their shares pending the issue of F AI' s Part B statement. 773 Turnbull also wrote to Landerer noting that ' the key tasks for Goldman Sachs at this stage' were to:

prepare an assessment of the offer and a draft recommendation to directors. This will largely consist of comparative analyses of both F AI and HIH (given the paper component) . This task is in hand and we should have a draft report ready for your consideration by mid next week. We are, of course, well up the learning curve in respect ofF AI. 774

The FAI board met on 28 September 1998. 775 Adler proposed that GSA be appointed as F AI' s financial adviser in connection with the takeover. Landerer advised that he had also participated in discussions with GSA. It was resolved to appoint GSA for a fee of $1.5 million. No mention was made of Project Firelight or GSA's role in connection with it. All of the non-executive directors of FAI gave evidence that they would have wanted to know that GSA had been engaged in Project Firelight prior to making any decision to appoint GSA. Three of them said that, having regard to what had occurred in Project Firelight, they would have been unlikely to approve the appointment of GSA had they known the facts. 776

One of the tasks immediately undertaken by GSA was an effort to stimulate a rival bid for F AI. Turnbull sent Adler reports on 1 and 2 October 1998 concerning his efforts in this regard. Turnbull reported that there was unlikely to be any rival bid. Many of the other insurance companies which Turnbull contacted expressed the need for, and the condition of, extensive due diligence on F AI before they would

consider making an offer. 777

GSA also commenced work at this time on 'Preliminary valuation materials for Project Fireside'. Fireside was the codename given by GSA to the FAI takeover. 778 This work was the genesis of the written and oral presentations which were given by Turnbull and Pillemer to the board of F AI on 13 and 23 October 1998 for the

purpose of advising the board on FAI's response to HIH's takeover offer. The 'preliminary valuation materials' contained a wide range of valuation analysis and data about F AI needed to evaluate HIH' s takeover offer. 779

Some consideration was given by the board ofF AI to the question whether it should appoint an independent expert to prepare a report on HIH's takeover offer. Turnbull wrote to Adler recommending against the appointment of an independent financial expert for FAI. 780 As Turnbull pointed out to Adler, FAI was under no statutory obligation to prepare an independent expert's report (as was the case) and

it may not be in F.A.I's shareholders interest to have such an investigation conducted and presented in the Part B ... [FAI's] directors will have the benefit of an advice from Goldman Sachs upon which to rely. 78 1

In the event, the F AI board decided to commission Deloitte to advise it on the value of HIH shares alone and not to prepare an independent expert's report for inclusion in the Part B statement. 78 2

Th e failure ofHIH Insurance 219

14.5.25 FAI board meetings of 13 and 23 October 1998

The board of F AI met on 13 October 1998 to consider its response to the takeover. 783 Turnbull and Pillemer made a presentation to the board based upon a written document. 784 The document analysed various financial and commercial aspects of the takeover bid and assessed it against other comparable takeovers of general insurance companies in the past. It was noted that HIH's bid assumed that there would be a $60 million write-down of F AI' s assets 'as forecast by the market'. 785 A number of the F AI directors disagreed that such a write-down was necessary. 786

The non-executive directors of F AI gave evidence that they were particularly interested in that part of the presentation entitled 'Alternatives for [FAI]'. This part of GSA's presentation set out what was described as a 'decision tree analysis'. 787 In the options contained in the 'decision tree', the non-executive directors considered that there was one other possible transaction which deserved detailed consideration because, as GSA's presentation noted, it 'could generate returns to shareholders in excess of [HIH's] bid'.788 The strategy was described in GSA's document as involving the sale of a so-called:

'clean' general insurance business, as well as the break up and sale of remaining Holding Company assets of FAI over a 2-3 year period. 789

A more detailed financial analysis of this proposal appeared in a complicated chart on a separate page of the GSA presentation. 790 This chart analysed three different valuation scenarios identified as 'base case', 'best' and 'worst'. It postulated the sale of the non-core assets of FAI General so as to create a 'clean' general insurance company. Amounts of recapitalisation ranging from $44 million to $196 million would be required so that the clean general insurance company would satisfy the solvency requirements to remain a licensed insurer.

In the GSA analysis, various multiples of the net tangible assets of the clean general insurance company were then applied to estimate the net disposal proceeds of sale of that company. Those proceeds were then added to the assumed recoverable sale values of other assets of the holding company, including cash, Sydney and London property interests and various other assets. The total proceeds ranged from a 'best case' of$526 million to a 'worst case' of$149 million, the present values ofwhich were calculated at a 12 per cent discount rate over two and three year periods respectively. Assuming a three year 'work out' period the present values disco·unted at 12 per cent were $374 million on the best case and, on the worst case, $106 million.

The non-executive directors ofF AI gave evidence that they paid particular attention to the 'break-up and sell clean general insurance company' alternative and that considerable discussion about that alternative occurred at the F AI board meetings on 13 and 23 October 1998. 791 The non-executive directors considered that this transaction might provide a better return to the F AI shareholders than accepting HIH's take-over offer.

220 The impact of the FA! acquisition

At no time in the written or oral presentations given by GSA to the F AI board was any mention made of the fact that GSA in the course of Project Firelight had itself analysed in detail a very similar transaction to the 'break up and sell clean general insurance company' alternative which it presented to the FAI board on 13 and 23 October.

The Project Firelight privatisation proposal and the 'clean general insurance company proposal' were practically identical. It would have been of assistance to the directors of F AI to have known that GSA had spent considerable time in the course of 1998 analysing a very similar proposal in which Goldman Sachs might

invest its own money but had decided not to proceed with it. 792

Such information should have been revealed to the F AI board by a financial or corporate adviser like GSA because it would have assisted the directors to decide whether to appoint GSA as their financial adviser on the takeover. It would also have assisted the directors in forming their opinions about the viability of the ' break

up and sell clean general insurance company' proposal which presented by GSA as a potentially more attractive alternative than the takeover. The fact that these matters were not revealed to the board ofF AI is regrettable. This is particularly so in light of the evidence of some directors that it might have affected their attitude to the appointment of GSA.

Some of the non-executive directors gave evidence that had they known of the 7 September 1998 memorandum and the conclusions reached in it they would have called on F AI's auditors to clarify matters. It has to be borne in mind that the presentation made by GSA to the directors on 13 and 23 October 1998 concerning the 'break up and sell' option included reference to large write-downs in the value ofF AI' s assets. At least two of the directors said there was no need for a write-down because the audit had considered the carrying value of the assets. 793 The other issue was the perceived need to buy out the reinsurance contracts. Both the auditors and the board knew of the premium overhang. On the evidence it is too speculative to

say what would have happened had disclosure been made.

The FAI board met on 29 October 1998. 794 Turnbull provided an update on developments in the takeover. A report by GSA recommending that shareholders of F AI accept the HIH takeover offer in the absence of a higher offer was tabled. GSA reported that it had considered and taken into account advice received from Deloitte

in making its recommendation to the board. The minutes record that GSA's report and recommendations were ' extensively discussed'. Adler advised that he had spoken to the F AI directors who were not present at the meeting- Keys, O' Connell and Hill- all of whom had indicated that they were in favour of recommending the HIH offer to FAT's shareholders. It was then resolved that, based upon GSA's and Deloitte's advice, the FAI board would recommend that FAI shareholders accept

HIH's offer in the absence of a higher offer.

The failure ofHIH Insurance 221

14.5.26 FAI's Part B statement

It was agreed to make an appropriate announcement to the Australian Stock Exchange. The draft F AI Part B statement was then tabled and approved. The ASX was notified by a letter later that day. 795 Formal approval of the FAI Part B statement was given at a board meeting held on 10 November 1998 796, following the despatch ofHIH's Part A statement and offer documents to the FAI shareholders on 2 November 1998. 797

FAI's Part B statement was dated 13 November 1998 and was signed by Harris and Adler pursuant to a circular resolution of FAI's directors passed on

13 November 1998. 798 The Part B statement referred to a unanimous

recommendation of the FAI directors that FAI shareholders accept HIH's offer in the absence of a higher offer. It also advised that the directors intended to accept the offer in respect of the shares each of them owned. The Part B statement contained the following representation:

There is no other information material to the making of a decision by F AI shareholders whether or not to accept the Offer, being information that is known to any of the Directors of F AI and has not previously been disclosed to F AI shareholders in this Statement or the Part A Statement other than [the sale by Lader Pty Ltd of 45 million FAI shares on 24 September 1998 and the existence of 25 million converting preference shares in the capital ofF AI which could be converted into ordinary shares by resolution of the board ofF AI which had not yet decided whether to resolve to convert the converting preference shares]. 799

No mention was made in the Part B statement of the financial implications for F AI of the under-reserving which was not accounted for in its audited accounts for the year ending 30 June 1998. Nor was mention made of the future adverse financial impacts of the GCR and National Indemnity reinsurance contracts. These matters were known to Adler and Mainprize.

14.5.27 HIH directors' attitude if fully informed

Some of the HIH directors gave evidence to the effect that had they known the true financial position of F AI, HIH' s takeover bid would not have proceeded in the manner that it did, or at all.800 Williams, Abbott, Fodera and Head said that (at the least) they would have proposed a significantly reduced 'purchase price'. Williams and Fodera gave evidence that they would have been interested in HIH acquiring F AI' s businesses by a trade acquisition. Richardson also expressed similar views in his capacity as an adviser to the HIH board. 801

14.5.28 Treasurer's approval

One of the conditions of the HIH's takeover bid for FAI was that the Treasurer's approval of the takeover be obtained, as required by the Financial Sector (Shareholdings) Act 1998. APRA prepared an analysis of the takeover at the request

222 Th e impact of the FA! acquisition

of the Commonwealth Treasury. The Treasurer accepted Treasury' s recommendation that the takeover be approved. 802

14.5.29 Adler's termination payment and consultancy

On 2 December 1998 Adler wrote to Williams803 enclosing a copy of his service agreement with F AI dated 25 August 1995 and a legal opinion which he had procured from Clayton Utz on the previous day. 804 Adler inquired of Williams what was required of him as a consultant to HIH and for how long HIH would require his services. Adler suggested that four months' work would be required at a cost of ' say $40 000 per month' .

HIH sought advice from Minter Ellison concerning whether any payment could be made to Adler in respect of the termination of his employment contract at F AI or in respect of consultancy services which he might render to HIH following the takeover. Minter Ellison advised in writing that it was not lawful to pay Adler any amount in respect of the termination of his employment at F AI without risking a breach of s. 698 of the Corporations Law.

Williams believed that HIH was obliged to pay Adler a termination payment under his employment contract. He attributed this belief to something Lo had told him about the advice received from Minter Ellison. Williams conceded that such advice was never received by him in writing.805 He also agreed that no document was ever prepared setting out Adler's obligations under his consultancy arrangement with HIH.806 Nor was any obligation imposed upon Adler to provide any account of the time he spent on HIH's behalf in return for the $40 000 per month that he received. Nor did Williams seek explicit approval from the HIH board before agreeing to pay Adler $3.8 million in respect of the termination of his employment with F AI. 807

Acceptances of the HIH takeover bid were very low in early December 1998. Williams denied that there was any connection between his agreement that FAI pay Adler the $3.8 million termination payment or enter into the consultancy agreement and the perceived need to induce Adler to accept the takeover bid. 808 At the same

time (early December 1998) he communicated to Richardson in terms showing that the F AI bid would not be successful unless Adler accepted the bid. 809

Adler had a very unusual employment contract which contained terms concerning termination which were very beneficial to him. 810 The effect of the agreement was that either party could terminate Adler's employment on not less than 12 months written notice, provided that the period of notice expired 'on the day before an anniversary of the date of this deed'. In other words, the period of not less than

12 months written notice had to expire on 24 August in any given year. If notice was given on any date other than 24 August in a given year Adler would be entitled to a period of notice extending until the second subsequent occurrence of the date 24 August after the giving of the notice. Accordingly Adler might be entitled to

payment of his salary for up to two years less one day under the notice provision.

Th e failure ofHIH Insurance 223

Furthermore, Adler was entitled both to such a payment and a retirement benefit of three times his current salary, even if he himself gave notice of termination to F AI. This was the combined effect of clauses 9.1 and 9.4 of the employment agreement.

8 11

Clayton Utz's advice to Adler8 12 which was provided by Adler to Williams in December 1998 was that the combined effect of these provisions entitled Adler to a total payment of $3.8 million following termination of his employment.

In their advice to HIH, Minter Ellison pointed out81 3 that Clayton Utz had overlooked the constraints imposed by s. 23 7 of the Corporations Law on the proposed payment to Adler for termination of his contract in lieu of notice. Minter Ellison advised that s. 23 7 prohibited such a payment unless it fell within one of the exceptions under s. 237. The only exception that could apply was that for a 'genuine payment by way of damages for breach of contract'.

Minter Ellison observed that the proposed payment in lieu of notice to Adler was not intended to be a payment of damages, but was a payment of an amount expressly provided for under his employment agreement. Nor was there any suggestion that Adler's employment contract had been or would be breached by FAI.

Minter Ellison then advised that a breach of Adler's employment contract might be brought about by F AI summarily terminating it without notice. By these means Minter Ellison advised that the payment to Adler could be dressed up as a payment by way of damages for breach of contract.

This was a most inappropriate suggestion for a solicitor to make to a client in providing legal advice. When providing advice to a corporate client, a solicitor should consider his duty to the law and the interests of the corporation as a whole. It is difficult to conceive of circumstances in which a solicitor could properly advise a client that it should deliberately breach a binding and enforceable contract-that is, commit an unlawful act-in order to avoid a clear statutory prohibition on making an otherwise unlawful payment. The advice would in any event probably be nugatory because the law recognises only matters of substance, not form and because of the operation of the principle that the law will not permit wrongdoing to found a cause of action. I do not pause here to consider further whether that would have been so.

In so advising HIH, I consider that Minter Ellison was involved in and contributed to an undesirable corporate governance practice ofHIH.

In the event Adler's employment contract was not terminated in the manner suggested by Minter Ellison. Rather, Williams simply authorised the $3.8 million payment to be made to Adler without regard to s. 237 of the Corporations Law.

By letter dated 7 December 1998, Adler thanked Williams 'very much for agreeing to the payment philosophy of my Service Agreement'. 814 Adler's letter went on to state that 'in regard to the consultancy arrangement, I accept that this is a matter that

224 Th e impact afthe FA! acquisition

is currently undetermined and will be reviewed by you in later months'. It is clear therefore that Adler and Williams had agreed the question of Adler's termination payment sometime between 2 December 1998 when Adler first raised the issue with Williams815 and 7 December 1998.

Lo's evidence was that he never received advice from Minter Ellison orally or in writing that F AI was obliged to pay out Adler's contract in the manner suggested by Williams. 816 Lo's evidence was that after he gave Williams Minter Ellison's letter of advice dated 4 December 1998 817 he did not recall Williams asking him to do anything else in respect of the proposed payments to Adler. 818 This evidence was

supported by the evidence of the two Minter Ellison lawyers who were involved in the provision of the advice, Martin Bennett and Gareth Jolly. Both Bennett and Jolly said that they had no contact with anyone from HIH concerning their advice dated 4 December 1998.819

Williams agreed to make the payment to Adler directly contrary to the written advice of Minter Ellison, and without informing the board of HIH. The payment was made despite the fact that shortly afterwards Williams engaged Adler to act as consultant to HIH. Therefore there was no reason to terminate Adler's employment with F AI because HIH desired to retain Adler's services.

The authorisation which Williams gave FAI to pay Adler some $3.8 million in lieu of notice under his employment contract involved an undesirable corporate governance practice.

On 11 December 1998, four days after wntmg his letter to Williams820, Adler informed the Australian Stock Exchange that his 'personal company' Lader Pty Ltd had decided to accept the takeover offer from HIH in respect of the 16 per cent of the shares in F AI which it still owned. He stated that Lader had decided to take the

share alternative 'as we believe the positioning and strength of HIH (post acquisition) is not fully appreciated by the overall investment community' .821 Adler reported at the meeting of the F AI board on 16 December 1998 that the Adler family had accepted the HIH takeover offer, thereby increasing HIH's interest in F AI to approximately 36 per cent and that:

I am highly confident that once my accegtance is public knowledge that [HIH] will move over 51% fairly rapidly. 8 2

Adler's consultancy arrangements with HIH were never reduced to a written contract. Instead, Williams orally agreed with Adler that HIH would pay $40 000 per month to Adler and the arrangement was ratified at the HIH board meeting on 3 June 1999. 823 In February 2000 Williams unilaterally and without reference to the HIH board extended Adler's consultancy arrangement on the same terms.

824

At a meeting of the F AI board held on 9 February 1999, the board voted in favour of each of its members receiving various retirement allowances and termination packages from F AI. 825 In Adler's case the amount approved by the F AI board was $3.8 million. Reference is made in the minutes to a letter from Adler to Williams

dated 8 February 1999 826 in which Adler wrote: 'I would like to bring my personal

The failure ofHIH Insurance 225

arrangements to a head as I must plan the next stage of my life'. Adler enclosed with his letter copies of his service agreement with F AI, the opinion received from Clayton Utz on the appropriate amount of his termination payout and a calculation of the payout figures as at 28 February 1999. Adler commented that the payout was:

completely in line with the documents I sent you a few months ago and is only updated for the additional time between now and our last conversation.

Would you be kind enough to approve the attached so I can make the appropriate arrangements.

The minutes of the F AI board meeting record that Adler advised the board that the proposed termination payment to himself had been reviewed by Williams who had requested that it be approved by the board of F AI. Somewhat oddly, this entry appears in the minutes after reference to the fact that Adler had left the meeting 'as required under Section 232A of the Corporations Law'. 827

14.5.30 Andersen's involvement in the FAI takeover

Andersen was the auditor of both FAI and HIH for many years prior to the takeover. Davies was the audit engagement partner for the F AI audits for eight consecutive financial years between 1990 and 1997. Davies' evidence was t}:lat during the year ending 30 June 1997, there commenced a planned transition whereby Scrivens would become Andersen's engagement partner for the FAI audit in 1998.

Scrivens was the engagement partner for the audit of the F AI group for the year ending 30 June 1998. Because of Davies' prior experience, he remained the engagement partner for the audits of Oceanic Coal and F AI Workers Compensation (NSW) Ltd and the statutory fund managed by it. In order to assist with the transition from Davies to Scrivens, Davies was designated as the relationship partner for the 1998 FAI audit. 828 In Andersen's presentation to the FAI audit committee on 18 June 1998, Davies' role as relationship partner was described as acting as an alternative contact point if Scrivens was not available. 829

Davies described his involvement in several important discussions and meetings concerning key issues which arose in the F AI 1998 audit, including his attendance at the FAI audit committee meetings held on 18 June 1998830 and 3 September 1998. 83 1

Apart from auditing Oceanic Coal and F AI Workers Compensation (NSW), Davies was consulted several times by Scrivens and by other Andersen partners and staff in connection with the proposed accounting treatment of the GCR and NI reinsurance contracts in the F AI 1998 audit. 832 Scrivens gave evidence that he regarded Davies' opinions on the F AI audit matters in respect of which he consulted Davies during

1998 as very valuable to him. 833 's audit ofF AI' s accounting treatment for the GCR and NI reinsurance contracts in 1998 is considered in Section 14.4.

During the course of the F AI takeover, Fodera asked Davies to review with Fodera F AI' s 1998 annual report. 834 Davies denied that he said anything to Fodera about FAI's 1998 accounting treatment of the GCR and NI reinsurance policies. 835 Davies

226 The impact of the FA! acquisition

said that Fodera asked him questions about the reinsurance policies that FAI had entered into and referred Davies to the reinsurance note in F AI' s 1998 financial statements.836 Davies' evidence was that Fodera asked him what Davies knew about the meaning of that note. 837

Davies' best recollection was that he didn't answer Fodera's question but instead asked Fodera what prompted his question. Davies' evidence was that Fodera responded that HIH was aware of certain reinsurance contracts that had been entered into by F AI but Fodera did not reveal what reinsurance contracts HIH was aware of.

When Davies asked Fodera how he became aware of the FAI reinsurance contracts, Fodera told him that Williams had been advised of the contracts by Gardener. 838 Davies said nothing else on this subject during HIH's takeover ofFAI.839

Davies didn't reveal his knowledge of the GCR and NI contracts and their treatment in FAI's 1998 accounts to Fodera because he didn't believe it was appropriate to reveal to Fodera information about FAI that wasn't public knowledge. 840 He believed that there was a legal limitation upon him discussing with Fodera anything that was not public concerning F AI and that he had an obligation of confidentiality to FAI. 841 During the course of the 1999 HIH audit Davies still felt bound by his

obligation of confidence to F AI, even though F AI was then a wholly-owned subsidiary of HIH. 842

Fodera gave evidence that when he endeavoured to discuss FAI's 1998 accounts with Davies, Davies stated that he felt that there was a conflict arising from Fodera's request for information. If Davies was to disclose any information to Fodera, it had to be approved by FAI's management. 843

Davies could not recall when or by whom it was first proposed to him that he might serve as a member of HIH's due diligence committee for the purposes of the takeover. 844 He did not recall discussing with anybody else whether it was or was not appropriate for him to serve on that committee. He expected he would have discussed it with Gooley, Andersen's audit practice director, because 'it was clearly

a potential for conflict ... and therefore a matter that I was required to clear with the practice director before accepting the appointment'. 845 He did not recall any misgivings being expressed to him by Gooley or anybody else about the proposal that he serve on the committee. Davies identified the source of the 'potential

conflict' as:

My previous involvement and knowledge of F AI and possession of information- FAI client confidential information and my inability to share that information with HIH. 846

Davies considered that if Andersen was:

were able to establish satisfactory terms of engagement with HIH, which limited the scope of [Andersen's] involvement such that the conflict did not arise, then it would be appropriate for [Andersen] to accept the engagement.847

Th e failure ofHIH Insurance 227

Davies agreed that when he served on the HIH due diligence committee he knew a good deal of confidential information about F AI arising out of the 1998 F AI audit. 848

Davies agreed that he would have been in a position to tell the HIH due diligence committee at its first meeting on 2 October 1998 that the 1998 audited accounts of FAI (which had been released approximately three weeks earlier) accounted for the beneficial financial effects of the GCR and NI reinsurance contracts and not the 'premium overhang' and also contained the high stock market price of PAl's shareholding in Home Security International reached on 30 June 1998. 849

Davies considered that the scope of Andersen's involvement on the HIH due diligence committee was limited in such a way as to prevent a conflict arising between the duties it owed to HIH and FAI respectively. Davies thought that Andersen's report to the committee:

made it quite clear what the terms of engagement were . . . that the information that we were dealing with was restricted to public information and basically information that could be derived from FAI's annual report and our involvement was so limited. 850

Careful examination of the records of the HIH due diligence committee, including the reports of the committee which Davies signed on 28 October 1998, does not in my opinion support Davies' views. There were two reports of the due diligence committee. One was a report on the prospectus for the convertible notes offered by HIH Holdings (NZ) Limited to raise the cash consideration payable under the takeover. 851 The second was a final report of the due diligence committee on the Part A statement for the purposes ofHIH's bid. 852

Both due diligence committee reports contained references to the fact that Andersen was the accounting adviser to the due diligence committee. Andersen's tasks were described as having 'included ' various matters set out in the reports. The reports of the due diligence committee do not say that Andersen's work on the committee, its obligations to the committee, to its members or to HIH were limited only to the tasks so described in the reports. Nor do the reports say that Andersen had access or recourse only to publicly available information about F AI or any other subject. Significantly, in the sign-off paragraphs of the reports, which appear immediately above the signatures of the members of the committee (including Davies), it was stated that:

228

Nothing has come to the attention of any other member of the Due Diligence Committee .. . that causes that member to believe that the Part A Statement [prospectus] contains a material statement that is false or misleading.

Each member of the Due Diligence Committee is satisfied .. . that the question of whether the Part A Statement [prospectus] contains a material statement that is false or misleading has been approgriately considered in the preparation of the Part A Statement [prospectus ]8 3

The impact of the F AI acquisition

The due diligence committee's prospectus report also contained the representation that no member of the committee had any reason to believe that there was a material omission from the prospectus or that its issue may involve conduct that was misleading or deceptive.

854 The making of these representations does not sit easily with the proposition that, in the case of Andersen or Davies, their sources of relevant information, knowledge or belief were limited to publicly available

infonnation about F AI. The representations in the reports quoted above imply that all the members of the committee had consulted all their available sources of knowledge, information and belief, public or private.

14.5.31 Andersen's letter of advice

Andersen wrote a letter of advice to the directors of HIH and the members of the due diligence committee dated 26 October 1998 in connection with the Part A statement and the prospectus.855 The letter was annexed to both reports of the due diligence committee. The letter said that Andersen had performed:

a limited examination of the aggregation of the proforma consolidated balance sheet of [HIH] and [FAI) ... We have also performed a limited examination of the profit and loss accounts of HIH and F AI for the years ended 30 June 1997 and 1998.

Andersen's letter went on to note in detail the work that Andersen had done.

In my view, nothing in Andersen's letter of 26 October 1998856 stated or implied that the only information about F AI to which Andersen had recourse for the purpose of its service on the due diligence committee was publicly available information. Indeed, having regard to the sign-offs which Davies gave on behalf of Andersen in the reports of the due diligence committee, it would not have been correct to have

made such a statement in Andersen's letter. Nor did Andersen's letter suggest that any information or belief held by Andersen which was relevant to the consideration of offerees who read the Part A statement or the prospectus would not have been revealed or considered by Andersen using all its sources of knowledge or belief for the purposes of the due diligence exercise. This was the case even though the due

diligence planning memorandum said this would happen and the reports of the due diligence committee said it had in fact happened.

There is no reference in the minutes of the due diligence committee of Davies having clearly stated to the committee that the only information to which Andersen would have recourse whilst serving on the committee was publicly available information about F AI. Nor is there any record in the minutes of the committee of

Davies having described to the committee the practical implications of such a limitation of Andersen's sources of information. For example, Davies did not explain to the committee that he would not reveal what he knew about FAI's 1998 accounts, including the profits which had been booked by F AI from the GCR and NI contracts or the guaranteed accounting losses (the premium overhang) which

would arise from those contracts thereafter, or the high valuation of FAI's shareholding in Home Security International.

The failure of HIH Insurance 229

14.5.32 Final due dil igence reports an d sign-offs

Section 7 of the planning memorandum noted that it was intended that the final due diligence reports would contain confirmation signed by each member of the committee that it was satisfied with the due diligence procedures adopted and that

in relation to its particular areas of expertise, nothing has come to its attention that causes him [sic] to believe that the Prospectus and the Part A Statement contains a material statement which is false or misleading or omits material information.85 7

Davies' signature appears with those of all other members of the due diligence committee858 as having read and adopted the planning memorandum under which the work of the committee had been done.

Davies was unsuccessful in his attempt to limit the scope of Andersen ' s or his role on the HIH due diligence committee in the manner which he says he intended. Having regard to his knowledge of FAI's affairs, and in particular the significant issues which had arisen in the course of the audit of its 1998 accounts, Davies was placed in a position of almost inevitable conflict between the duties he owed to F AI and HIH respectively. I consider that Davies' participation in the HIH due diligence

in these circumstances involved or contributed to undesirable corporate governance of HIH because it undermined the purpose of HIH's due di ligence and the understanding of the members of the committee about the scope and nature of its work.

14.6 T he 'total re t urn swap' transaction

The cash consideration paid by HIH for the takeover ofF AI was raised by an issue of $155 million of converting notes by HIH Holdings (NZ) Limited, a wholly­ owned subsidiary of HIH, pursuant to a prospectus dated 26 October 1998. 859 A significant collateral transaction was entered into between HIH and Societe Generale Australia Limited (SocGen) the effect of which was not fully understood by the HIH board and was not disclosed in the prospectus.

14.6.1 The note iss ue and the collateral transactions

SocGen was a joint underwriter with Macquarie Equities of the converting note issue. It also received a $35 million priority allocation of the notes. Negotiations for the terms of the priority allocation were conducted from about 1 October 1998 and involved Fodera, John Harvey of SocGen and Richardson. 860

From the outset, SocGen's proposal included a collateral transaction between HIH and SocGer. called the total return swap (the swap). As part of the swap, a collateral deposit was to be paid by HIH to SocGen of $35 million, equal to the amount of the priority allocation underwritten by SocGen. The collateral deposit entitled HIH to receive interest at a rate higher than the deposit market rate. 86 1 It was also designed to ' secure HIH's obligations under the swap'. The deposit served to protect SocGen

230 The impact of the FAJ acquis ition

in the event that the notes declined in value, in which event SocGen had the right to sell the notes and apply the deposit to indemnify itself against any loss which it thereby suffered.

Draft terms for the swap, including the collateral deposit, were set out in a letter from Richardson to Fodera, Harvey and Brown dated 13 October 1998. 862 Key terms included that the deposit could be withdrawn by HIH at any time on 30 days written notice; that on the giving of a withdrawal notice SocGen could sell the notes; that any profit on sale of the notes was HIH's; and that SocGen had a right of

set-off against the deposit to recover any loss on the transaction. HIH was to receive interest on the notes less a margin payable to SocGen. The amount involved was expected to be $35 million. 863

Harvey confirmed these arrangements in a letter to Fodera dated

22 October 1998.864 The letter also contained a further requirement that the collateral deposit equal the notional amount of the swap. The notional amount was identified as $35 million. 865

The fax message covering Harvey's letter noted that

SocGen required HIH to agree to the swap at the same time as the underwriting agreement between HIH and SocGen in respect of the notes. 866

14.6.2 The purpose of the swap

The evidence shows that SocGen required the collateral deposit in order to protect its position. This is confirmed by an internal memorandum from Harvey to SocGen's managing director, Michel Macagno, which stated that SocGen entered into the swap 'as a means of reducing the overall underwriting exposure to the Issue'. 86 7 The internal SocGen memorandum went on to make clear that the purpose of the transaction was to protect SocGen against any loss which it might suffer as a result of subscribing for the notes. The SocGen internal memorandum and the

requirement that the swap transaction be entered into at the same time as the underwriting agreement show clearly the connection between the two transactions. In effect SocGen's underwriting was conditional upon HIH agreeing to the swap.

Fodera confirmed the essential terms of the transaction- that the collateral deposit was received under the swap by SocGen as security against any future loss on the notes and that SocGen had the right to sell the notes and apply the deposit to recover any loss which it had suffered. He agreed that the transaction meant that SocGen was not at risk of any loss on the notes it subscribed for because it had recourse to the collateral deposit. 868

Brown also agreed that the effect of the transaction was to put SocGen in a different and more favourable position in relation to the $35 million of notes it subscribed for than other noteholders. 869 By contrast, other noteholders faced the possibility of a capital loss in the event of a decline in the value of the notes or of HIH shares into which they converted.

The f ailure ofHIH Insurance 231

14.6.3 Disclosure to the HIH due diligence committee and the board

The due diligence committee formed to approve the prospectus for the note issue and the Part A statement for the takeover comprised representatives from HIH, Andersen, SocGen, Hambros, Minter Ellison, Atanaskovic Hartnell and Macquarie Equity Capital Markets Limited. Fodera, Cohen and Stitt represented HIH. Davies represented Andersen. Richardson represented Hambros. Brown represented Minter Ellison. Harvey represented SocGen and Pistilli represented Atanaskovic Hartnell.

Due di]jgence committee meetings were held on 2 October870, 9 October87 1, 13 October 872 , 16 October

873 and 22 October 1998. 874 At none of these meetings

was any reference made to the swap. The evidence of Richardson, Stitt and Brown confirmed that the swap was not considered by the due diligence committee. 875 Consequently, the fact that SocGen as underwriter was indemnified against any loss because it held security in the form of the collateral deposit was not disclosed to or addressed by the due diligence committee so that it could decide whether the swap should be disclosed in the note prospectus or the Part A statement. In the event the swap was not disclosed in either document.

Fodera submitted that the relevance of the failure to inform the due diligence committee was not apparent because the HIH board was informed of the swap and all the non-board members of the due diligence committee were involved in the preparation of the swap as financial advisers or lawyers and were therefore aware of

it. I do not accept that that was correct in the cases of Andersen or Atanaskovic Hartnell, although I accept that Fodera had reasonable grounds for his belief that Pye and Davies of Andersen knew about the transaction. The issue, however, is not the knowledge of individual members of the due diligence committee. Rather it is the consideration by the committee as a whole of the question whether the transaction ought to have been disclosed in the prospectus and the Part A statement. The due diligence committee did not consider that question because the issue was never placed before it.

Despite the fact that it was not placed before the committee, the swap was raised by Fodera at the HIH board meeting on 16 October 1998. 876 Fodera tabled a discussion paper entitled 'New Converting Notes' at the meeting877 which was based on written advice from Richardson.878

The minutes of the 16 October 1998 HIH board meeting record that Fodera 'advised that the Company intended to assume the investment risk on $35 million of Notes through the use of derivatives with SocGen as the counter party. HIH's reason to enter into that transaction was to capture the above average interest margin' .879 The discussion paper said that 'it is proposed that of the $176 million (maximum) to be raised through the issue of a new series of converting notes, $35 million are to be purchased by Societe Generale Australia Limited . .. but with the investment risk eventually borne by HIH (similar to credit insurance) through the use of derivatives'. 880

232 The impact of the FA! acquisition

Other than what appears in the minutes, the nature and scope of Fodera' s explanation to the board is unclear. There is no indication from the board minutes that Fodera explained the commercial intent or effect of the collateral deposit as eliminating any risk for SocGen arising out of its subscription for notes. Stitt said that his understanding of the swap was limited to the explanation given by Fodera at the board meeting, namely that the swap would permit HIH to invest some of its money at a rate higher than it could otherwise obtain in the market. 88 1 Stitt said he did not understand that SocGen was indemnified against any risk of loss on the

notes it held. 882

Similarly, Abbott said he did not understand the nature and effect of the swap. He was unable to recall any director asking Fodera for an explanation of the transaction in terms that he could understand. 883 Abbott said he was unaware that the effect of the transaction was to eliminate any risk for SocGen in subscribing for notes. 884

Williams said that he did not realise that SocGen was not taking any risk on the $35 million subscribed. Indeed Williams only became aware in October 2000 that SocGen was at liberty to sell its notes, realise the loss and then indemnify itself from the collateral deposit because the share price of HIH had fallen below $1.00 for five consecutive trading days. 885

Cassidy, Cohen, Head and Sturesteps had no specific recollection of the terms of the swap or of any discussion about it at the board meeting on 16 October 1998. 886

Counsel assisting submitted that Fodera might have breached the law by not giving a comprehensive presentation to the board about the total swap transaction. But I accept that Fodera may not have appreciated that board members did not understand the full ramifications. I do not find that Fodera withheld information either

negligently or deliberately.

14.6.4 The HIH board approves the documents

The swap documentation was tabled at the HIH board meeting on

22 October 1998.887 As was required by s. 205 ofthe Corporations Law, the minutes record that the 'board noted that in the opinion of each of the directors, after taking into account the financial position of the Company (including future liabilities and contingent liabilities of the Company), that the giving of financial assistance as entailed in Derivative Contract [the swap] was not likely to prejudice materially the

interests of the Company or the interests of the creditors or members of the Company or any class of those creditors or members. No director of the Company voted against the resolutions of the directors approving the giving of financial assistance .. . and adopting the above statement' . 888

The board resolved to appoint a committee comprising Cassidy and Stitt to settle, sign and execute the note deed, the guarantee deed, the underwriting agreement, the loan agreement, the Part A statement and offer and the swap.88 9 A later resolution was made at a board meeting on 26 October 1998 which authorised Fodera and

Williams to settle, sign and execute those same documents on behalf ofHIH. 890

Th e failure of HJH Insurance 233

14.6. 5 The converting note prospectus

A short-form prospectus under s.1022AA of the Corporations Law was issued on 26 October 1998. ASIC granted HIH exemption from compliance with the requirements of s. 1022 for a full prospectus. The prospectus did not mention the swap or its effect.

89 1 A declaration was made in the prospectus pursuant to s. 1 008A of the Law that the statements in the prospectus were true and not misleading. 892

SocGen was identified on the front page of the prospectus as one of the two 'joint lead managers and co-underwriters to the note issue'. 89 3 Several of the representations made in the prospectus were contrary to any suggestion that a transaction along the lines of the swap was proposed. For example, in the section describing the 'Key Terms of the Issue'894, the 'effect of the Note Issue on HIH' was said to be described in schedule 1 to the prospectus. Schedule 1 made no reference to the swap or the collateral deposit, or any transaction along similar lines.

The prospectus also contained a 'Summary of Terms of the Notes'.895 Again, there was no mention of the swap or the collateral deposit or any transaction to similar effect in that summary. Nor were any such transactions described in schedules 2 or 3 to the prospectus which set out the conditions of issue of the notes and the terms of the deed of guarantee for the notes entered into by HIH. In the section describing the ranking of the notes amongst HIH's obligations896, the prospectus stated ' the Notes will rank equally with other unsecured obligations of [HIH]' . No mention was made of the fully secured position which had been achieved by SocGen as a result of receiving cash collateral for its notes by the deposit of $35 million from HIH.

Part 3 of the prospectus was entitled 'Additional Information' and included a summary of the terms of the underwriting agreement which had been entered into between HIH, SocGen and its co-underwriter Macquarie Equity Capital Markets Limited. 89 7 There was no mention in the summary of the underwriting agreement of the swap or the collateral deposit or any transactions to similar effect. I accept the submission of counsel assisting that this may have resulted in a breach of s. 1 022AA of the Corporations Law which required that the prospectus contained:

all such information as investors and their professional advisers would reasonabl y require, and reasonably expect to find in the prospectus, for the purpose of making an informed assessment of:

(i) the effect of the offer or invitation on the disclosing entity and

(ii ) the figures attaching the securities.

In my opinion, the swap was relevant to both these matters. The question that arises is who (if anybody) bore responsibility for that breach.

14.6.6 Minter Ellison's advice on the prospectus

On 26 October 1998 Brown (HIH' s legal adviser on the note issue) provided written advice to the HIH board that, to the best of Minter Ellison' s belief, the prospectu s did not have any material statement that was false or misleading, did not contain a

234 The impact of the FA I acquisition

material omission having regard to s. 1022AA of the Law and its issue did not involve conduct that was misleading or deceptive or was likely to mislead or deceive. 898

Stitt said he relied on the advice of Brown as to any disclosure which was required in the prospectus. 899 Brown did not give any advice at the board meeting suggesting that there was any legal requirement that the swap be disclosed in the prospectus or that there was a connection between the swap and the underwriting agreement

which would require disclosure of the swap in the prospectus or in the Part A statement. Brown did not suggest that the failure to disclose the swap, including the collateral deposit, in the prospectus or the Part A statement would involve any misleading or deceptive conduct.90° Fodera, Cohen, Abbott, Head, and Gorrie all submitted that they were entitled to rely on Brown's advice. I accept these submissions because I consider that this was a complex transaction the commercial and legal ramifications of which should have been explained. Richardson also made a similar submission. In his case, by virtue of his position as the financial adviser to

HIH, I consider that he should have seen clearly the commercial and financial implications of the swap and the need for its disclosure. However I do not consider that he ought to have done more than to bring these matters to the board' s attention. He was entitled to rely on Brown's legal advice in this regard, assuming (as he was entitled to do) that Brown understood the commercial and financial ramifications of the transaction and the law relating to prospectus disclosure.

14.6.7 Andersen 's knowledge

The question also arises whether Andersen were made aware of the terms of the swap and the collateral deposit. In its submissions901 , Andersen said that the evidence did not show that Pye or Davies were informed of the details of the swap. I accept that the extent to which Andersen was made aware of the swap and the collateral deposit is not clear on the evidence. Pye was present at the HIH board meetings of 22 and 26 October 1998. The minutes of the first meeting record that he

left the meeting before matters concerning the swap were discussed. 902 At the second meeting, on 26 October, the swap documents were tabled and approved but this was a short formal meeting and it does not appear that any matter of substance about the documents was discussed. Pye might have become aware of the swap at

that meeting but he denied any knowledge of the swap. 903 It might have appeared to

those who attended the board meeting of 26 October that Andersen was aware of the swap. I do not consider that any director or adviser of HIH was or ought to have been aware that Andersen was ignorant of the swap, assuming that to be the case. The evidence on this issue was not sufficient to enable me to make any finding.

14.6.8 Accounting issues

An issue before the due diligence committee was Andersen's opm10n on the consolidated pro forma balance sheets ofF AI and HIH which were included in the

Th e failure ofHIH Insurance 235

prospectus and the Part A statement. The financial implications of the swap were plainly relevant to the presentation of the consolidated pro forma balance sheets.

Davies was a member of the due diligence committee but he did not attend the 16 October 1998 HIH board meeting. He did not recall having seen Fodera's 'New Converting Notes' discussion paper904 , or any discussion about the transaction.905 Pye attended the 22 October 1998 board meeting although as noted earlier he departed before the issue was raised. In an internal memo of 23 October 1998906 , Pye gave consideration to whether the notes should be classified as debt or equity in HIH' s accounts. That memo revealed no knowledge by Pye of the details of the swap which were known to Fodera and others.

AASB 1033 907 regulated the classification of financial instruments such as converting notes as debt or equity. In accordance with AASB 1001 the classification must have regard to matters of substance not form and the substance is not to be ignored because it is contained in more than one agreement. Clause 4.1 provided that classification was to be made in accordance with the substance of the contractual arrangements in accordance with the definitions of 'financial liability ' and 'equity instrument' .908 A 'financial liability' meant any contractual liability to deliver cash or another financial asset to another entity.909 As cash was to be placed on deposit with SocGen and be available to meet any loss arising from the notes, at

least $35 million of the notes subject to the swap should in my opinion have been classified as a liability and not as equity.

The consolidated pro forma balance sheet in the prospectus showed that all of the converting notes to be issued were treated as equity. 910

In the 1999 and 2000 consolidated financial statements of HIH all of the 1998 converting notes were classified as equity not as liabilities. 911

The statements in the pro forma balance sheet and in subsequent financial statements of HIH were wrong in this respect.

On 2 November 2000 the HIH board decided to seek legal advice on the swap. 912 Minter Ellison (Bennett) advised HIH on 14 November 2000 that the swap should be accounted for as a financial instrument and if the amount potentially payable by, or to, HIH on the termination of the swap was material, it should be accounted for as such. Minter Ellison proposed a note disclosure to the 30 June 2000 accounts. It advised that the 'form and extent of this disclosure should take should be considered by ... the Board and the auditors of HIH'.913 Lo sent a memo to Fodera on 15 November 2000 enclosing the legal advice.9 14 He noted that the annual accounts for HIH had already been signed and that the HIH annual report was about to be

despatched to shareholders. He said that any change to the accounts would require Andersen' s concurrence. At its meeting on 15 November 2000, the HIH board considered the matter and decided it was not necessary to make disclosure in the 2000 accounts and that the issue would be reconsidered in 2001.

9 15 The opportunity

for that of course did not arise.

236 Th e impact of the FA! acquisition

14.6.9 The role of the legal advisers

Minter Ellison was the legal adviser to the HIH board on this transaction. Brown said he did not tum his mind to the question whether it was material for prospective investors in the notes to know that SocGen was itself subscribing9 16, nor to the is sue of disclosure of the swap in the prospectus at or before the 16 October 1998 board meeting. No consideration was given by him or any of the directors to the question of disclosure of the swap. 917 Brown appreciated that the HIH board needed to be satisfied that the prospectus was true and not misleading and that the members of the board relied upon the due diligence committee, of which he was member, for reassurance about this. 91 8 He said that the main focus of the 16 October 1998 board meeting, at which the swap was first discussed, was whether s. 205 of the Corporations Law (financial assistance) was complied with. 9 19 This is consistent with the evidence of others present at that meeting.

However, as the legal adviser to HIH and its directors, Brown should have given consideration to the board's need to be fully informed about the impact of the swap on HIH's disclosure obligations in the prospectus. The transaction was complex and it had a significant effect on the rights of HIH, SocGen, other converting

noteholders and other creditors of HIH (who were entitled to assume that SocGen' s notes were quasi-equity, not in effect a secured loan as a result of the collateral deposit). More attention should have been given by Brown to this critical issue.

In his submissions, Brown raised a question whether s. 1 022AA of the Corporations Law required disclosure of the swap for reasons relating to the construction of the section and noted that no expert evidence was taken on the point. I do not believe that I would have been assisted by expert evidence because I regard the issue as

straightforward. In my view subscribers and prospective subscribers for notes would reasonably have expected that all of the material terms on which SocGen as joint­ underwriter was to subscribe for $35 million of notes, including the swap and the collateral deposit, would be explained in the prospectus. On his evidence the matter was not even considered by Brown as an issue on which legal advice might be

needed.

14.7 The financial consequences of HIH's ta keover of FAI

Counsel assisting and many of the witnesses and parties before the Commission identified the takeover of F AI as a contributing cause of the HIH collapse. By contrast a number of parties submitted that HIH suffered either no loss or no material loss because of its takeover ofF AI. In this part of the report I examine the evidence and make findings about the financial consequences of the takeover. It is

impossible to be precise. The evidence does not permit an exact calculation to be made. I have done the best I can on the avai lable material, but it is very much an estimate.

Th e failure of HIH Insurance 237

In the corporate governance Chapter 23 address some other less tangible (but no less real) adverse effects of the FAI takeover upon HIH's systems of management, governance and control.

14.7.1 FAI's under-provisioning as at 31 December 1998

Any analysis of the financial implications of HIH's takeover of FAI must take into account the extent of the under-reserving in FAI's claims provisions at the time of the takeover; that is, FAI's reserves as at 31 December 1998. I have found that there was substantial under-reserving in FAI's claims provisions at all times after 30 June 1997.

Section 14.1 of this chapter addressed the level of under-reserving in F AI as at 30 June 1998, the last date on which audited accounts ofF AI were published prior to the takeover by HIH. I noted the work done by and the communications passing between Spratt, Moran and Wilkie about the levels ofFAl under-reserving ofwhich they were aware during the course of the takeover in the last months of 1998.

Because Spratt was Peter employed by HIH following the takeover various HIH employees including Thompson, manager of HIH's risk management and reinsurance division, soon became aware of Spratt' s work and his conclusions about the level of under-reserving within F AI. HIH then began a reassessment of all of FAI's claims reserves. In particular HIH focused upon FAI's international liability

'problem portfolios' which had been so seriously under-reserved in FAI's accounts since at least 1997.

14.7.2 Martin's work on FAI's 31 December 1998 reserves

More work was undertaken later by Martin and others at HIH to value the full extent of FAI's under-reserving at the time of the takeover. This work also covered the post-takeover deterioration in FAI's reserves for the purpose of calculating HIH' s reinsurance recoveries under the Hannover and Swiss Re reinsurance contracts.

Martin gave evidence920 of the work that he did between September 1999 and December 2000 as financial controller - international and group reinsurance for HIH. In that capacity, Martin was one of the five financial controllers reporting to William Howard, then the general manager of the financial services division of HIH. Martin's role included analysing and reporting the performance of HIH's international divisions, accounting for HIH's group reinsurance contracts, reviewing group management reporting and various one-off projects, including the Allianz joint venture.92 1 In relation to HIH's accounting treatment of the Hannover

reinsurance contracts for the year ending 30 June 2000, Martin gave evidence that:

238

A substantial deterioration in reserves for claims incurred on a date prior to 31 December 1998 occurred in the period ending 30 June 2000.922

The impact of the FA! acquisition

Due to this deterioration, $110 million of undiscounted claims recoveries were processed by HIH in relation to the two Hannover reinsurance contracts for the year ending 30 June 2000. Martin' s evidence was that the $110 million:

was primarily attributable to a deterioration in FAI and overseas claims reserves. The calculation of FAI reserves came both from FAI's AEGIS system and the stand-alone IRIS system which Peter Thompson's RMR Division used (which recorded FAI runoff/problematic claims such as MIPI, ALAS, ALAC, MDU and Rutty).923

Martin produced924 the final version of a paper which was originally drafted by Simpson925 in respect of recoveries claimed by HIH under the Hannover reinsurance contracts for the year ending 30 June 2000. Martin's paper identified the following deterioration in F AI' s pre-takeover reserves:

Since 31 December 1998, FAI Australia had reported adverse development of $245M. FAI International had reported adverse development of $108M. Further Peter Thompson has set IBNR's of $63M for F AI International.

F AI Inwards had deteriorated by $40M since 31 December 1998. On top of this a $40M general reserve has been taken up for F AI issues including Agnew and a prudential reserve ... There has been claim losses from the UK F AI 1236 syndicate of $69 .SM for the twelve months to 30 June 2000. This resulted in losses of $36.1 M which was predominantly prior year

losses. 926

FAI's reserves had thus deteriorated by $532 million since 31 December 1998.

Martin also gave oral evidence about this report. 927 He confirmed that he was satisfied in making this report that, as at 30 June 2000, there had been sufficient adverse development in the F AI and HIH reserves since 31 December 1998 to justify taking up all of the recoveries under the Hannover reinsurance contracts

which were then proposed. 92 8 He confirmed that his opinion was that the total deterioration of F AI' s 31 December 1998 reserves to 30 June 2000 was more than $500 million. 92 9 There was no challenge made by any party to Martin's evidence of the $532 million deterioration of FAI's 31 December 1998 reserves. I accept Martin's uncontradicted evidence about that reserve deterioration.

Martin calculated the $532 million deterioration on an undiscounted basis. That was appropriate in my view because the majority of the reserves which he analysed were for F AI run-off portfolios and all of the F AI profitable lines were sold into the Allianz joint venture shortly after Martin completed his analysis in October 2000. Therefore there was no basis for a discounted (or going concern) assessment of FAI's reserves for 30 June 2000 or thereafter.

14.7.3 Evidence of Cassidy

In his witness statement, Cassidy also addressed the deterioration of HIH's outstanding claims liabilities across various portfolios.93 0 In relation to Martin's assessment of the F AI pre-takeover claims reserves, Cassidy said:

Th e failure ofHJH Insurance 239

As a final comment, this assessment of actuarial outcomes does not appear to me to be inconsistent with the quantification of 1998 and prior (undiscounted) claims development subsequently reported by HIH by June 2000 and provided by Robert Martin in Annexure 2 of his Statement (WITS.0037.0034). My analysis of the portfolios of business reporting this development and the consulting actuaries who had previously assessed those portfolios, is as follows :

(a) F AI $532 million PWC

931

In hi s closing submissions, Cassidy referred back to this evidence, putting the submission that these adverse developments in FAI's outstanding claims liabilities were contributing factors in the failure of HIH. 932 Again, no party challenged Cassidy's evidence or submissions in these respects. I accept Cassidy's evidence as confirming Martin's F AI reserve deterioration analysis.

14.7.4 Andersen's February 2001 review of FAI reserve deterioration

Martin's assessment of F AI' s 31 December 1998 reserve deterioration was also confirmed by audit work undertaken by Andersen. On 15 February 2001 Alan Docherty sent Pye, John Fanning and Sandor Helby an email entitled ' F AI Unders and Overs ' .933 Docherty' s email stated that the unders and overs constituted a

'reconciliation ofF AI June 1998 audited net assets, back to approximated F AI Sub­ Consolidated Net Assets as at 31 December 2000. ' This analysis showed total deterioration for F AI reserves as at 31 December 1998 of $532 million.

It does not appear that the 'FAI Unders and Overs' exercise described above was concluded by Andersen because HIH went into provisional liquidation one month after Docherty' s email. However it is noteworthy that in relation to the realisation of the assets which were accounted for in the F AI balance sheet as at 30 June 1998, Docherty arrived at a figure that was practically a break-even. He noted a loss of $50 million in relation to the sale of Oceanic Coal ($125 million received on sale, against a 31 December 1998 book value of $ 175 million). This loss was offset by gains on the sale ofF AI Life ($26 million profit over book value of $68 million), Godfrey Pembroke (estimated at $10 million) and the St Moritz Hotel (said to be $30 million). Also offset was a write-down (loss) in the value ofHSI of $20 million, a figure that Docherty said required checking. These F AI asset recovery figures are addressed further below.

14.7.5 The KPMG phase II report

The KPMG phase II report934 provides an assessment of the claims reserves for the HIH group as at 15 March 2001 , the date of the provisional liquidation. The phase II report suggests that the $532 million adverse development of FAI's outstanding claims liabilities as at 31 December 1998 was probably a conservative estimate of that adverse development at the time of Martin's analysis.

240 Th e impact of the FAI acquisition

The phase II report identified total HIH under-provisioning of approximately $ 1.9 billion; that is, a $1.9 billion difference between the outstanding claims liabilities in HIH's draft accounts as at 31 December 2000 and KPMG's estimate of those liabilities on a discounted going concern basis as at 15 March 2001. HIH's OCP as at 31 December 2000 included the $532 million of FAI reserve deterioration which Martin had calculated earlier in 2000. It is most unlikely that any significant

part of the $1.9 billion shortfall represented adverse development in the ten weeks between 31 December 2000 and 15 March 2001 . Rather, the likelihood is that a substantial part, if not the whole, of the $1.9 billion difference ought to have been brought to account in HIH's OCP as at 31 December 2000.

A reasonably accurate estimate can be made of the extent to which the additional $1.9 billion of claims reserves arrived at in the phase II report is attributable to portfolios of business acquired by HIH from F AI. This estimate can be made by comparing the portfolio breakdown schedule setting out the outstanding claims

liabilities in HIH's draft accounts as at 31 December 200093 5 with the portfolio breakdown of the KPMG estimate in the phase II report. The results of this comparison ofFAl portfolios are summarised in Table 14.1.

Table 14.1 Comparison of FAI portfolios: HIH and KPMG

($million)

HIH reserve as at KPMG reserve as at

FAI portfolio 31 December 2000 936 15 March 2001 Difference

Professional indemnity 74.2 164.1 89.9

Liability 95.9 164.1 68.2

Workers compensation 71 .6 81 .1 9.5

Compulsory third party 713.8 638.0 (75.8)

Marine (0.3) 24.6 24 .9

Short-tail 8.1 28.5 20.4

Lloyd 's 27.5 38.3 10.8

Asia 23.0 19.2 (3.8)

Run-off 172.3 207.7 35.4

MDU 49.8 35.1 (14.7)

FCHC (5 per cent) 0.0 70.0 70.0

Total 1 235.9 1 470.7 234.8

The comparison in the table suggests that $235 million (the difference between the figures of $1471 million and $1236 million) of the total HIH reserve shortfall was attributable to F AI portfolios. This does not include any allowance for bad reinsurance debts or overstated reinsurance recoveries on the HIH or F AI contracts.

Several of the FAI portfolios (for example PI, Liability, Lloyd' s Syndicate 1236, and 'Run-Off) were in run-off at the time of the takeover. The total of the reserve shortfalls on those portfolios calculated in the phase II report is $204.3 million. Credit has to be given for the positive developments which appear in the F AI portfolios for CTP, Asia and Medical Defence Union. If the positive developments

Th e failure ofHIH Insurance 241

in those portfolios are set off against the negative developments in the run-off portfolios, a net reserve shortfall of $110 million is arrived at. If a 5 per cent allowance for future claims handling costs is made in respect of the run-off portfolios $63 million would be added. Therefore the total of the reserve shortfall and the allowance for future claims handling costs is $173 million.

Unlike the Martin analysis, the phase II report was not limited to the deterioration in FAI's outstanding claims liabilities on business written prior to 1 January 1999 (the effective date of HIH's takeover of FAI). If only the $173 million shortfall calculated above was included this would be likely to understate the true cost to HIH of under-reserving and adverse development in the F AI portfolios. This is

because neither Martin nor the phase II analysis took into account any payments which were made after 31 December 1998 in respect of under-reserved claims in any of the FAI portfolios. Following the takeover, those payments were made out of the HIH group ' s assets and thereby reduced the outstanding claims provisions required in the relevant F AI portfolios. It was not possible for the Commission to value accurately the claims payments made by HIH on the F AI portfolios after the takeover. Accordingly I have made no allowance for such payments.

14.7.6 Adler's submissions on the phase II report

In his submissions on the causes of the failure of HIH, Adler estimated 'that the contribution of F AI business written within the HIH portfolios to the total reserve shortfall identified by KPMG is $261.6 million, or 13.7 per cent of HIH's total reserves shortfall as at the date of its failure. ' 937 This is nearly $90 million more than I have calculated. As I have explained above, this is not a precise science. Adler submitted that 'at least some of this figure is likely to be attributable to deterioration in claims in the FAI portfolio after the date of FAI's takeover by HIH'.938 Adler appeared to allude to Martin's analysis939 and submitted that a figure of $500 million of under-reserving within FAI as at 31 December 1998 should be rejected. It is not clear to what Adler's submission actually refers because the $532 million arrived at by Martin was not a calculation ofF AI under-reserving as at 31 December 1998. Rather it was a valuation of the subsequent adverse development of F AI' s claims reserves as at that date, as the evidence considered above makes clear. In effect, Adler's submissions ignored the $532 million of adverse reserve development in FAI, which Martin' s analysis showed had already been taken into account in the valuation of HIH's reserves as at 31 December 2000 and which did not form part of the $1.9 billion deficiency identified in the phase II report.

Adler's supplementary submission940 explained his methodology for calculating what he submitted was F AI' s overall contribution to the HIH reserves shortfall identified in the phase II report. No attempt was made in Adler's analysis to distinguish those F AI reserves which had deteriorated since 31 December 1998 (as Martin had done) from the overall shortfall ofHIH's reserves calculated in the phase II report. Whilst Adler's calculation of an F AI reserve shortfall of $262 million might be accepted as an estimate arising from the phase II report, it is inadequate to

242 The impact of the FA/ acquisition

assess the true financial cost of the F AI takeover to HIH. This is because it ignores both the $53 2 million of adverse F AI reserve development after 31 December 1998 valued by Martin and the cost of F AI claims payments made by HIH after the takeover but before the phase II report.

Limiting the losses known to have been incurred to the deteriorations in the outstanding claims of the FAI portfolios of business written prior to 1 January 1999, it is likely that such deterioration was at least $705 million. I have arrived at this figure by adding the $532 million identified in the Martin analysis (which was included in HIH's 31 December 2000 valuations shown in the table above) to the further $173 million I derived above from the phase II report (which was not included in HIH's 31 December 2000 reserves). The value of payments made by HIH on under-reserved claims in PAl's portfolios after 1 January 1999 would need to be added to the $705 million to arrive at the total overall cost to HIH. However, the Commission has not been able to calculate the value of the paid claims and so I have not taken it into account. I therefore consider the $705 million figure for the F AI reserves shortfall to be appropriately conservative.

14.7.7 The realisation of FAI's investments and other assets

Cassidy's witness statement941 explained how HIH's investment committee managed HIH's investments, including those which it acquired as a result of the FAI takeover. Cassidy's statement described the role of the HIH investment committee and the establishment and implementation of its investment guidelines. Cassidy's evidence was that:

investment policies were originally established after taking account of [HIH's] investment obiective of maximising returns on assets within prudent risk parameters. 42

As Cassidy observed:

the F AI acquisition brought changes to the way HIH managed its investments. FAI's investment portfolio was heavily weighted towards property investment and was largely inconsistent with the direction being pursued by the Group's Investment Committee. In addition, the F AI portfolio included substantial exposure to associated company investments which lead to some early confusion as to which F AI assets would be

consolidated with those of HIH for Investment Committee reporting purposes. 943

FAI's 'associated company investments' included FAI Life, Godfrey Pembroke, F AI Finance, HSI and Oceanic Coal. All of these investments, which HIH reported under the caption 'non-core assets', were the responsibility of Howard. He was also responsible for FAI's property, commercial loan and mortgage assets and its

un listed investments. FAI's listed equities and fixed interest assets came under the control of John Ballhausen who managed all similar HIH investments. 944

The failure of HIH Insurance 243

Howard reported to Fodera and Williams in relation to the management of the FAI assets which he controlled. Cassidy's evidence was that:

exit strategies had been put in place with regard to many of these investments. In addition, a separate committee was formed to monitor the progress of the strategies adopted in respect of the F AI property investments. 945

Cassidy's statement also provided a general overview of HIH's asset allocation practices and the internal and external asset, equity and funds management arrangements which applied. There was added complexity arising from the management arrangements for several F AI assets, in particular the F AI property assets, including the St Moritz Hotel and the Emu Brewery development. 946

14.7.8 Oceanic Coal and the St Moritz Hotel

As Andersen observed in its presentation to the HIH audit committee for the year ending 30 June 1999:

the total loss on the sale of OCAL was $70 million of which $20 million has been recognised as pre-acquisition (in goodwill) and $50 million as an extraordinary loss. 94 7

There was a separate management paper explaining why HIH considered that there was justification for recognising the loss on the sale of OCAL as an extraordinary loss. Andersen concurred with this accounting treatment.

The $70 million loss on the sale of OCAL in 1999 was offset to some degree in the following year by the booking of an upward revaluation of $30 million in respect of HIH's interest in the St Moritz hotel. A paper prepared by Frank Wolfl48 explained why in his view it was appropriate to increase the valuation of St Moritz from US$136 million to US$153 .5 million, being an increase of US$17 .5 million (A$30 million). This revaluation was adopted by HIH.

14.7.9 Shares in One.Tel

Michael Maroney's report to the HIH investment committee (Howard, Ballhausen and others) of 21 February 2001 949 illustrated the difficulty of attributing profits and losses from movements in the market values of F AI' s listed securities held for its general insurance purposes and marked to market under AASB 1023. For example, Maroney's schedule of movements in the value of FAI's One.Tel shareholding950 showed that as at 31 December 1998 FAI owned 9.8 million One.Tel shares. This holding had an historical cost of $1.2 million and a market value of $58.7 million at that time. Following the takeover and before 30 June 1999, One.Tel made a bonus issue pursuant to which HIH received 39.6 million additional shares in One.Tel. During the same period HIH sold One.Tel shares valued at $49.3 million. Taking

into account a decline in One.Tel's share price during this period, HIH booked a total gain in respect of its One. Tel shareholding of $34.2 million for the six months to 30 June 1999. HIH sold further One.Tel shares in the six months to

244 The impact of the FA! acquisition

31 December 1999, at which date it held 25 million One.Tel shares. Realised and

unrealised gains in respect of that holding were $56.4 million as at

31 December 1999. However a substantial decline in One.Tel's share price in early 2000 eliminated most of the unrealised gains recorded as at 31 December 1999 so that realised and unrealised gains as at 30 June 2000 were reduced to $25.4 million. A further decline in One.Tel's share price meant that by 10 November 2000 the cumulative total of all realised and unrealised gains on the One.Tel holding was $53.7 million. HIH's remaining shareholding in One. Tel lost all its value when One.Tel went into liquidation shortly afterwards.

14.7.10 Home Security International Inc (HSI)

Maroney's report951 also documented the write-down of the value of FAI's shareholding in HSI. from $4.9 million to $1 million progressively during the course of 2000. A number of other write-downs of F AI - related assets were referred to in Maroney' s report.

External valuations undertaken by HIH of a number of the 'non-core' FAI assets as at 30 June 1999 indicated that they were not materially under or over-valued at the date of the takeover. They included the St Moritz Hotel, the Emu Brewery site and OCAL.

HSI was an exception to this rule. As at 31 December 1998, it was valued in FAI's books at US$11.25 per share. 952 HSI's shares spiked up in December 1998. The price rose from a low of US$7.75 on 1 December 1998 to a peak of US$11.25 on 31 December 1998, being its highest price that month and reached on the last trade

of the year. The price immediately declined on the next trading day

(4 January 1999) to US$10.50 and had further declined to US$9.38 by the end ofJanuary 1999. HSI never regained the share price peaks of 1998 and it steadily declined in value thereafter.

FAI's shareholding in HSI as at 31 December 1998 was valued at $39.6 million on the basis of the US$11.25 price reached on the last trade in 1998.953 Less than two years later, on 8 December 2000, HIH sold its entire shareholding in HSI to Bradly Cooper for $1.25 million.

The history of the investment by F AI, and then HIH, in HSI is dealt with in detail in Chapter 22 . In broad terms the conclusion reached is that there was a loss of value on the asset acquired on takeover of approximately $38 million, with a further $47 million in cash having been injected for little or no commercial benefit.

14.7.11 Project In House

I.n January 2001 Gregory Waters, HIH's internal actuary, presented his report

entitled 'Project In House' .954 A number of transactions undertaken and accounting treatments adopted by F AI prior to the takeover were described in Waters ' report. Section 3 of his report was entitled 'Commercial Loans' .955 Waters noted that the

The failure ofHJH Insurance 245

source for this section of the report was an HIH internal audit report prepared by his department dated 23 July 1999 following a thorough review of F AI commercial loans. 956

Waters noted that at the time of the F AI acqUisition there were a number of commercial loans in the F AI portfolio which were in default or in arrears. Furthermore inadequate provisions had been made for them in the accounts ofF AI. Waters described F AI' s loan approach as high-risk, not prudentially managed, poorly documented and without adequate follow-up.

Waters noted that the FAI commercial loans which were in default at the time of his internal audit report dated 23 July 199995 7 totalled $56.1 million. As a result he concluded that a further $33.65 million of provisions were required. A number of these loans was analysed and reported on in the Project In-House report. Many were

made to companies which had been liquidated years earlier or which were no longer on the ASIC register of companies and which, in Waters's view, ought to have been written off at the time of the F AI acquisition.

Section 3.2 of the Project In House report analysed loans which were recorded in the F AI general ledger but not in its commercial loans accounting system. These totalled $33.3 million, of which loans approximating $7 million were in default but had not been provided for.

Elsewhere in the report there were many references to F AI loans, investments and other transactions which Waters considered to be suspicious, undesirable or insecure for various reasons. No allowances for bad or doubtful F AI debts arising out of these transactions appear to have been made in HIH's calculations of the goodwill on the acquisition of F AI, nor were they addressed in any of the parties' submissions on the value ofFAl to HIH.

14.7.12 Summary of approach to FAI assets

Having regard to the evidence analysed above I consider that it is inappropriate to take into account any changes (whether positive or negative) in the values ofFAl's 31 December 1998 assets after the takeover by HIH. There is no clear evidence that the overall value of F AI' s assets was materially misstated at the time of the takeover. Unlike the subsequent F AI reserves deterioration, neither a subsequent increase nor a decrease in the value of those assets can be attributed sensibly to F AI. Instead, responsibility for losses and credit for gains post-takeover should go to HIH. It alone had management responsibility for all the FAI assets following the takeover and the ability to exploit the relevant market conditions at the time of sale. To suggest otherwise is to imply that the management of HIH had no active role in reviewing the carrying value of the F AI assets or market conditions. It would also imply that HIH should bear no responsibility for losses resulting from the sale of FAI assets (other than where it can be proven that they were materially overvalued at the date of takeover) or take any credit for any subsequent profit on sale.

246 Th e impact of the FA! acquisition

This argument is stronger in relation to those F AI assets which were held for the general purposes of its insurance business, such as the shares F AI held in One. Tel. Those shares were marked to market as at 31 December 1998 in accordance with AASB 1023. Therefore any profit on their subsequent sale was solely attributable to HIH's management who had to determine from time to time what their current worth was and if and when they should be sold.

Having regard to the numerous 'swings and roundabouts' in relation to the valuation of FAI's assets following its takeover by HIH, some of which are described above, I consider it to be reasonable to regard the net realisation value of those assets as break-even. Accordingly I have not attributed any profits or losses in respect of those assets in my calculations.

14.7.13 Value of the FAI businesses sold in the Allianz transaction

An amount of$125 million was claimed by HIH following the exercise of its put option in respect of its share of the Allianz joint venture. That amount has not been paid and the liquidators and Allianz remain in dispute about the amount that should be paid by Allianz for that share. Adler's submissions focused on the $400 million

'valuation' performed by Deutsche Bank in October 2000 (as to which see Chapter 17), but it is clear that HIH will not receive more than $325 million, being the $200 million paid by Allianz into the joint-venture claims reserve trust and possibly an additional $125 million in due course.

The FAI lines of personal business no doubt represented the major part of HIH' s contribution to the Allianz joint venture. Those F AI lines were assumed by HIH for goodwill write-down purposes to be 86 per cent of the value contributed by HIH to the joint venture. It is not possible to say whether the management ofHIH enhanced the value of the F AI personal lines businesses after the takeover ofF AI and prior to the Allianz joint venture.

14.7.1 4 Bottom line

If the realisation values of all F AI assets acquired upon takeover are treated as 'break-even' (which for the reasons give above I consider to be a fair assessment), then my estimate of the directly quantifiable loss suffered by HIH due to the acquisition ofFAl is as shown in Table 14.2.

Th e failure of HIH Insurance 247

Table 14.2 Estimate of directly quantifiable loss of HIH due to the FAI takeover

Item

HIH's total enterprise value acquisition cost (cash and shares)

Add FAI 's audited net assets as at 30 June 1998

Less FAI 's June-Dec 1998 trading loss

Less fair market value write-down of OCAL made upon acquisition

Less overstated NI/GCR recoveries and understated future premiums in FAI's 30/6/98 net assets

Less FAI pre-takeover reserve deterioration to 31 December 2000

Less phase II estimate of reserve deterioration to 15 March 2001

Add highest value of Allianz joint-venture share (86 per cent of $400M)

Total

Amount ($m)

(300)

224

(67)

(20)

(67)

(532) (173) 344

(59 1)

To err on the side of caution I have not included in these calculations any part of the premiums paid by HIH for the Hannover and Swiss Re reinsurance policies. Those premiums are due in part to the losses arising from the F AI reserves deterioration. Nor have I attributed any cost to the adverse cash flow impact of the Allianz transaction. Instead, I have credited F AI with $344 million of value arising from that transaction despite the fact that only $200 million has been received to date by HIH. There may also be, no doubt, a measure of double-counting in relation to the Allianz transaction because Adler identified that the book value of the F AI personal lines business was $30 million at the time of the takeover. 95 8

I have also ignored any potential future income tax or other tax benefits (such as franking credits) because in the events which transpired, I considered them too remote and of no realisable value. The reserves referred to above have been calculated in some cases on a discounted and in other cases an undiscounted basis. Given that HIH was not a going concern by the time of the phase II report, an

undiscounted (non-going concern) basis of reserve calculation is appropriate.

I have also ignored all integration costs following the takeover, although they were calculated at $55 million. 959

14.8 Possible contraventions and referrals

In this section I set out the findings that I have made about matters that might have been a breach of the law in relation to the subjects considered in this chapter. I also note those matters that, in my opinion, should be referred to ASIC for further consideration.

14.8.1 Arising from Section 14.1.4

Boulden Section 14.1.4 discusses Boulden's conduct in relation to the forwarding of data to PricewaterhouseCoopers actuaries in late 1997.

248 Th e impact of the FA! acquisition

The question that arises 1s whether, in the circumstances referred to in Section 14.1.4, Boulden might have contravened s. 590(1 )( c )(iii) of the Corporations Law. If, as I have found, Boulden:

• gave, or was privy to the giving of, data that did not include genume case estimates to PWC

• knew the data did not include genuine case estimates

and

• knew that PWC would rely on the data and might as a result provide an estimate that seriously undervalued outstanding claims liabilities

then, in my view, there might have been such a contravention.

The omission of genuine case estimates totalling $112 million in the data forwarded to PWC is a serious matter. It had an obvious impact on reported profit as at 31 December 1997 and on F AI' s solvency requirements pursuant to the Insurance Act 1973.

I recommend that this matter be referred to ASIC for consideration of whether proceedings should be instituted against Boulden.

Wilkie A separate question arises in connection with Wilkie's conduct in relation to PWC's interim valuation at 31 December 1997. I have already found that Wilkie was aware of the level of under-reserving within F AI, as identified by Boulden in early December 1997. That information was passed to him by Kamha. A question arises as to whether Wilkie's conduct, in omitting from his 19 February 1998 report to the board the information passed to him by Kamha in early December 1997,

might contravenes. 1309(1) of the Corporations Law. If, as I have found, Wilkie:

• furnished information to the board-namely, estimates of outstanding claims liabilities that included reserves for the professional indemnity portfolio

in circumstances where the information was false or misleading because it did not include Boulden's estimates

and

• knew that the information was false and misleading

then in my view, there might have been a contravention ofs. 1309(1). Alternatively, there might have been a contravention of s. 232(2).

I recommend that this matter be referred to ASIC for consideration of whether proceedings under s. 1309(1) or civil penalty proceedings under s. 232(2) should be instituted against Wilkie.

The failure of HIH Insurance 249

Adler

I have also found that Adler became aware of the potential for significant under­ reserving within F AI in September or October 1997 as a result of a conversation with Kamha. Adler asked Kamha to prepare a report and to come back to him in relation to the matter as soon as possible. Soon after the conversation Wilkie was appointed chief operating officer of F AI and assumed responsibility for both the general insurance division and the commercial and professional insurance division. In the months that followed, Adler did not ask Kamha or Wilkie what investigation had been made, what was the progress of the investigation, what was the magnitude or nature of the problem, how it was to be dealt with, how the problem arose and why it had to wait until February 1998 before it could be dealt with by an actuary. Further, Adler did not in any of his reports to the board in late 1997 or early 1998 make reference to the potential under-reserving.

On the basis of those findings, it is my opinion that Adler might have breached s. 232(4) of the Corporations Law because, in failing to make the inquiries just referred to, he did not exercise the degree of care and diligence as an officer ofF AI that a reasonable person in a similar position would have exercised.

As noted, it is important to ensure that at all levels the processes whereby outstanding claims liabilities are estimated are attended by attention to detail. It is also most important that the board is informed of all relevant matters that may affect the adequacy of reserves.

I recommend that the matter of Adler's conduct in this regard be referred to ASIC for consideration of whether civil penalty proceedings should be instituted against him.

14.8.2 Arising from Section 14.1.9

Wilkie

Section 14.1.9 deals with the circumstances surrounding the supply of data to PricewaterhouseCoopers to enable it to prepare its actuarial valuations as at 30 June 1998. In my opinion, Wilkie's conduct ensured that the data was incomplete. In that context, the question that arises is whether Wilkie might have contravened s. 232(2) of the Coq)orations Law. If, as I have found, Wilkie:

• issued to McCarthy and Trahair a direction that they were not to discuss the known under-reserving with PWC

and

• directed that Kamha would be the F AI contact point with PWC in relation to any questions concerning the data

250

in circumstances where he knew the data forwarded to PWC did not contain genuine case estimates and he knew, or should have known, that it was thus likely PWC would underestimate IBNR ·

Th e impact of the r-:41 acquisition

then, in my view, there might have been such a contravention.

It is a most serious matter when a senior executive of an insurance company misleads the external actuaries by fail ing to disclose to them the omission of genuine case estimates from data delivered to them for the purpose of enabling them to prepare valuations.

I recommend that this matter be referred to ASIC for consideration of whether civil penalty proceedings or criminal proceedings should be instituted against Wilkie.

14.8.3 Arising from Section 14.1.10

The state of affairs in relation to the failure to disclose the reserve shortfalls to the F AI board is dealt with in Section 14.1.1 0.

Adler

The question that arises is whether, in the circumstances described, Adler might have contravened s. 1309(1) of the Corporations Law. If, as I have found, Adler:

• was aware as at March 1998 from his discussion with Wilkie that there was a shortfall in reserves of between $50 million and $60 million

and

• did not report that matter to the board during the period March-June 1998

in circumstances where he knew that by omitting such information from his reports to the board those reports would be rendered misleading in a material particular

then, in my view, there might have been such a contravention. Alternatively, there might have been a breach of s. 232(2).

Again, I regard it as most important that all steps in the processes leading to the estimation of outstanding claims liabilities be followed conscientiously and with attention to detail. Given the importance of provisions within the accounts of a general insurer, it is vital that the board receive the most accurate information relating to reserves and any information that might suggest significant reserve shortfalls.

I recommend that this matter be referred to ASIC for consideration of whether proceedings under s. 1309(1) or civil penalty proceedings under s. 232(2) should be instituted against Adler.

Mainprize

A further question that arises is whether in the events that occurred Mainprize might have contravened s. 1309(1) of the Corporations Law. If, as I have found, Mainprize:

The failure of HIH Insurance 251

• was aware of a significant shortfall in reserves in the period March- June 1998

and

• failed to report the matter to the board

in circumstances where he knew that by failing to report the matter to the board his reports were false or misleading in a material particular

then, inmy view, there might have been such a contravention. Alternatively, there might have been a contravention of s. 232(2).

I recommend that this matter be referred to ASIC for consideration of whether proceedings under s. 1309(1) or civil penalty proceedings under s. 232(2) should be instituted against Mainprize.

Wilkie

The same question arises in relation to Wilkie.

He too might have contravened s. 1309(1) of the Corporations Law, against the background described in Section 14.1.1 0. If, as I have found, Wilkie:

• was aware of under-reserving in the order of $100 million since early March 1998, following his conversation with Kamha

and

• failed to refer to such a matter in his reports to the board during the period March- June 1998

in circumstances where he knew that his reports were false or misleading in a material particular'

then, in my view, there might have been such a contravention. Alternatively, there might have been a contravention of s. 232(2).

I recommend that this matter be referred to ASIC for consideration of whether proceedings under s. 1309(1) or civil penalty proceedings under s. 232(2) should be instituted against Wilkie.

14.8.4 Arising fro m Section 14.1 .11

As discussed in Section 14.1.11, in the months following 30 June 1998 Wilkie ' s reports to the board made no reference to under-reserving within F AI. The question that arises is whether, in these circumstances, Wilkie's conduct in omitting from his reports to the board between July and December 1998 his knowledge of significant under-reserving rendered his reports fa lse and misleading. In my opinion, there is sufficient evidence to say that he was aware his reports were false and misleading, such that his conduct might have contravened s. 1309(1) or s. 232(2) of the Corporations Law.

252 Th e impact of the FA! acquisition

Similar matters are, however, fully canvassed in other referrals. On this occasion I do not recommend that the matter be referred.

14.8.5 Arising from Section 14.2

Wilkie: negotiation of the GCRA AXOL contract Given the findings I have made concerning Wilkie's involvement in the negotiation of the General Cologne Re Australia aggregate excess-of-loss contract, a question arises as to whether Wilkie might have contravened s. 232(2) of the Corporations Law. If, as I have found, Wilkie:

• was involved in the negotiations that resulted m F AI entering into the transaction with GCR

and

• caused F AI to enter, or facilitated its entry, into the transaction

where, on a best-case outcome for F AI, the transaction was financially detrimental to F AI in that it involved the payment of more premium and fees than possible recoveries, did not involve a transfer of risk, could not confer any accounting benefits on F AI unless the auditors were misled, and did not confer any legitimate commercial benefit on F AI

and

in circumstances where Wilkie knew that it was financially detrimental to FAI

then, in my view, there might have been such a contravention.

Section 232 can be contravened only by an 'officer' of a corporation. Wilkie was a director of F AI General Insurance Company Limited and, given the nature of his duties as chief operating officer, there is a reasonable argument he would satisfy the definition of' executive officer' in relation to F AI Insurances Limited.

The matters to which I have referred are serious. If they are made good, they represent a serious breach of the trust placed in Wilkie by virtue of the position he held as a senior executive ofF AI.

I recommend that this matter be referred to ASIC for consideration of whether criminal charges or civil penalty proceedings, or both, should be brought against Wilkie in respect of this matter.

Wilkie: the F AI accounts and the APRA return A question that arises is whether Wilkie might have contravened the Corporations Law in his dealings with the auditor and the board in relation to the first and second GCRA AXOL contracts. If, as I have found, Wilkie:

The failure of HIH Insurance 253

• in meetings with Scrivens on 21 and 26 August 1998, furnished to Scrivens information about risk transfer under the contracts or about their accounting treatment and made available to Adler and the board generally information about reinsurance contracts

• deliberately refrained from disclosing to Adler, the board and the auditor ofF AI the six 1 May 1998 contracts, their connection with the second AXOL contract, the 1 May 1998 side letter and the 26 June 1998 side letter

· where, by reason of that omission, the information was false and misleading

and

in circumstances where he knew that

they were all highly relevant to an assessment of whether there was a transfer of risk under the arrangements between F AI, GCRA and ultimately whether F AI' s proposed accounting treatment was appropriate

and

the information was false and misleading

then, in my view, there might have been a contravention of s. 1309(1) or s. 232(2) of the Corporations Law. Alternatively, if Wilkie took no reasonable steps to identify and rectify the omission and thus to ensure that the information was not false and misleading then he might have breached s. 1309(2) or s. 232(4).

I recommend that this matter be referred to ASIC has for consideration of whether proceedings, or both, under s. 1309(1) or s. 1309(2), or civil penalty proceedings under s. 232(2) or s. 232(4) should be instituted against Wilkie.

Further, if it be the case, as I have found, that in signing the APRA return in July 1998 Wilkie intentionally provided a false answer to the question just described­ and if he did so, as I consider he did, in order to divert APRA' s attention from the true terms of the GCRA policies and especially from the side letter- then, in my view, there might have been a breach of s. 232(2) of the Corporations Law or s. 128(2) of the Insurance Act 1973.

I recommend that this matter be referred to ASIC and APRA for consideration of whether proceedings under s. 128(2) of the Insurance Act or civil penalty proceedings under s. 232(2) of the Corporations Law should be instituted against Wilkie in relation to his signing of the APRA return.

I have also found that Wilkie falsely advised Scrivens on 21 August 1998 that F AI entered into the NI contract to protect against ' future volatility' when it was actually done to obtain an immediate recovery. I do not propose to refer any question arising out of this conduct.

254 The impact of the F AI acquisition

Mainprize: negotiation of the GCRA AXOL contract

I pose the same question and reach the same conclusion in relation to Mainprize as I do concerning Wilkie and the negotiations concerning the GCRA AXOL contract. Mainprize might have breached s. 232(2) of the Corporations Law.

The matters I have found are serious. If they are made good, they represent a serious breach of the trust placed in Mainprize by virtue of the position he held as a director ofF AI Insurances Limited and as its highest ranking financial officer.

I recommend that this matter be referred to ASIC and APRA for consideration of whether proceedings under s. 128(2) of the Insurance Act or civil penalty proceedings under s. 232(2) should be instituted against Mainprize.

Mainprize: the FA/ accounts

If, as I have found, Mainprize:

• in meetings with Scrivens on 21 and 26 August 1998, furnished, or permitted to be furnished, to Scrivens information about risk transfer under the contracts or about their accounting treatment

• deliberately refrained from disclosing to the auditor ofF AI the six 1 May 1998 contracts, their connection with the second AXOL contract, the 1 May 1998 side letter and the 26 June 1998 side letter

where, by reason of that omission, the information was false and misleading

in circumstances where he knew that

they were all highly relevant to an assessment of whether there was a transfer of risk under the arrangements between F AI, GCRA and ultimately whether F AI' s proposed accounting treatment was appropriate

and

the information was false and misleading

then, in my view, Mainprize might have contravened s. 1309(1) or s. 232(2) of the Corporations Law. Alternatively, if Mainprize took no reasonable steps to identify and rectify the omission and thus to ensure that the information was not false and misleading then he might have breached s. 1309(2) or s. 232( 4).

I recommend that this matter be referred to ASIC for consideration of whether proceedings under s. 1309( 1) or s. 1309(2), or civil penalty proceedings under s. 232(2) or s. 232(4) should be brought against Mainprize.

If, as I have found, Mainprize signed the F AI accounts for the year ending 30 June 1998 when he knew they contained a false entry, being the amount of recoveries booked under the GCRA contract, and the entry was material to the reported profit, he might have contravened s. 1308(2) or s. 232(2) of the Corporations Law.

The failure of HIH Insurance 255

I recommend that this matter be referred to ASIC for consideration of whether proceedings under s. 1308(2) or civil penalty proceedings under s. 232(2) should be brought against Mainprize for signing off on the accounts.

Peiris Counsel ass1stmg submitted that Peiris deliberately failed to advise the audit committee meeting on 3 September 1998 that Andersen's estimate of the future guaranteed accounting loss as being $23.65 million assumed that the earthquake trigger of the NI contract had occurred.

I am not satisfied that Peiris' conduct meets the threshold for a possible contravention. There was insufficient evidence to form the view that he was an officer for the purposes of s. 232 of the Corporations Law. Otherwise, there is no evidence that in his dealings with the audit committee he 'furnished information' to them such that the question of a possible contravention of s. 1309(1) or s. 1309(2) could be said to arise.

Burroughs In relation to Burroughs I pose the same question as I have in relation to Wilkie concerning the negotiation of the GCRA AXOL contract-allowing for the fact that Burroughs was only involved in the negotiation of the first AXOL contract. By his actions in preparing and transmitting material for and to GCRA, and with the knowledge he possessed as to the purpose of the transaction, Burroughs facilitated F AI' s entry into the transaction. I am doubtful whether, as reinsurance manager, Burroughs fell within the definition of 'executive officer' at the relevant time and I make no finding that he might have contravened s. 232(2) of the Corporations Law as a principal offender. But his conduct might have constituted a knowing involvement in Wilkie's breach of s. 232(2).

I recommend that this matter be referred to ASIC for consideration of whether civil penalty proceedings should be brought against Burroughs.

General Cologne Re Australia In the light of the findings I have made, a question that arises is whether GCRA might have contravened s. 232(2) of the Corporations Law. If, as I have found , GCRA- through its officers Smith, Byatt and Self, to the knowledge of its managing director, Barnum, and with the assistance of its holding company GCR­ proposed, negotiated and documented the transaction with F AI in circumstances where, either by one or more of its officers, it was aware that:

• the transaction was being pursued by FAI so that it could, if it so chose, stage an increase in its reserves over time because of historic under-provisioning- that is, so that F AI could disguise its past and future breaches of the accounting standards

256 Th e impact of the FA! acquisition

• although one element of the transaction involved the AXOL contract, which on its face suggested risk transfer to GCRA, overall the transaction did not involve any risk transfer

• the transaction was documented for the purpose of enabling the officers ofF AI to present only the AXOL contract to the auditor of F AI and hide those elements that removed risk transfer

• F AI proposed to account for the AXOL contract as a contract of reinsurance by booking a substantial profit for the year ending 30 June 1998 and that to do so F AI would have to withhold from the auditor the six 1 May 1998 contracts, the fact of their connection with the AXOL contract and the 1 May 1998 side letter

and

• the transaction was financially detrimental to F AI in that it involved the payment of more premiums and fees than total recoveries, with no legitimate commercial benefit being conferred on F AI thereunder

then, in my view, it might be found that GCRA knowingly assisted FAI's executives in not acting in the best interests of the company or in accordance with their duties under the Corporations Law. If that were the case, GCRA might have been knowingly involved in Wilkie's possible contravention of s. 232(2) .

Important concerns arise for the regulatory process-and for the reinsurance industry generally.

I recommend that this matter be referred to ASIC for consideration of whether criminal charges or civil penalty proceedings, or both, should be brought against GCRA.

General Cologne Re

A similar question arises concerning the conduct of GCR, through its officers who formed part of its worldwide 'Alternative Solutions Division'-namely, Ellingsen, Vukelic, Houldsworth and Byrne. For similar reasons, I think GCR might have been knowingly involved in Wilkie's possible contravention of s. 232(2).

I recommend that this matter be referred to ASIC for consideration of whether criminal charges or civil penalty proceedings, or both, should be brought against GCR.

Barnum Between early April 1998 and 30 June 1998 Barnum was aware that the GCRA employees who reported directly to him-Smith, Self and Byatt-were negotiating and documenting the transaction with F AI. He was aware of some aspects of the

transaction. He was not involved in the negotiations to the same extent as Smith, Byatt and Self. He may not have been aware of all the relevant circumstances. He was aware that the negotiations involved the use of side letters, a practice he did not condone. Yet he did nothing to prevent their use.

The failure ofHIH Insurance 257

Barnum's failure to direct GCRA employees to cease negotiating with FAI management a no-risk transaction when he was aware of the matters just noted is regrettable. But if in so doing he failed in a duty imposed on him by the Corporations Law-and I make no finding that there was such a breach- it would be a duty imposed on him as an officer of GCRA. A question could arise as to whether that was within the terms of reference; even if it were, I consider it too remote to justify a referral in the context of this Commission. No finding is made against Barnum in this respect or otherwise in relation to the GCRA AXOL contract.

Ellingsen Ellingsen is a foreign national. He was asked to appear before the Commission to give evidence. He decided not to do so. 960 He did not provide a witness statement. Along with Houldsworth, Ellingsen was the primary architect of the GCRA transaction. For reasons similar to those that I have outlined in relation to GCRA and GCR, Ellingsen might have been knowingly involved in Wilkie's possible contravention of s. 232(2).

I recommend that this matter be referred to ASIC for consideration of whether criminal charges or civil penalty proceedings, or both, should be brought against Ellingsen. In so doing ASIC will no doubt take into account that Ellingsen is a foreign national and the difficulties this would necessarily bring to any proceedings.

Houldsworth Houldsworth's conduct is not relevantly distinguishable from that of Ellingsen. He too decided not to give evidence. But one critical piece of evidence concerning Houldsworth's conduct was hearsay-namely, the evidence that he required the other business to be claims-free. Self told the Commission that Ellingsen said that Houldsworth required it. In view of that, and taking into account that Houldsworth is a foreign national and the difficulties that would bring in relation to proceedings, I decline to make a finding against him.

Smith, Byatt and Self Each of Smith, Byatt and Self was involved at various times in the negotiation and documentation of the transaction between FAI and GCRA. For reasons similar to those that I have outlined in relation to GCRA and GCR, each of them might have been knowingly involved in Wilkie's possible contravention of s. 232(2).

I should add that I consider each of them was forthright and helpful in his evidence. Although they were at times heavily involved in the negotiation of the transaction, its structure originated overseas.

I recommend that this matter be referred to ASIC for consideration of whether civil penalty proceedings should be brought against Smith, Byatt and Self. My omission of a reference to criminal proceedings is deliberate.

258 Th e impact of the FAI acquisition

14.8.6 Arising from Section 14.3.5

Tuck field

In light of the findings I have made concerning the NI contract the only possible question for referral that arises concerns Tuckfield and the provisions of ss. 300 to 305 of the New South Wales Crimes Act 1900, which are set out in Appendix G.

I have already found that the cover note was 'false', in that it was not made on the date it bears. I have found that Tuckfield knew it to be misleading and considered that F AI management might choose to use it to mislead their auditors into accepting it as genuine. I am also satisfied that Tuckfield considered that if it was so used it would be done in order to mislead the auditor in connection with its consideration of the proposed accounting treatment for the NI contract for the year ending 30 June 1998. If, as I have found, Tuckfield:

• made an instrument-the cover note

and the instrument was false

in circumstances where he

knew the instrument was false

made it with the intention that another person (any of F AI' s management) might use it to induce a third person (the auditor) to accept it as genuine-that is, as having been created on 4 July 1997

and

intended that, if so accepted, Andersen would undertake a further act­ being the acceptance of the NI contract as genuine reinsurance or the proposed accounting treatment for the NI contract-to its prejudice being the result of its having accepted the instrument in the course of its duty as auditor

then, with one caveat, I consider that Tuckfield might have contravened s. 300(1). The caveat relates to the fourth point. It is a matter for the trier of fact whether Tuckfield knew not simply that the cover note might be used for that purpose but that it would be so used. In my view, there is sufficient evidence to warrant consideration of that matter.

I have not overlooked Tuckfield's submission that this would be a peculiar result because there is no suggestion of a possible contravention by F AI management for 'using' the allegedly false instrument. But my findings in this regard stem from Tuckfield's preparation ofthe cover note, not its use.

I recommend that this matter be referred to the DPP for consideration of whether proceedings should be instituted against Tuckfield.

The failure ofHJH Insurance 259

4

6

T7326.

WITS.0094.001 at 002 pars 14 to 20.

T5073 , T5321.

HI.0026.0002.0444A, HI.0026.0002.0445 .

WITS.0086.001 at 006.

T5234 to 5326, T5442.

For more detail about the Big Six line, see WITS.0091.001 at 002 to 004.

T5077 to 5078. 9

For more detail about the work carried out by PWC, see WITS.0196.001 at 003 to 009 . 10 WITS.0196.001 at 015 par. 38. II

12

13

See, for example, HI.0018 .0100.0002A at 0015A to 0017 A; SBB.020.875 _OOOA at _002A; SBB.017.765_001A.

See also WITS.0050.00 1 at 004 to 005 - secondary evidence that in 1994 and 1995 Imperial Re was critical ofF AI provisioning practices.

WITS.0093 .001 at 001 par. 2. 14 WITS.0088.00 I at 032 par. K.1 . 1. 15

T5478.

16

T5484 to T5485- preferred over Kamha's evidence at T5721. 17 T5484. It was usual that an internal actuarial report would be reviewed by PWC. See WITS.0196.001 at 003 to 009. 18

BKEM.0004.050. 19 BKEM.0004.051. 20

T5485 .

2 1

BKEM.0004.064. 22 BKEM.0007.095. 23

BKEM.0004.067 . 24 BKEM.0004.028. 25

See also T5246, T5490 to T5492. 26 TRAH.0001.140. 27

WITS.Ol96.001 at 019 par. 59. 28 T5162, T5736. 29

BKEM.0004.064. 30 T5723 to T5724. 31

BKEM.0007.095 . 32 T5485 . 33

T5738 .

34

T8047.

35

T5739 to T5740.

260 Th e impact of the FA! acquisition

36

WITS.0086.001 at par. 76. 37 T7950. 38

T5724.

39

WITS.008l.001 at 004 to 005. 40 WITS.0081.001 at008; Tl5129. 41

T5427.

42

T5427 to T5428. 43 T8077 to T8078. 44

WITS.0081.001 at 009. 45 The spreadsheet is at F AlP. 001 I . 062A. 46

T2829/50. 4 7 WITS.0090.00 I at 015 par. 30. 48

T5608.

49

T5608.

50

BKEM.0003.002; SBA.l92.266_001. The figure of$67.5 million did not purport to canvas the under-reserving identified by Boulden in other lines of account. The dealings with Stow are discussed in greater detail in Section 14.2. 51

WITS.0081.00 1 at 009. The spreadsheet containing the updated review is at WITS.0043 .013 . 52 WITS.0090.001 at 016 pars 33 to 34. 53

WITS.0090.001 at 016 par. 35; WITS.0081.001 at 010. 54 T5613. 55

WITS.0074.001 at par. 2 and .003 at pars 6 to 7. 56 BKEM.0003.002, BKEM.0003.001, BKEM.0003 .004, BKEM.0003 .005A, BKEM.0003 .003. 57

T5072.

58

T5095, T5322, T5355. 59 WITS.0092.001 par. 5; at 003 pars 12 and 13; T5473 . 60

T5473 , T5474. 61 T5473, T5474. 62

WITS.0091.001 at par. 2, 006 to 007. 63 T5074. 64

The PWC report is at PWCA.0034.175. See also BKEM.0004.069, BKEM.0004.071, BKEM.0004.073, BKEM.0004.074, BKEM.0004.032. 65 T5141; T5169. 66

T5177 to T5178. 67 T5170. 68

PWCA.0015.016. 69 BKEM.0007.100.

The failure of HIH Insurance 261

70

T5854.

7 1

T5854.

72

T5181.

73

T5183.

74

Kamha's report to the board on 11 November 1997 is at FAIB.0009.154A. 75 FAIB.0010.154. 76 T8077.

77

T5427/45. 78 T8075. 79

T8083/46 to 59; T8084/26 to T8085/37; T8098/32 to T8100/20. 80 T5419 to T5420. 8 1

T5420.

82

T5420.

83

WITS .0082.001 at 004 par. 23. 84 FAIP.0003.002A. 85

T5193.

86

T5318.

87

T5192, T5198, T5209. 88 T5351 , T5209. 89

T5421.

90

T5423.

91

T5421 to T5422. 92 T7263. 93

T7346.

94

T5608 .

95

T5609.

96

WITS.0095.001 at 021 (Ascani). 97 WITS.0025.001 at 021. 98

FAIB.0010.057. 99 BKEM.0013.012. 100

BKEM.0013.012. 10 1 T7794. 102

T7795 .

103

T7795 .

104

WITS.0082.001 at 005 par. 27; T5208. 105 T7796. 106

T7786, T7787. 107 T6223.

262 Th e impact of th e FAJ acquisition

108

ROY .0125.0014A. 109 T5230; WITS.0088 .001 at 038 par. Nl. 110

T5510.

111

T5865 and T5866. 112 BKEM.0013 .001. 11 3

T7789.

114

T7355, T7360. 115 T7259. 116

BKEM.0013 .005. 117 WITS.0088.001 at 042 pars 0.7.5 to 0.7.7. 11 8

WITS.0043.00 1 at 008 to 009. 11 9 T5614. 120

T5614 to T5615. 121 WITS.0086.001 at 008 par. 53. 122

WITS.0086.001 at 009 par. 58 . 123 WITS.0086.001 at 009 par. 59. 124

WITS.0086.001 at 009 par. 60 . 12 5 WITS.0088.001 at 048 par. P.4.2; T5280 to T5281. 126

T5280 to T5281. 127 T5655.

128

BKEM.0002.012. a further draft of the memo, at BKEM.0002.006, is materially the same. See also the statement of David Cheal (WITS.0083 .001 at 002) as to why the From and Date fields are different from the original. 129

T5641.

130

T5645 .

131

T5645/28 to T5645/34. 132 T5649 to T5650. 133

T5650.

134

T5689.

135

BKEM.0002.003 , BKEM.0002.004. 136 GUYC.0003.007. 137

GUYC.0003.008. 138 This matter is dealt with in more detail in the Section X.X. 139

BKEM.0009.006; WILK.OOO 1.001 at 037 . 140 T5592. 141

T5572.

142

T5573.

143

T5575.

Th e failure ofHJH lnsurance 263

144

WITS.0090.001 at 002 par. 7. 14 5 T5290. 146

T5582 to T5583 . 147 T2577/41 to 53 ; T2578/21. 148

MART.0002.001 at 080; T15959. 149 WITS.0221.001 at 002 par. 12. 150

T8137, T8139. 15 1 T8143. 152

T8143 .

153

T8176.

154

T8176.

155

FAIB.0014.008. 156 T8179/24. 157

AARC.0002.043. 158 T6617. 159

FAIB.0012.000. 16 ° FAIB.0015 .003 at 004. 16 1

T8190/4. 162 T8195/22 to T8195/32 . 163

FAIB.0015 .000. 164 T8197/39. 165

T8198/35 . 166 FAIB.0011.143, FAIB.0012 .116, FAIB.0013.073, FAIB.OOI4.054, FAIB.0015 .104. 167 T3940.

168

T2923/42 to T2924/25 . 169 GCRC.002.317. 170

GCRC.004.135 . 17 1 This contract is addressed in Section 14 .2. 172

DETT.0002.00 1-004. 173 APRA.017 1. 0120; APRA.Ol95 .0316 at 0319; APRA.0195.0181 at 0184. 174

FAIB.001 8. 046, FAIB.0022.044. 175 SBA .070.777 001. 176

HI.0026.0002. 0274A. 177 T5114. 178

T5663 .

179

T5665.

180

T5674.

18 1

T5361126 to T5361155.

264 The impact of the FA! acquisition

182

WITS.0095.001 at 013 to 034. 183 WITS.0095.001 at 009. None ofMDU, the Big Six, ALAS or Crane were the subject of reinsurance recoveries. 184

WITS .0098.00 I. 185 WITS .0098.001 at 004. 186

WITS.0098.001 at 005. 187 WITS.0200.001 at pars 3 to 4. 188

SUBM.0006.001 at 185. 189 See WILA.0001.038; Tl644 to Tl646; WITS.0007.001 at par. 9. 190

T2845 to T2846. 191 WILA .0001.038. 192

T2847/40 to T2847/50. 193 WILA.OOO 1.042. 194

T2852/l 0 to 2852115; T2314/23 to T2314/3 7. 195 WILA.OOOI.036 . 196

WILA.0001 .036 at 037. 197 T2856; T2858/55, T2859/5; T2854/59 to T2858/5;T2859/15 to T2859/30. 198

T2319toT2320. 199 HI.0026.0002.0422 . 200

WILA.0001.023. 201 WILA.OOO 1.023 202

WILA.OOOI.006. 203 CIV.019.001 460. 204

WILA.0001.021. 205 T 1665/7. 206

WILA.OOO 1.117 207 WILA.OOOI.092 208

T292 1140 to T2921145 (Wilkie); Burroughs could not recall this being discussed: T2343/43 to T2343/55 . 209 WITS.OO 16 .00 I at 002 par. 7 (Self). 210

WITS .0016.001 at 002 par. 8 (Self); T2903 to T2904; T7367/20. 211 GCRC.002.317; Houldsworth, Vukelic, Ellingsen, Smith, Barnum and Byatt. 212

T2909 to T2913 esp. T2913115 . 213 T7274/40 to T7274/50. 214

GCRC.004.135 . 215 GCRC.004.134. 216

GCRC.002.307 (email) and GCRC.002.300 (draft slip). 217 WITS.0017.001 at par. 6; Tl092 to Tl094.

The f ailure of HIH Insurance 265

218

He said that the position changed in late April I998: TI 095/5. 219 GCRC.002.305. 220

T7278/40 to T7279/J 0. 22 1 T2929/4I to T2929/46. 222

GCRC.002.291. 223 GCRC.002.289. 224

GCRC.004.031. 225 GCRC.004.026. 226

T2362/55; T2364/50 to T2364/60. 227 GCRC.004.024 to 029 and see T2642/I9 to T2642/22. 228

GCRC.004.116. 229 Tl129. 230

T2651 and T2632 (Trahair). 23 1 GCRC .004.019, an incomplete version ofwhich was provided to the Commission. 232

GCRC.004.0I9. 233 GCRC .004.028 at 029. 23 4

GCRC .004.0I9 at 023. 235 Tl29I /4. 236

Tl293/4 to TI293/34. 237 WITS .0025.001 at 002 par. 5. 238

Tl295110. 239 Tl296. 240

WITS .0025.001 at 002 par. 5. 24 1 T295 1/5 to T2951 /30. 242

T2951/20 to T2951/35 . 243 T7284 to T7381. 244

T7381/I5 to T738 1/25. 245 Tl307/4 to TI307/23. 246

GCRC.004.I32. 247 TJ137/7 to Tll47/34 (Smith) and T2371/25 to T237 1/30 (Burroughs). 248

See BKEM.0006.020; TI299/7; T2371 /33 ; Tl139/2. 249 Tll4I /7 (Smith); T2370 to T237I esp. T237I /50 to T237 1/60 (Burroughs). 250

BKEM.0006.020. 25 1 BKEM.0006.02I. 252

HI.0004.0002 .0206, HI.0004.0002 .0 I9I , HI.0004.0002.020 I, HI.0008.0002.00 I5 , HI.0004.0002. 0 I84 and HI.0004.0002.0 186 . 253 They were in fact paid on 29 May I998 (SBB.OJ 0.920 _ 00 I) and 9 July I998

(GCRC.O 16.007).

266 Th e impact of the FA! acquisition

254

HI.0004.0002.0173. See also T2374/50 to T2374/55. 255 See Tll56/2. 256

HI.0014.0001.0175. 257 GCRC.004.002. 258

Tll39.

259

Tl150/ 1 (Smith). 260 GCRC.001.151. 26 1

T2962.

262

BKEM.0006 .023 . 263 T2963/34. 264

T2964/3 to T2964/ 18. 265 T2965/ l9 to T2965/33. Wilkie stated that he thought that Self was present when Mainprize signed it. 266

T7387 /36 to T7387 /40. 267 WITS.0016.001 at 004 par. 17 (Self); WITS.0017.001 at 004 par. 16 (Smith). 268

Tll63.

269

HI.0014.0007.0214 at 0219. 270 Tl223. 27 1

T1386 to Tl388. 272 Tl388/4. 273

WITS.0036.001 at 002 par. 8. 274 T2983/35 to T2983/55; T2984/25 to T2984/30. 275

T2984/36 to T2984/37. 276 T2908/40 to T2908/55; T2984/40 to T2984/56. 277

T2986/25 to T2987/35 . 278 T2986/50. 279

T7286.

280

T7385 .

28 1 Tl311.

282

T2971140 to T2971 /45. 283 Tl311/7; but as to Smith see Tl155. 284

GCRC.004.013. 285 GCRC.009.010, GCRC.OI0.005 , GCRC.011.012, GCRC.012 .012 , GCRC.008.009, GCRC.007.013 and GCRC.009.009. 286

SBB.010.920 001. 287 SBB.010.921 001. 288

Tl331 /9toT1332/3. 289 GCRC.001.048 at 050 (email) and GCRC.001.067 (draft slip).

Th e failure ofHIH Insurance 267

290

GCRC. 001.048 at 049 (email) and GCRC.001.059 (draft slip). 29 1 Tl286/35. 292

GCRC. 001.048 to 049. 293 GCRC .001.051. 294

Tl738/7. 295 Byatt' s draft slip (mistakenly) did not provide for the removal of $13 million premium but if it had then the gap would have been $25 million ($87 million minus $62 million). 296

GCRC.001.048. This letter was to be similar to the 1 May 1998 side letter: Tl745/2. 297 GCRC.001.129. 298

GCRC.001.077. 299 The Commission did not receive a copy of'draft 5'. 300

GCRC.001.048 . 30 1 Tl761 (Byatt); T7292; T7397 (Mainprize). 302

See WITS.0016.001 at 005 at par. 20 (Self). 303 Tl345/3 (Self). 304

GCRC.004.014. 305 GCRC .004.003 and see WITS .0016.001 at 005 par. 20 . 306

T3067 /5 to 3067 I l 0. 307 $5 million was said to be due on 1 January 1998 (i.e. six months before the contract was entered into) and $2 million was due on 30 June 1998: GCRC.004.004 at 006. The

rationale for thi s was set out in Byrne's email. 308 Payable 30 June 1998, 1 January 1999, 1 January 2000, 1 January 2001 and 1 January 2002: GCRC.004.004 and 006. 309

GCRC.004.004 at 006. 3 10 WITS .0028.001 at 012 par. 2.23. 3 11

Tl907.

312

Tl221 /54. 3 13 Tl143. 31 4

WITS .0026.001 at 009 . 3 15 Self agreed with this (Tl339/5 to Tl340/ 1) although he insisted the 'design of the product' came from overseas: Tl340/28 to Tl340/29. 3 16

T1310 to Tl311 and Tl393 to 1T394; Tl155 (Smith); T1694 (Byatt); Tl962 (Barnum). 3 17 T297 l/40 to T297l/45. 3 18

Tl961/38 to Tl962/21 (Barnum); Tl693 (Byatt); Tl252 to Tl253 (Self); T2965/55 to T2966/5 and T2969/ 15 to T2969/20 (Wilkie); T2351/45 to T2351 /50, T2353 /45 to T2353 /55 and T2355/25 to T2355/30 (Burroughs); Tl147 to Tl158 and T1152 to Tll53 (Smith); T7383 to T7385 (Mainprize). 3 19

Tl257.

320

Tl967 /34.

268 Th e impact of the FA! acquisition

321

T1725/3. 322 WITS.0015 .001 at par. 7. 323

T2356 to T2357. 324 T2356/3 (Burroughs). 325

T2401 /1 5 to T2401/25. 326 T2974/40 to T2974/45; T2975/25. 327

T2966/55 to T2967/24. 328 T2967/1 5 to T2967/25; T2976/20 to T2976/30. 329

T3310/48 to T3310/56; T3098/21. 330 T7387/52 . 33 1

T7388/1 0. 332 T7391 /1 to T7391!10. 333

BKEM.0006.020. 33 4 See GCRC.004.002 and T7399/20 to T7399/49. 335

WITS .0026.001 at 010. 336 The email is BKEM.0006.021 . 337

Tl067 (Smith), Tl249 (Self), T2297 and T2305 (Burroughs), Tl904 to Tl906 (Barnum), T7328 to T7330 and T7396 (Mainprize) and T2835 to T2386 (Wilkie). 338 T2386/l to 5 (Burroughs); Tll55/3 (Smith); Tl353 (Self); T2981/48 to 54 (Wilkie);

T7431 /25 to T7431/45 (Mainprize). As to Barnum see below. 339 See, for example: T2981 /55 to T2982110 (Wilkie); see also Til 54 to Tll55/23 and Tll97/6 (Smith); Tl353/5 to Tl353/7 (Self); T238611 to T2386/5 (Burroughs). 340

Tl723/51 . 341 GCRC.002.317; sent to Houldsworth, Vukelic, Ellingsen, Smith, Barnum and Byatt. 342

GCRC.001.151; copied to Ellingsen, Houldsworth, Smith, Barnum and Byatt. 343 Tl092 to T1094 (Smith). 344

WITS.0016.001 at 004 par. 17. 345 WITS.0017.001 at 004 par. 16. 346

See WITS.0009.001. 347 T1726 to Tl754. 348

Tl732.

349

Tl753toT7154. 35 0 GCRC.OO 1.006. 351

T2012/45 to T2012/55 . 352 T2016/45 to T2016/51. 353 Tl971.

354

T2049 to T2050. 355 Tll89/ l.

The failure of HIH Insurance 269

356

Tl188/4. 357 WITS.0008.001 at 006 par. 33 . It seems that Adler had also contacted Tuckfie1d: T2123 (Tuckfield) and GUYC.0003 .001. 358

GUYC.0003 .00 1. 359 T8169/30. 360

T905/ 18 to T905/51; WITS.0008.00 1 at 006 par. 33. 361 GUYC.0001.128. 362

GUYC.0001.129. 363 GUYC.0001.136. 364

GUYC.OOO 1.139 and see T950/3 explaining handwritten notes on the draft. 365 GUYC.0001.117; the accompanying fax cover sheet is GUYC.0001.116. 366

GUYC.0003.048. 367 GUYC.0003 .050. 368

GUYC.0003.051. 369 T965 (Bromley) and T2148/1 (Tuckfield). 370

T2103.

371

GUYC.0001.136; T2113/35. 372 T2113/50. 373

The draft slip bears a handwritten note in Tuckfield's handwriting: 'back date August last year. Premium payable next year?': GUYC.0001.139; T2114/10. 374 T2105/20 . . 375

T2104/20. 376 T3160/35 to T3160/55. 377

T7294/27 to T7294/32. 378 GUYC.OOOl.lOO. 379

GUYC.OOOl.lOl. 380 GUYC.0001.101 at 104 and GUYC.0001.105 at 108. 381

GUYC.0001.139; GUYC.0001.116 and GUYC.0001.111. 382 GUYC.0001.129. 383

GUYC.0001.139 at 141 , GUYC.0001.111 at 113 and DUYC.0001.129 at 131. 384 GUYC.0001.154. 385

GUYC.OOOI.I53. 386 GUYC.0001.091 at 094, GUYC.OOOI.081 at 084, GUYC.0001.075 at 078,

GUYC.0001.063 at 066, GUYC.0001.043 at 046 and GUYC.0001.054 at 057 . These slips were also either on blank letterhead or Carpenter Bowring letterhead except for GUYC.0001.081 which was on Carpenter Bowring letterhead but had one page (086) on Guy Carpenter letterhead. 387

T21 05/40 to T21 06/30. 388 GUYC.OOOl.lOI and GUYC.0001.105 .

270 The impact of the FAI acquisition

389

GUYC.0001.091 , GUYC.0001.081 , GUYC.0001.075, GUYC.0001.063. The final version of the placement slip and the cover note used the reference code 'FAI/389/97'A'. GUYC.0001.043, GUYC.0001.054 and AS .0002.0002.0391. 390

GUYC.OOOl.l29. 39 1 GUYC.0003.053 . 392

WITS.0008.102 at 111. 393 T1003. 394

GUYC.0001.081 at 086. 395 GUYC.OOO 1.090. 396

GUYC.0001.074. 397 GUYC.OOOl.lOO. 398

GUYC.OOO 1.075. 399 GUYC.0003 .055. 400

GUYC.OOO 1.072. 401 T898 and T993/1 . 402

T2248/34. 403 T7295/33 . 404

T7600/47 to T7600/58. 405 T7407/7 to T7407110 . 406

T3077/1 0 to T3077 /20. 407 GUYC.0001.072 at 073 . 408

GUYC.0001.063. 409 GUYC.OOO 1.062. 410

GUYC.0002.001. 41 1 T2152/50 to T2153/10. 4 12

GUYC.0001.053. 413 GUYC.0003 .060; T979/34. 41 4

Wilkie told the Commission that the 70 per cent attachment point was selected to ensure sufficient recoveries to produce the betterment of $30 to $40 million that was required: T3051 /40. 4 15

T3080/ 10 to T3080113. 4 16 GUYC.0001.054. 417

GUYC.0001.043; the correspondence relating to the forwarding of these slips­ GUYC.0003.059; GUYC.0003.060. 4 18 See for example WITS.0029.001 at par. 14; T215511 to T2155/5; T2248; T3080/36 to

T3080/38 . 41 9 The date 30 June 1998 appearing on the slip before signed on behalf ofFAl: GUYC .OOO 1.054. 420

GUYC.OOO 1.043 at 044.

The f ailure of HIH insurance 271

421

GUYC.0001.043 at 044. 422 GUYC.0001.043 at 049. 423

GUYC.0001.043 at 046 to 047 . 424 GUYC.0001.043 at 048 . 42 5

GUYC.0003 .051. 426 GUYC.0001.043. 42 7

GUYC.0001.043 at 046. 428 GUYC.0001.043 at 047. 429

See T3003/55 to T956/58. 430 See WITS .0029.001 at par. 4; WITS.0008.001 at 012. 431

GUYC.0001.043. 432 GUYC.0001.043 at 047. 433

HI.0014.0001.0176. 434 See GUYC.0003 .050 and GUYC.0003.051. 435

WITS.0026.001 at 006. 436 WITS .0026.00 1 at 004. 437

GUYC.0003 .061. 438 GUYC.0003 .061. 439

AS.0002.0002.0391 and T2104/50. 440 T2821 (Peiris ). 441

AS.0002.0002.0391 at 392 (first premium payable in arrears) and 395 (alteration of definition of experience account balance). 442 T2107 to T2108. Tuckfield made various other changes. Nothing turns on them: they

are described in the statement at WITS.0029.001 at 006 to 007. 443 AS .0002.0002.0391. 444

See Tuckfield's statement at WITS.0029.001 at 004 par. 17 and Guy Carpenter's oral submissions at Tl9096/18 . 445 T2161130 to T2160/35. 446

See HI.0014.0008.0007 at 0013. 447 WITS.0029.001 at 003 par. II. 44 8

GUYC.0003.053, GUYC.0001.089, 090 and 091 , GUYC.0003.055 and GUYC.0002.001 and T2152/10. 449 T3942 to T3943. 450

See APRA.0171.0006; APRA.Ol71.0007; APRA.Ol71.0008; APRA.Ol71.001 0; APRA.Ol71.0011; APRA.Ol71.0012; APRA.Ol71.0013 ; APRA.0171.0014; APRA.Ol71.0017; APRA.Ol71.0018; APRA.0171.0019; APRA.Ol71.0020; APRA.Ol71.0021; APRA.017l.0022; APRA.Ol71.0026; APRA.0171.0027. 451

WITS.0048.001 at 005 par. 35.

272 The impact of the FA! acquisition

452

They had the codes 'BONXLI' , 'PIXULI ', 'PIXUL2' and 'MISCSLJ ' : See APRA.0171.0027 and WITS.0048.001 at 003 pars 23 and 24 (Colomb). 453 Identified by the code ' SLGEN 1 ': T3412/20 to T3412/24 (Colomb). 454

Identified by the code 'SLGEN 2' : T3414/ 19 to T3414/29 (Colomb). 455 APRA.O 171.0021. 456

APRA.0171.0010. 457 T3214/46. 458

T3214/51 to T3214/53 . 459 T3 207 /55 to T3208/2; T3208/42 to T3208/44 ; T3209/ 18 to T3209/2 1; T3 2 10/58 to T321 1/9. He took comfort from the fact that Colomb had prepared the return and so it

was appropriate for him to sign it. 460 See HI.0014.0007 .0150 . 46 1

Trahair accepted this: T2683/38 to T2683/56. 462 BKEM.0006.035 and BKEM.0006.034. 463

See TRAH.0001.041 and WITS.0009.001. 464 HI.0014.0001.0176. 465

WITS.0009.001. 466 WITS.0035.001 at 003 par. 12; T3753/44 to T3753/55 (Vanderkemp). 467

HI.0014.0001.0176. 468 WITS .0035 .001 at 003 par. 13. 469

WITS.0035.001 at 004 par. 16. 470 WITS.0035.00 1 at 004 pars 16 and 19 (Vanderkemp). T2511 /43 to T2512/ 11 (Peiris). 471

WITS.0035 .035; WITS.0035.040; WITS.0035.044. 472 WITS.0035 .044 at 046 and 049. 473

WITS .0035.050. 474 WITS.0035.001 at 006 par. 26. 475

WITS.0035 .001 at 006 to 007 pars 31 to 32 (Vanderkemp); T3941 /25 to T394 1142 (Scrivens) 476 WJTS .0035.001 at 007 par. 32 (Vanderkemp); WITS.0034.001 at 004 to 005 pars 21

and 23 (Scrivens). 477 WITS.0034.001 at 005 par. 24. 478

WITS.0034.001 at 007 par. 31. 479 WITS.0034.001 at 007 pars 32 to 33. 480

WJTS .0034.001 at 007 pars 32 to 33 . 48 1 WITS.0034.001 at 008 to 009 pars 38 to 41. 482

T2512/13 to T2512/23 . 483 HI.0014.0007.0150. 484

HI.0014.0007.0 150. 485 SBA.070.768_001 ; BKEM.0006.054 and BKEM.0006.055 .

Th e failure ofHIH Insurance 273

486

SBA.070.768_001 at 768_004; BKEM.0006.055 at 055 to 056. 487 BKEM.0006.054 and BKEM.0006.055. 488

WITS.0034 001 at 009 to 011 pars 43 to 47 (Scrivens). 489 WITS.0034.00 1 at 0 I 0 to 011 par. 46; T2531!27 to T2531131. 490

WITS.0034.001 at 010 to 011 par. 47. 491 WITS.0034.001 at 009 pars 42 to 47. 492

WITS.0034.001 at 011 to 012; see T8000/16 to T8000/27 (Adler). 493 SBA.070.768_001; T2539/20 to T2539/24 (Peiris). 494

Andersen's assessment of the maximum benefit available reflected its belief that $80 million was available under the NI policy irrespective of any earthquake. 495 Some of the witnesses suggested that Burroughs attended, but I do not make that

finding.

496

WITS.0034.001 at 013 par. 58. 497 T3942/40 to T3942/49. 498

WITS.0034.001 at 013 par. 59. 499 WITS .0034.001 at 013 par. 59. 500

WITS.0034.001 at 014 par. 59 . 501 WITS.0034.001 at 014 par. 59. Scrivens added that it was his recollection that Barnum did most of the talking and Tuckfield merely nodded in agreement: WITS.0034.001 at

014 par. 60. 502 WITS.0034.188. 503

WITS.0033 .018. 504 T2490/42 to T2490/49 (Peiris); T2635/28 to T2635/35 (Trahair). 505

AARC.OOO 1.057. 506 AARC.OOOI.058 at 059. 507

AARC.OOOI.058 at 060. 508 AARC.OOO 1.058. 509

AARC.0001.058 at 060. 51 ° FAUD.0002.128. 511

FAUD.0002.128 at 135. 512 FAUD.0002.000. 513

WITS.0034.001 at 019 to 020 par. 80. 514 WITS.0034.001 at 019 to 020 par. 80. 515

WITS.0034.001 at 020 pars 81 and 83. 516 WITS.0034.001 at 020 par. 84. 517

WITS.0034.001 at 020 par. 85. 518 WITS.OI04.001 at031 to032par.IOO. 519

WITS.Ol04.001 at 032 par. lOO(g) . 520 WITS.O 104.001 at 033 par. 102.

274 The impact of the FA! acquisition

521

T7851/36 to T785l/39. 522 T7854/48 to T7855/2. 523

T7855/46 to T7855/47 . 524 T7851 /51 toT785l/53. 525

T785 l/54 to T7851/56 526 T2564/47 to T2564/58 . 527

T2569/3 to T2569/8. 528 T2562/33 to T2562/54. 529

T2562/56 to T2563/6. 530 Scrivens and Davies both thought the standard was ambiguous: T4566/33 to T4566/35 and T4570/22 to T4570/30 (Davies) and T4074/19 to T4074/57 (Scrivens). 531

FAUD.0002.000 at 000 005 . 53 2 FAUD.0002.000 at 000 004. 533

FAIB.00l7.000. 53 4 FAIB.00 17 .000 at 000 002 and 000 003 . - -535 FAIB.00 17 .000 at 001. 536 See CIV.Ol9.002 at 003 to 005. 537 CIV.019.002 at 006 . 538 CIV.019.002 at 033. 539 CIV.019.002 at 083 . 54° CIV.Ol9.002at003. 541 WITS .0009.001. 542 This was put to Mainprize and denied by him at T7441/30 to T7441 /33. 543 CIV.019.002 at 039. 544 CIV.019.002 at 040. 545 T2565/28 to T2565/41 . 546 WITS.0034.001 at 021 par. 88. 547 WITS .0026.00 1. 548 WITS.0028.001. 549 WITS.0057.001. 550 T4765/15 to T4765/22. 55 1 See clause 9 (reinsurance of short duration contracts) and clause 12 (reinsurance of long duration contracts) in FIS 113 : WITS .0028.001 at 052 to 055 pars 5.12 to 5.18 (Couttas). 552 T4753/22 to T4753/48. 553 T3717/ 15 to T3717/21 (Vankerkemp); T4000/l4 to T4000/42 (Scrivens). 554 See, for example, T4000/l5 to T4000/25 (Scrivens); T890/l to T890/58 (Bromley); T4735/34 to T4736/39 (Findlater). 555 SUBP.OO 18 .001 at 024 par. 55 . The failure ofHJH Insurance 275

556

T4074/30 to T4074/40. 55 7 WITS.0026.001 at 009 to 010. 558

Using a discount rate of 5.85 per cent per annum. 559 WITS.0026.001 at 008. 560

See T4768 to T4771 especially at T4771 /25 to T4771/31. 56 1 WITS.0028.001 at 020 par. 3.16. 562

WITS .0028.001 at 022 to 023 par. 3.24. 563 That is, on the basis the side letter of I May 1998 (HI.OO 14.0001.0175) and the letter of 6 May 1998 linking the contracts (GCRC.004.013) were effective. 564

WITS .0026.001 at 010. 565 WITS .0026.001 at 010 and T4763/25 to T4763/41. 566

WITS.0028 .001 at 029 to 030 par. 3.39. 567 WITS.0026.001 at 004. 568

WITS.0026.001 at 004. 569 WITS.0026.001 at 004. 570

See the clause ' Experience Account Balance' : HI.0014.0008 .0007 at 0011. 571 WITS.0026.001 at 005 as corrected in oral evidence at T4751 /30 to T4752/25. 572

WITS.0026.001 at 005 . The KPMG actuarial analysis is ROY.0023 .045. 573 WITS.0026.001 at 006. 574

WITS .0026.001 at 006. 575 WITS.0028.001 at 036 par. 4.12. 576

WITS .0028.001 at 045 par. 4.35. 577 WITS .0028.001 at 040 par. 4.24. 578

See Pearson ' s report of 5 February 2002 : WITS.0026.001 at 009 to 010. 579 WITS .0028.001 at 028 par. 3.35. 580

See Couttas ' s first report: WITS.0028 .001 at 029 par. 3.39. 58 1 AARC.0001.058 at 060. 582

T3196/27 to T3196/33 . 583 T3975/54 to T3976/2. 584

T3974/01 to T3974/15 . 585 T3 978/43 to T3979/35. 586

He merely agreed with a suggestion that it might have affected the way he approached the audit: T3944/6 to T3944/9. 587 See SBA.339.178 001 and T3962/55 to T3963/9. 588

Vanderkemp accepted this was a risk factor: T3744/7 to T3744/22. 589 Vanderkemp agreed this was 'unusual': T3764111 to T3764/20. 590

See T398011 4 to T3980/2 6. 591 T3 977 /31 to T3977 /36.

276 Th e impact of the FA! acquisition

592

T4062/36 to T4062/39. 593 T3 7 15/ 16 to T3715 /20 (Vanderkemp) and T3943/20 to T3943/30 (Scrivens). 594

Hl.0014.0008.0007. 595 T4840/1 6 to T4840/22. 596

WITS.0057.001 at 039 . 597 SAC4: ACCS.0001.310. 598

See WITS.0057.001 at 020. 599 T4050/6 to T4050/28 ; see above. 600

WITS.0034.001 at 008 to 009 pars 38 to 41 ; WITS.0034.160. 601 WITS .0034. 160 at 177 . 602

T4013/52 to T4015/5. 603 T4018/54 to T4018/56 . 604

WITS.0034.188. 605 WITS .0034.188 at 189. 606

PWCC.0001.057 at 068 to 069. 607 T3783/2 to T3758/4 (Vanderkemp). 608

T3782/54 to T3782/58 (Vanderkemp). 609 See clause 5.1.1 of AASB1023: ACCS.OOOI.145 at 157 . 610

See the 30 June 1998 accounts at CIV.019.002 at 067 . 611 Davies thought that there was ambiguity in the interpretation of the standard: (T4566/33 to T4566/35 and T4570/23 to T4570/29) and so did Scrivens (T4074/22 to T4074/28). 612

WITS.0034.190 at 197 . 613 T4062/36 to T4062/39. 614

WITS.0034.190 at 197. 615 CIV.019 .002 at 035. 616

T4974/30 to T4974/35 (Scrivens). 617 WITS.Ol55.001 at 016 to 019 pars 53 to 64 (Richardson). 618

Tl3527/36 to T13527/38 (Fodera). 619 WITS.0161.001A at 007 to 008A par. 2.21. 620

WITS.0170.001 at 020 par. 77 (Gorrie); WITS.Ol68 .001 at 090 par. 279 (Cohen). 62 1 SBA.336.202_001. 'Vitamin' was the codename assigned by Hambros to the proposed takeover ofF AI. 622

SBA.336.202 001 at 202 004. - -623 SBA.336.202 001 at 202 005. - -624 SBA.336.202 001 at 202 005 . - -625 SBA.336.202 00 l at 202 023f. - -626 SBA.336.202_001 at 202_034 par. lO(v) . 627 T13525/24 to T13525/48 . The failure ofHIH insurance 277

628

SBA.336.239_001; SBA.336.235_001. 629 SBA.336.235 001 at 235 001. - -630

WITS .0 156.001A at 133A par. 6. 631 T12240/28 to T12240/34. 632

SBA.336.232 001 at 232 001. - -633 T13526110 to T13526/28 . 634

SBA.336.215 001. 635 SBA.336.215 001 at 215 002. - -636

BRD.Ol3 .012 at 013. 637 WITS .0155.001 at 018 to 019 pars 62 to 63 (Richardson). 638

WITS.0155.001 at 019 par. 64 (Richardson). 639 SGHA.0001.073 ; SGHA.0001.195. 64

° FODE.0001.098 . 641 SBA .336.156 001. 642

SBA.336.154_001 ; SGHA.0001.006; SBA.336.002_001. 643 SBA .336.002 001 at 002 016. - -644

STUR.0001.045 at 054 par. 3.1. 645 STUR.OOO 1.045 at 057 par. 4.1. 646

STUR.OOO 1.045 at 059 to 060. 647 SBA.336.151_001; SBA.206.371 _001. 648

SBA.206.371_001 at 371 _002; SBA.336.062_001 at 062_002. 649 BRD.029.000. 650

BRD.037.001. 651 WITS .Ol56.001Aat 131Apar. 6(Williams). 652

WITS.Ol89.001 at 060 par. 186. 653 WITS.0189.001 at 061 par. 191. 654

WITS .0168.001 at 084 to 086 pars 263 to 267. 655 WITS.0179.001 at 055 to 057 pars 259 to 261. 656

BRD.029.000 at 000 001. 657 WITS .Ol56.001A at 136A (Williams). 658

SBA.070. 731 001. 659 WITS.OI56.001A at 136A (Williams); WITS .0195 .001 at 003 par. 11 (Adler). 660

T12257/22 to T12257/27 (Williams). 661 FODE.OOI4.223 at 226. 662

Tl3531 /52 to Tl3531/54. 663 Tl3531. 664

T13530/ 16 to T13530/53. 665 WITS.0031.001 at 003 to 007 pars 19 to 45 (Davies).

278 The impact of the FAI acquisition

666

FOD£.0014.223 . 667 Tl3533/43 to Tl3534/37 . 668

WITS .O l62.001 at 042 par. 5.2.3 (Fodera). 669 WITS.Ol62.001 at 042 par. 5.2.4 (Fodera). 670

WITS.Ol62.001 at 042 par. 5.2.4. 67 1 WITS .O l56.001A at 137A. 672

Tl2262/ 17 to Tl2262/31. 673 Tl2269/46 to Tl2269/55. 674

Tl2269/57 to Tl2270/5. 675 Tl7428/23 to Tl7428/32. 676

Tl7428/36 . 677 Tl2257 to Tl2259 (Williams). 678

Tl2260 (Williams). 679 Tl2260 to Tl2261. 680

Tl2263/2 to Tl2263/8. 681 Tl2264. 682

Tl2295/51 to Tl2296/5 (Williams). 683 BRD.037.001. 684

BRD.037.000. 685 WITS.Ol68.001 at 088 to 089 pars 276 to 277 (Cohen). 686

WITS.Ol68.001 at 089 par. 277 (Cohen). 687 Tl2274/35 to Tl2274/41. 688

Tl2275/46 to Tl2275/56. 689 Tl2276/13 . 690

Tl2277 (Williams). 691 WITS.Ol55 .001 at 020 pars 70 to 72. 692

WITS.Ol55 .001 at 020 par. 74. 693 BRD.037.002. 694

BRD.037.002 at 003. 695 WITS.Ol68.001 at 087 par. 273. 696

WITS .Ol68 .001 at 088 par. 274. 697 WITS.Ol68.001 at091 par. 282. 698

WITS.Ol68.001 at091 par. 284. 699 WITS.Ol62.001 at 043 par. 5.4.1; Tl3543 to Tl3544/27. 700

Tl3545/31 to Tl3545/36. 701 WITS.Ol79.001 at 057 to 058 pars 263 to 271. 702

WITS .Ol79.001 at 059 par. 274. 703 WITS.Ol79.001 at 059 pars 274 to 275; Tl5268/43.

The failure of HIH Insurance 279

704

WITS.O 179.001 at 058 par. 268 . 705 WITS .0189.001 at 066 par. 207. 706

WITS.0189.001 at 067 par. 210. 707 WITS.0189.001 at 067 par. 210; T15664. 708

T15663 /1 6 to T15663/31. 709 WITS.0171.001 at 011 to 012 par. 42. 7 10

WITS.0170.001 at 020 par. 77. 71 1 WITS.0170.001 at 020 par. 79. 712

FODE.0014.223. 713 FODE.0014.223. 7 14

FODE.0014.223 at 226 . 715 FODE.0014.223 at 226. 7 16

BRD.037 .002. 717 WITS.0162.001 at 042 par. 5.2.5. 7 18

WITS .0179.001 at 059 pars 272 to 273 . 719 T12255/4 to T12255/10. 720

T12255/24 to T12255/51. 72 1 Tl2258 to T12259. 722

T12283116 to T12283/25. 723 T12283 to T12284. 724

T12299/38 to T12299/58 . 725 T12281/33 to T12281/46. 726

Tl4745/3 to Tl4745/48 . 727 T14745/16 to T14745119; WITS.0121.001 at 020 par. 99. 728

Corporations Law, s. 631. 729 GSAC.0001.023 . 730

GSAC.0001.023 at 024. 731 GSAC.0001.012 ; T6348 (Turnbull); GSAC.0001.093 . 732

T8534 to T8535 (Adler). 733 GSAC.0007.008 at 009. 734

T6287/44 to T6287/45. 735 GSAC.0007 .008. 736

GSAC.OOO 1.1 04. 737 GSAC.0011.001. 738

T6064 and T6196 (Pillemer); T6373, T6377, T6382 and T6410 (Turnbull). 739 GSAC.0009.001 at 009. 740

GSAC.0008 .033. 74 1 T6116/6 to T6116/ 18 (Pillemer).

280 Th e impact of the FA! acquisition

742

GSAC. 0002.021. 743 GSAC. OOO 1.606. 744

GSAC.0009.001 at 011 ; BKEM.0012.010; T6413/37 to T6413/57 (Turnbull). 745 GSAC. OOO 1.325 . 746

BKEM.OOlO .OOl. 747 SBA.206.547 _001 at _048; T6125/ 18 to Tl6125/39 (Pillemer); GSAC.0002.011. 748

GSAC.0001.591; GSAC.0001.592. See also GSAC.0002.011. 749 BKEM.0012.026. 750

T6416/55 to T6417 /9 (Turnbull). 751 SBA.339.277 001. 752

SBA.339.277 001 at 001. 753 SBA.339.277 001 at 003 . - -754

T6169/58 to T6170/3. 755 GSAC.0008.119. 756

T6142/33 to T614311 ; T6145/ 18 to T6145/22 and T6149/8 to T6149/20 (Pillemer); T6439/9 to T6439/20 and T6514113 to T6514116 (Turnbull). 757 T6144/46 to T6145/22. 758

T6149 (Pillemer); T644011 to T6440/25 (Turnbull); T8609 (Adler). 759 HARR.2012.078. 760

BRAM.0007.079. 761 T7541 /50 to T7542115 . 762

T6291 to T6293; T6478 to T6481; T6501 to T6508. 763 OCNL.0001.007; LAND.0002.054; LAND.0004.002. 764

T8278/32 to T8278/43 (Adler). 765 T6812 (Hill); T7872/52 to T7872/59 (Harris). 766

WITS.0186.001 at 002 pars 6 and 7. 767 WITS.0186.001 at 002 par. 7. 768

WITS.0186.001 at 002 to 003 par. 8 (Atanaskovic). 769 Tl2288/30 to Tl2288/33 . 770

Tl2287112 to Tl2287/20. 771 SBA.070.732 001. 772

GSAC.0001.676 035 at 676 036. - -773 GSAC.0001.658 003 . 77 4

GSAC.0001.676 041 at 676 043 . - -775 F AIB.OO 1 7 A.OOO. 776 WITS.0104.001 at 039 pars 133 to 134 (Hill); T6763/50 to T6765/33 (Hill); WITS.0103.001 at 053 par. 155 (Landerer), T6548/2 to T6548/39 (Landerer); T7210/3 to T721 0/40 (O ' Connell). Th e f ailure ofHJH Insurance 281

777

GSAC.0001.658_001; GSAC.0001.657. 778 GSAC.OOOI.477 . 779

GSAC.OOOI.477 esp. at 522 to 527. 780 GSAC.0001.674. 78 1

GSAC.OOO 1.674. 78 2 DETT.0003 .049. 783

FAIB.OOI8 .000. 784 GSAC.OOOI.381. 78 5

GSAC.OOOI.381 at 385 footnote (e). 786 T6921 /34 to T6922/40 (Hill); T7878/36 to T787911 0 (Harris). 787

GSAC.0001.381 at 392; T7200/26 to T7200/46 (O'Connell); T7882/48 (Harris). 788 GSAC.OOOI.381 at 392; T6713/42 to T6713/56 (Landerer); T720l/40 to T7202/25 (O'Connell). 789

GSAC.0001.381 at 392. 790 GSAC.OOOI.381 at 397. 791

GSAC.OOO 1.403 . 792 SBA.339.277 001. 793

T7203/34 (O 'Connell); T7878/36 to T7878/43 (Harris). 794 F AIB.0020.000. 795

PURE.0005.007. 796 F AIB.0021.000. 797

TEND.OOI6.001. 798 CIV.001.174. 799

CIV.001.174 at 183 to 184 . 800 WITS.O 179 .001 at 059 par. 276 (Abbott); Tl5270/21 to Tl5270/59 (Abbott); Tl5534/28 to Tl5535/9 (Head); WITS.OI89.001 at 073 par. 230 (Stitt); T13549/7 to

Tl3549/27 (Fodera); Tl2289/52 to Tl2290/35 (Williams). 801 Tll689/7toT1 1690. 802

CMTH.0003 .005. 803 SBA.070.752 001. 804

SBA.070.754 001. 805 Tl2320/3 to Tl2321 /29. 806

Tl2323/35 to Tl2323/38. 807 Tl2324/40 to Tl2324/52. 808

Tl2326/29 to Tl2326/35. 809 Tl2326/ 10 to Tl2326/24. 8 10

SBA.070.752_001 at 752_007 to 752_008 . 8 11 SBA.070.752_001 at 752_007 and 752_008 .

282 Th e impact of the FA! acquisition

812

SBA .070.754 001 at 754 003 . - -813 SBA.070.755 001. 81 4

WATS.0006.173. 8 15 SBA.070.752 001. 816

WITS .0291.001 at 002 par. 1. 8 17 SBA.070.755 001. 818

WITS.0291.001 at 002 par. 2. 819 WITS.0217.001 at 001 par. 5 (Bennett); WITS.0210.001 at 001 par.5 (Jolly). 820

W ATS .0006.173. 82 1 SBA.070.758 001. 822

FAIB.0022.012. 823 BRD.050.000 at 003. 824

Tl6ll8/35 to Tl6ll8/49 (Adler); SBB.066.65l _OOJ. 825 HARR.2012.066. 826

WATS.0006.170. 827 HARR.2012 .066 at 068 . 828

WITS.0031.001 at 003 par. 24 (Davies). 829 FAUD.0001.123 at 125 . 83

° F AUD.OOO 1.000. 83 1 F AUD.0002.000. 832

WITS.0031.00 1 at 004 par. 25 . 833 T4143/7 to T4143/49. 834

T4475/20 to T4475/32 (Davies). 835 T4475/34 to T4475 /37. 836

T4475/49; CIV.019.002 at 039. 837 T4476/1 to T4476/9. 838

T4476/ 12 to T4476/31. 839 T4476/33 to T4476/38 . 840

T4476/52 to T4476/55 . 84 1 T4478/28 to T4478/41. 842

T4479/7 to T4479/11. 843 TJ3538/33 to Tl358/48 ; Tl3736/26 to TJ3736/59. 844

Tl7419/36 to Tl7419/44. 845 Tl7419/46 to Tl7420/30. 846

Tl7420/34 to Tl7420/37. 847 Tl7421 /6toTJ7421 /9. 848

Tl7426/17 to T17426/25. 849 T17426/48 to Tl7426/58 .

The failure of HIH Insurance 283

85 0

Tl7421114toTl7421 /23. 85 1 SBA .027.809 001. 852

WEIN.OOOl.OOl. 853 WEIN.OOOl.OOI at 003 to 004 pars 8.2 and 8.3; SBA.027.809 _001 at 809 _005 to 809 _006 pars 8.1 and 8.2. The word 'prospectus' and some further text appear in place

of the words 'Part A Statement' in the prospectus report. 85 4 SBA.027.809 001 at 809 006. - -855

SBA.027.809_001 at 809_155 to _ 157; WEIN .0001.001 at 122 to 124. 856 WEIN.OOOl.OOl at 122 ; SBA.027.809 _001 at _ 155. 857

WEIN.OOOl.OOI at 062 par. 7.2; SBA.027.809_001 at 809_097 par. 7.2. 858 WEIN.0001.001 at 063; SBA.027.809 _001 at 809 _098 . 859

BRD.039.025. 860 AARA.0519.0299. 86 1

Tl3553/59 to T13554/36 (Fodera). 862 HI.0016.0033.0005 . 863

Hl.0016.0033.0005 at 0006. At HI.0016.0033.0005, the letter said that 'following discussions today, the amount is likely to be raised to $35 million', apparently meaning that the amount likely to be raised was $35 million. 864

SBA.21 0.429 _ 001 ; The covering fax shows the letter was sent on 23 October 1998 . 865 SBA.21 0.429 001 at 429 002 to 429 003 . - - -866 SBA.21 0.428 001. 867 SGHA.0023.023. 868 Tl3554117 to Tl3554/36. 869 Tl5792/41 to Tl5792/46 and Tl5792/58 to Tl5793/9. 870 SBA .l55.223 001. 87 1 FODE.0015 .209. 872 SBA.155.238 001 at 238 002. - -873 PCKB.OOO 1.206. 874 PCKB.OOO 1.221. 875 Tl7327/20 to Tl7327/23 (Richardson); WITS .Ol90.001 at 002 par. 2 (Stitt); Tl5786/40 to Tl5787/4 (Brown). 876 BRD.038.000 at 000 002 . 877 BRD.038 .124. 878 SUBP.0054.001 at 033 par. 6.3. 879 BRD.038.000 at 000 002. 880 BRD.038.124. 88 1 WITS .Oi90.001 at 002 to 003 par. 5 and Tl5670/7 to Tl5670/ 13 . 882 Tl5671 /27 to Tl567 l/31. 883 T 15273/50 to Tl5273/54. 284 The impact of the FA! acquisition

884

Tl5273/56 to Tl5273/59 . 885 Tl230911 0 to Tl2309/ 17. 886

Tl4426/47 to Tl.4427/38 (Cassidy); Tl4947/22 to Tl4948/20 (Cohen), Tl51 70/42 to Tl51 70/55 (Sturesteps) and Tl5536/50 to Tl5536/52 (Head). 887 BRD.039.000 at 000 002 . 888

BRD.039.000 at 000 002. 889 BRD.039.000 at 000 002 . 890

BRD.040.000. 89 1 BRD.039 .025 . 892

BRD.039 .025 at 049. 893 BRD.039 .025. 894

BRD.039.025 at 032 . 895 BRD.039 .025 at 032 . 896

BRD.039 .025 at 034 . 897 BRD.039.025 at 040. 898

DEUB.0005 .012 at 015 . 899 Tl567 8/23 to Tl5678/31. 900

WITS .Ol90.001 at 003 par. 8. 90 1 SUBP .0097.001 at 028 to 029 par.77. 902

BRD.039 .000 at 000 001 . 903 Tl6678/4 to Tl6678/9 and T1917811 4 to Tl9178/20 . 904

Tl746911 7 to Tl7469/53 . 905 Tl7474111 to Tl7474/19. 906

AND.l387 .0023 .0087 . 907 ACCS .0002.00 1. 908

ACCS.0002.001 at 008. 909 ACCS.0002.001 at 035. 9 10

BRD.039.025 at 065. 9 11 CIV.OO 1.838 at 876 and 900 and CIV.001.926 at 952 and 980. 912

HI.0016.0025 .0152. 9 13 HI.0016.0027.0040 at 0041. 914

HI.0016.0025 .0152. 915 BRD.063.000 at 002 . 916

Tl5797/45 to T15797/49. 917 Tl5800/8 to Tl5800/1 6. 9 18

T15799/10 to Tl5799/27 . 919 T15800/23 to T15800/26. 920

WITS .0037.001 at 002 par. 8.

The failure of HIH Insurance 285

921

WITS.0037.001 at 002 par. 8. 922 WITS.0037.001 at 007 par. 37. 923

WITS.0037.001 at 007 par. 41. 924 WITS.0037.036. 925

WITS.0037.001 at 007 par. 38. 926 WITS .0037.036 at 037. 927

T3554/34f. 928 T3554/28 to T3354/32. 929

T3555/29 to T3555/31. 930 WITS.Ol61.001A at 053A par. 6.42. 931

WITS.Ol61.001A at 056A to 057A par. 6.53. 932 SUBP.0050.001 at 013 par. 4.1.1. 933

AND.6134.0433. 934 CORE.0009A.295. 935

HIHP.OOlO.OOl. 936 Taken from HIHP.OOlO.OOl. 937

SUBP.0030.001 at 133 par. 2.2. 938 SUBP.0030.001 at 133 and 134 par. 2.2. 939

SUBP.0030.001 at 133 to 134 par. 2.3 . 940 SUBP.Ol05 .001 at 001 to 002 pars 5 to 8. 941

WITS.Ol61.001A at 099A to 110A pars 14.1 to 14 .37. 942 WITS.Ol61.001A at lOlA par. 14.9. 943

WITS.Ol61.001A at lOlA to 102A par. 14.10. 944 WITS.Ol61.001A at 102Apar. 14.11. 945

WITS.Ol61.001A at 102 par. 14.12. 946 WITS.Ol61.001A at 104A to 107A pars 14.17 to 14.28. 947

AUDC.015.099 at 101. 948 BRD.059.261 at 266. 949

SBA.321.590 001. 950 SBA.321.590 001 at 590 005 . - -951 SBA.321.590 001 at 590 002 and 590 003 . - - -952 SBB .088.772 001. 953 SBB.088 .7 72 001. 954 SBA.188.450 001A. 955 SBA.188.450 001A at 450 013A. - -956 W A TS.0006.13 1. 957 WATS.0006.131. 958 SUBP.0030.001 at 142 par. 3.17(b). 286 The impact of the FA! acquisition

959

AARA.0217.0081. 960 See the letter from his solicitors at CMC0.0033 .225 at 226.

The failure ofHJH Insurance 287

15 Under-provisioning for claims

Generally speaking, the most significant item in the accounts of a general insurer is its provision for outstanding claims. The under-provisioning for HIH' s outstanding claims was a major cause of both the collapse of HIH, and the extent of that collapse. It warrants careful attention.

15.1 Overview of HIH's business and provisions

15.1.1 The business

The HIH group operated a range of businesses, but its principal operations were in general insurance. By the time of its collapse, the HIH group was the second largest insurer in the Australian market. Its main operating arms were HIH Casualty and General Insurance Limited, CIC Insurance Limited and F AI Insurance Limited. It

wrote business in the following key portfolios. 1

The professional indemnity or PI portfolio was written across three companies, namely C&G, FAI and CIC. The C&G component covered a range of occupations, including accountants, builders warranty, real estate, directors and officers, barristers, solicitors, engineers and hospitals. It also contained several schemes such as NSW LawCover, the Hong Kong Hospital Authority, and Hong Kong lawyers. The F AI component was not as dominated by schemes and, relative to the others, the CIC component was quite small.

The liability portfolio was largely written across the same three companies. The C&G component consisted of many classes of business, including material damage, general liability, product liability, contract works and council. The FAI component included corporate liability and general liability. The CIC component was smaller than the others and included contract works, household liability, product liability and public liability. In addition, the liability portfolio included the Colonial Mutual Group (CMG) liability business written through a binder with Jardines.

The workers compensation portfolio included current business and business recently placed into run-off. It involved state-based workers compensation schemes in Western Australia, Northern Territory, Australian Capital Territory and Tasmania (current); New South Wales, Victoria and South Australia (run-off business); and Northern Territory (old run-off business). The current and run-off business was

spread across portfolios from C&G, F AI and CIC. Old run-off included business from FAI, CIC, C&G and World Marine General Insurances Pty Limited.

The failure of HIH Insurance 289

The compulsory third party insurance, or motor vehicle accident bodily injury insurance, was written by FAI in New South Wales and Queensland, and by CIC in New South Wales. C&G had a portfolio in this area, but its portfolio was transferred to Allianz with effect from 1 January 2001.

C&G also had a small portfolio of salary continuance insurance business, providing an income stream to insured who became temporarily disabled.

A number of lines of business were grouped into a broad category referred to as short tail. Not all of these liabilities were short tail in the actuarial sense of the term. The key lines of business classified as short tail included motor, household, pleasurecraft, commercial property, travel, engineering, trade credit and contingencies (bloodstock and livestock). HIH wrote its short tail business through a number of different divisions. Many of the divisions wrote similar lines of business, but the operations were accounted for by division, rather than portfolio. These divisions included regional, metropolitan, financial institutions, motor, property, householders and F AI. In September 2000 many of the personal lines were sold into the joint venture with Allianz. After the liquidation of HIH. Allianz agreed to assume the liabilities in respect of any policy in force on 1 January 2001. QBE has assumed liabilities on the various travel insurance books of business.

The marine portfolio comprised five distinct areas of business, namely marine US (C&G run-off), marine Australia (C&G), Hong Kong marine, F AI Australian marine and marine group stop loss.

HIH ' s UK operations consisted of both its UK branch and Cotesworth business.

The UK branch business comprised a number of classes of insurance split between marine and non-marine. The UK branch ceased writing new direct business on 31 December 1998 and new reinsurance business (treaty and pro rata) in September 1999. The only exception was for non-marine direct where some professional indemnity and travel business was written during 1999.

Cotesworth & Company Limited was the managing agent for the six Lloyd's syndicates in which the HIH group had a participating interest, namely Syndicates 228, 535, 536, 1069, 1236 and 1688. The business of those syndicates may be summarised as follows:

• Syndicate 228 wrote marine insurance. It was placed in run-off at the end of the 2000 year of account.

• Syndicate 535 wrote marine insurance and a marine excess-of-loss account, which included some business written in previous years by Syndicate 536.

• Syndicate 536 wrote marine excess-of-loss insurance and a non-marine excess of loss account. Syndicate 536 was placed in run-off at the end of the 1999 year of account.

290 Under-provisioning/or claims

• Syndicate 1069 wrote a wide range of non-marine business. It was placed in run-off from the end of the 2000 year of account.

• Syndicate 1236 wrote direct property, catastrophe excess-of-loss, property per risk excess-of-loss, retrocessional property, space, marine and professional indemnity business. It was placed in run-off at the end of the 1999 year of account.

• Syndicate 1688 wrote marine, personal accident and travel business in 1999. It continued to underwrite other classes of business and actively underwrote in the 2001 year of account.

HIH's US operations consisted largely of business written by CareAmerica Compensation and Liability Insurance Company (renamed HIH America Compensation and Liability Insurance Company) and Great States Insurance Company. The business was primarily Californian workers compensation business.

The run-off treaty business was written by C&G, CIC, CMG, World Marine and F AI. It consisted of high risk variable treaties written over many years by several companies within the HIH group. The common characteristic of the treaties within this account was that they had all been placed in run-off. The risk management and reinsurance division of HIH was responsible for the administration of the majority of the portfolios within this account.

The Medical Defence Union was an F AI portfolio that incorporated a number of reinsurance treaties covering Australian Medical Insurance Limited (AMIL), formerly United Medical Protection (UMP), formerly NSW Medical Defence Union (MDU).

HIH's Asian operations consisted of liabilities written through FAI and C&G throughout Asia, but in particular in Hong Kong.

HIH had a comprehensive reinsurance programme, the administration of which was supervised by a reinsurance committee comprising five executive directors from the main board, various senior managers and the group reinsurance manager. In 1998, HIH set up a reinsurance arrangement referred to as HIH Re. HIH Re was not a separate entity, but rather was operated as a division within the HIH group. The arrangement provided reinsurance protection for the various underwriting divisions within HIH for claims between the respective divisional net retentions and

$5 million. In addition to HIH's 'traditional' reinsurance programme and HIH Re, HIH also wrote (and inherited from FAI) several large whole of account stop-loss reinsurance contracts in the nature of financial reinsurance.

In summary, HIH's insurance business was managed through separate divisions broadly corresponding with the classes or portfolios of business to which I have referred. In some instances, for example, each of the professional indemnity, liability and workers compensation divisions and the United States, United Kingdom and Asian operations, the businesses were separately managed through

individual subsidiary companies.

The failure of H/H Insurance 291

15.1.2 The HIH group OCP

The HIH group's insurance operations grew rapidly from 1995 as it made a number of acquisitions both locally and overseas. HIH's outstanding claims provision (OCP) grew correspondingly. Table 15.1, extracted from the company's published accounts, shows this.

Table 15.1 Outstanding claims provisions, December 1995 to Ju ne 2000

Discounted Undiscounted

Reinsurance Rei nsuran ce

recoveries OCP (net of recoveries OCP (net of

OCP receivable reinsurance) OCP receivable reinsu rance) Date ($m) ($m) ($m) ($m) ($m) ($m)

31 December 1995 1 153.4 250.7 902.7 1 430.9 318.5 1112.4

31 December 1996 1 359.1 317.9 1 041.2 1 678 .8 373.5 1 305.3

31 December 1997 1 956,6 446.1 1 510.5 2 377.3 527.4 1 849.9

30 June 1999 3 699.5 1 050.2 2 649.32 4 246.7 1198.9 3 047.8

30 June 2000 4 430.9 1 819.9 2 611.0 4 922.9 1 991 .6 2 931.3

In its published accounts (for each of the periods ending 31 December 1997, 30 June 1999 and 30 June 2000), HIH included the following statement of accounting polici:

292

Provision for Outstanding Claims

Full provision is made for the estimated cost of claims notified but not settled at balance date, by review of individual claims on the basis of current information and in the light of existing court awards and levels of compensation and likely inflationary trends. Provision is also made for claims handling costs, and for the costs of claims incurred by balance date, but not reported until after balance date, based on an analysis of claims reported after balance dates in prior years.

The provision for outstanding claims is measured at the present value of expected future payments. Discount rates used are based on investment opportunities available on funds sufficient to meet future claims.

An actuarial estimate of outstanding claims for long tail classes of business is made and taken into consideration in determining the liability at balance date. Such actuarial review includes an allowance for un­ notified claims, an allowance for claims inflation, claims handling costs and anticipated investment earnings.

The provision for claims is classified into current and non-current liabilities according to estimated time of payment which has been based on a statistical analysis of payment patterns in prior years. Net claims expense includes claims settlement costs plus claims handling costs, less reinsurance recoveries received.

Under-provisioning for claims

The portfolio breakdown schedules Schedules were prepared within HIH setting out the breakdown of the group OCP on a portfolio-by-portfolio basis as at the various balance dates

between December 1997 and December 2000.4 These schedules were prepared by interrogating the general ledger and hence reflect the provisions within the general ledger (and ultimately HIH's consolidated accounts) for the individual portfolios of business at the relevant dates. 5 During the Commission hearings these schedules became known as the portfolio breakdown schedules.

The KPMG phase II report Following the demise of HIH, a KPMG actuarial team conducted a review of the claims reserves maintained by the HIH group. The review was conducted at the request of the liquidators and was headed up by Richard Wilkinson, a partner of KPMG in London with overall responsibility for KPMG's international insurance actuarial practice. 6

The first assessment of the outstanding claims reserves of the HIH group, referred to as the phase I review, provided an assessment as at 31 December 2000. It was completed on 10 May 2001 and the findings were embodied in a report ('the phase I Report'). 7

Subsequently KPMG undertook a second and more detailed assessment, referred to as the phase II review, which provided an assessment of reserves as at 15 March 2001.8 The phase II review was completed on 5 July 2001 and its findings were also embodied in a report ('the phase II report'). 9

In preparing the phase I and II reports, Wilkinson was the lead actuary, but was assisted by actuaries and staff from KPMG in London and Australia, Gary Ankcorn (a London actuary employed by HIH), Trowbridge Consulting and Ernst & Young. Trowbridge and Ernst & Young provided both peer review and actuarial support.

The methodology for the preparation of the phase I and II reports is described in the letters of instruction referred to above and in the introduction to the phase II report. Broadly, Wilkinson assigned an actuary or a team of two actuaries to undertake the valuation of each of HIH's major classes of insurance business. The actuary (or actuaries) provided an assessment which was then peer reviewed by a further actuary.

The results of both reviews were summarised in a table which appears in the phase II report.10 The table sets out the HIH group reserves as at 31 December 2000 and 15 March 2001 for the various classes of business. It has five columns, the first four of which are HIH's estimates as at 31 December 2000, a phase II going concern estimate as at 15 March 2001, a phase I break-up estimate as at 31 December 2000 and a phase II break-up estimate as at 15 March 2001. The final column contains a

recommended prudential margin. The end result (after allowing for reinsurance) can be summarised as follows:

The failure ofHJH Insurance 293

• HIH estimate (31 December 2000) $3.1 billion

• phase II going concern estimate ( 15 March 2001) $5.0 billion • phase I break-up (31 December 2000) $4.5 billion

• phase II break-up (15 March 2001) $5.7 billion

• prudential margin $1.7 billion

It can be seen from this summary that the results of the phase II valuation (ignoring any discrepancy which may be attributable to the different dates of valuation used by HIH and KPMG) suggest that the HIH group's OCL was undervalued by $1.9 billion on a discounted (or going-concern) basis. A portfolio by portfolio breakdown of this undervaluation can be obtained from the table to which I have referred.

If one were to assume the break-up of the HIH group, and use the undiscounted figures, then the phase II report valuation suggests HIH's OCL was undervalued by some $2.6 billion. The phase II report also recommended a prudential margin of about $1.7 billion. HIH did not customarily apply a prudential margin.

Wilkinson provided a further statement to the Commission in December 2002 in which he set out what he referred to as a 'Phase II High Level Actuarial Update' .11 He wrote that it was his view by that time and on the information which he and his team had reviewed, that the reserve requirement remained within the range originally indicated in the phase II report. But his view of the best estimate had

increased significantly to well above the mid range of that original forecast.

It was on the basis of the phase I and phase II reports that the liquidators opined that the estimated net asset deficiency of the HIH group at 15 March 2001 was in the range of $3.6 billion (on a central estimate) to $5.3 billion (applying a prudential margin). 12

15.1.3 The provision setting process

As I have already said, HIH' s insurance business was managed more along divisional lines than by reference to subsidiary companies.

At a senior management level there was some overlap in the personnel involved in the management of each of the divisions. For example, at all times relevant to this report Terry Cassidy and George Sturesteps were the managing director (Australia) and managing director (International) respectively. In those capacities they were

involved in the management of various divisions. Similarly, Greg Brown was the group general manager of both the professional indemnity and liability divisions, and later became group general manager of the corporate division (which assumed responsibility for various divisions, including hoth the professional indemnity and liability divisions). At the middle and lower management level, and general staff level, the responsibilities assumed by individuals tended to be portfolio specific. To

294 Under-provisioning for claims

the extent that it becomes relevant the personnel involved in the management of individual divisions are considered in the later sections of the report.

The provision setting process within HIH differed in some respects from portfolio to portfolio. Essentially, however, the process was as follows.

From the year ending 31 December 1997 David Slee was retained as the HIH group consulting actuary to provide his assessment of the central estimate of the HIH group's OCL. He did so by way of a series of six monthly reports. Slee ' s total figure for each balance date was intended to be a figure against which the reasonableness or adequacy of the total booked figure or provision proposed for HIH's consolidated

accounts could be tested. As Slee' s total figure was a central estimate it included no prudential margin. This was in accordance with his instructions.

For each balance date, senior management put forward their estimate of the OCL for each division. At a regular six monthly series of meetings, both the management and actuarial estimates of each division's OCL were discussed and considered. The financial services division (FSD) staff prepared schedules setting out the various figures to assist the discussion at these meetings. As I have mentioned, during the Commission hearings, these schedules became known as the portfolio breakdown schedules. Those in attendance at the meetings differed over time, and from division to division. However, at least for the Australian portfolios, those in attendance

included Dominic Fodera, Cassidy and Slee. Following these meetings, senior management, and in particular Fodera, would settle on a central estimate figure fo r each portfolio of business. The total of these figures was the figure which it was proposed that HIH book as its group OCP at the relevant balance dates. It was said to represent management's assessment of the central estimate of the group' s OCL.

In addition to the work carried out by Slee, Andersen carried out work in relation to HIH's OCL. Its work was limited in scope and was essentially corroborative in nature. Andersen did not have, or otherwise retain, any actuarial expertise to assist it in its audit ofHIH. As a result of its analysis ofHIH's OCL, however, Andersen did arrive at its own range of likely OCLs for the various portfolios of business­ expressed by reference to AA low, AA likely and AA high figures .

At each balance date Andersen prepared a table which was included in their regular presentations to the audit committee. The table set out for each portfolio the proposed booked figures (referred to as the HIH figure), the Slee estimate (referred to as the actuary figure) and the Andersen range (by reference to the AA low and

AA high figures). The object was to present to the directors (or at least those directors in attendance at the various audit committee meetings) information to enable them to assess the adequacy or reasonableness of the proposed OCP.

My impression from the evidence is that the consideration given to the adequacy of HIH' s proposed OCP at the various audit committee meetings was generally brief and seldom descended to particularity. From time to time there was some discussion of the discrepancies between the HIH, actuary and AA low and high figures. However, given the correspondence between the total group HIH and actuary

The f ailure ofHIH Insurance 295

figures, the proposed HIH figures were the provisions usually included in the draft consolidated accounts ultimately approved by the board.

15.2 Introduc tion to Slee and his retainer by HIH

At the time of the collapse of HIH, David Slee of David Slee Consulting Pty Ltd (Slee Consulting) was its consulting actuary. He provided extensive actuarial services to HIH. In particular, from the end of 1997 he provided six monthly reports in relation to the value ofHIH's group or global OCL.

15.2.1 Background

Slee is, and has been since 1967, a Fell ow of both the Institute of Actuaries of Australia and the Institute of Actuaries (United Kingdom). 13 He has approximately 40 years experience in the insurance industry, both as a consulting actuary and as a d. f b f " . 14 Irector or manager o a num er o msurance compames.

In 1989 Slee incorporated his actuarial practice and since that time has performed all his work through, and as the consulting actuary of, Slee Consulting. Although some of the actuarial work for HIH was performed by employees ofthat company (and in particular Robert Cameron), Slee accepted responsibility for all actuarial work done by that company.15 Slee signed the six monthly reports provided by Slee Consulting to HIH.

When Slee Consulting was incorporated, CE Heath Casualty and General Insurance Limited initially took a 49 per cent shareholding in the company, with Slee (through his family) taking the remaining 51 per cent. 16 However, Slee's evidence was that at no stage did anyone from CE Heath hold a position on the board of Slee Consulting, nor have any involvement iri the management of the company. In 1992 he suggested to Ray Williams that because by that time Slee Consulting was carrying out a reasonable (and increasing) quantity of workforCE Heath, they should dissolve the relationship between the two companies so that they would not only be independent, but also be seen to be independent. Slee's family then purchased the shares in Slee Consulting held by CE Heath.17

Prior to undertaking any work for CE Heath, Slee was an acquaintance of both Williams and Michael Payne. 18 Although he had no business dealings with Payne 19, he did have reasonably extensive dealings with Williams.Z0 It is sufficient for my purposes to note that from 1970 to 1980 Slee provided both managerial and actuarial services to various insurance companies associated with Williams. In 1980,

Slee accepted an invitation from Williams to join the then CE Heath group as an actuary? ' During the period 1980 to 1986, Slee' s work for CE Heath was principally in the area of workers compensation. It involved a wide range of tasks which were generally actuarial in nature. But following the legislative changes to the workers compensation system in Victoria in 1985 (and South Australia the

following year), CE Heath was effectively deprived of 90 per cent of its workers

296 Under-pro visioning for claims

compensation business. Slee informed Williams that as a result of these changes, CE Heath did not have sufficient work to occupy him full time, and he resigned. 22

From 1986 to 1988 Slee held the position of managing director of Royal Life Insurance. Upon the sale of that company in 1988 Slee proposed that CE Heath establish a company providing actuarial services for both the CE Heath group of companies and other insurers. 23 It was in this context that Slee Consulting was

incorporated.

During the period 1989 to 1995, Slee (through Slee Consulting) advised a number of clients on a range of insurance related subjects including the sale of companies, pricing, reserves, reinsurance, investments, management and marketing strategy. Slee estimated that the work he performed for CE Heath during this period accounted for about 20 per cent of his work. 24

15.2.2 The extent of Slee's work for HIH from 1996 to 2001

From 1996 to the collapse ofHIH in March 2001 Slee continued to provide actuarial services to clients other than HIH. During the same period, HIH took up a growing proportion of Slee's time. Slee estimated that during the year ending 30 June 1998 somewhere between 50 per cent and 80 per cent of Slee Consulting's revenue was derived from HIH. The percentage rose and Slee estimated that by the year ending 30 June 1999 it would have been around 70 per cent, and by the year ending 30 June 2000 around 80 per cent. 25 In a large part the increase was a consequence of

Slee's retainer to provide six monthly reports in relation to HIH's group OCL from 31 December 1997.

Slee was appointed consulting actuary for the entire HIH group in March 1999? 6

15.2.3 Slee Consulting's employees

Two employees of Slee Consulting were involved in performing actuarial work for HIH. 27

Cameron worked for Slee Consulting in Melbourne from 1 July 1997 until mid 2000. About half of his time was spent on HIH work, mainly with the liability business. In mid-2000 Cameron became an employee of HIH but continued in much the same role, assisting Slee in preparing his valuations. Cameron's work in relation to some of the liability valuations is considered later in this report.

Mark Kral worked in Sydney for two years from 1 July 1997 to 21 April 1999. Following discussions between Slee and Cassidy, he was hired by Slee to produce models for the HIH group domestic business. 28 He worked almost full time on HIH work, but ceased to be an employee of Slee Consulting and to do any work for HIH after the F AI takeover in 1999.

The failure ofHIH Insurance 297

15.2.4 Slee's work for HIH in relation to matters other than outstanding claims liabil ities

In addition to his work on HIH outstanding claims liabilities (see below), Slee was asked to provide advice and assistance in relation to a number of other discrete matters .Z9 He provided HIH with some limited actuarial assistance and advice in relation to pricing and premium levels in particular portfolios. He also assisted HIH in relation to various acquisitions and mergers (including CIC and CMG). In relation to reinsurance, he did some work on retention levels in 1994 and 1997. Slee also participated in several training programmes for HIH staff to explain the actuarial process and the general method of operation of an insurance organisation. In February 2001 , at the request of Randolph Weinand Jock McAdie, Slee prepared a planning document. It was only in draft when HIH ceased operations.

15.2.5 Slee's work for HIH in re lation to outstanding claims liabilities

Up to the time of the preparation of his 31 December 1996 report, Slee's instructions had been to assess the value of the specific classes of business identified in the terms of reference of his reports. 30 His valuations were primarily in respect of the Australian business, although he did carry out some work on the New Zealand and US businesses.31

In January 1998, Williams (in the presence of Fodera and Cassidy) raised with Slee the possibility that he be instructed to prepare valuations for the purpose of 'testing reserves on a global basis'. Slee was told by Fodera that he was not to include a prudential margin in his valuations, as this would be determined by the directors on a global basis. Slee agreed to undertake this task, commencing with his report as at 31 December 1997. 32

After the initial instruction from Williams, Slee received his day-to-day instructions from Cassidy and, particularly in later years, from Fodera.33 Slee explained that the terms of reference for his valuations were agreed orally (usually with Fodera) and then set out in his reports. There were no lengthy discussions on this topic. He understood, and believed the HIH executive directors understood, that his task was to test the aggregate of the net discounted claims reserve which it was proposed be booked. 34

Slee accepted in his evidence that his task was to arrive at a central estimate of HIH's group or global OCL. In his statement, he wrote that 'the hypothesis put was that the booked value proposed had a 50 per cent probability of adequacy ' .35 Hi s task was to determine an unbiased central estimate which would later be compared to the proposed booked value.36

Slee emphasised that he understood his instructions to be to test the aggregate or global booked value, and not that of any of the individual companies or divisions. 37

He acknowledged, however, that the aggregate or global nature of the test was not apparent .from either the notes to the HIH consolidated accounts (which referred

298 Under-provisioning for claims

simply to consideration having been given to 'an actuarial estimate of outstanding claims') or his six monthly reports.

Slee also emphasised the fact that his instructions were merely to test the booked provision, and not to make any recommendation as to the reserve or provision which ought to be booked. He regarded the latter as a separate task. He explained that if he were asked to make a recommendation as to an appropriate reserve or provision he would have recommended a prudential margin.

15.2.6 Instructions and assumptions in the Slee reports

In addition to the overall instruction to test the proposed booked provision, Slee was instructed to make a number of assumptions in preparing his valuations. Those instructions were provided orally or in writing?8

The instructions or assumptions were disclosed in a relatively consistent manner in each of Slee's reports. For example, in the 30 June 2000 report, under the heading 'Terms of Reference', after referring to his instruction ' to test the booked value of the proposed provision in the balance sheet of the Holding Company for all outstanding claims including claims incurred but not reported (IBNR), and future claims handing costs (FCHC) as at 30 June 2000', Slee set out the following

'conditions [which] were determined by the company'39 :

• the rate of discount to be used is 6.42 per cent per annum

• the allowance for FCHC is generally to be 2 per cent of gross liabilities

• all reinsurance is to be considered fully recoverable

• where it is inappropriate to use actuarial methods, to use case estimates (particularly where there are individual items subject to law suits and in some short-term business)

• no prudential margin is to be included, nor should any allowance be made for mismatch of assets to liabilities

• the portfolio is worldwide and in the time frame available other actuaries reports may be used to assist in the testing process and

• the report should be as simple as possible, so that the essential features are easily understood.

Although there was some variation, broadly equivalent instructions and assumptions were disclosed in each of the earlier Slee reports.

The appropriateness and effect of a number of these individual instructions and assumptions are considered below. But, there are some matters of relevance to them all. For example, PS 300 relevantly provides:

The failure ofHIH Insurance 299

• the actuary's report should state the extent of compliance with this standard, the reasons for not complying fully with this standard, and any restrictions on the actuary 40

• where the principal requires the actuary to use specific assumptions the actuary must clearly state the circumstances, discuss whether or not the assumptions are reasonable and consistent with this standard, and discuss the implications of divergence from this standard41 and

• the actuary should not recommend or support a provision which is less than the central estimate of the liabilities. 42

Estelle Pearson (an actuary from Trowbridge, retained by the Commission to provide expert actuarial assistance) summarised the effect of these provisions of PS 300 to be that if an actuary is constrained in a way which leads to the valuation result being less than the central estimate of the liabilities, the actuary must clearly disclose that the result is less than the central estimate of the liabilities, and preferably quantify the shortfall. 43

Slee did not suggest anywhere in his reports that his valuations of the OCL were less than a central estimate. On the contrary, in the 30 June 1999 report, for example, Slee wrote ofhis estimate of$2 698 million44 :

I understand that this figure is to be booked in the balance sheet, and it represents [a] liability which has [a] 50% probability of being adequate.

In the 30 June 2000 Slee report, Slee referred to a total discounted reserve for the group of $2 609 million and wrote45 :

I consider this to be a fair and reasonable estimate of the liability, with a 50% chance of adequacy in the aggregate, on the assumption that certain stop loss premiums are accounted for elsewhere and that the data and information provided to me is considered to be accurate.

In the circumstances, Pearson concluded that a reader of Slee's reports would understand them to contain his central estimate of HIH's OCL, notwithstanding the assumptions or constraints imposed by the company. 46

15.2.7 Unsigned Slee reports

Slee did not sign his 30 June 2000 and 31 December 2000 reports. Nevertheless, he did provide HIH with draft reports which on their face were complete.

In relation to his 30 June 2000 report, Slee said that his reason for not signing the report was that he had not a received an auditor or management sign-off as to the veracity of the data provided to him and the sustainability of the discount rate which he had been asked to assume. 47 In the body of that report Slee wrote that he had

used the 6.42 per cent discount rate requested by the company 'on condition that the auditors certify that adequate assets so exist and yield that amount'. He added that 'the auditors had not so certified this to me before I was asked to give a preliminary

300 Under-provisioning for claims

opmwn on the reserves at the end of August and that opinion was thus given conditionally' .48

PS 300 is silent on the issue of unsigned advice. Pearson proffered the opinion that in these circumstances it was reasonable to expect the 30 June 2000 unsigned valuation report to comply with PS 300, with the exception of the reliance upon the auditors (and management) as to the data veracity and the discount rate adopted. 49

It is not clear to me why Slee did not sign his 31 December 2000 report. While Slee did not obtain the auditor's sign-off on the data, he wrote in his report that he had been 'assured by management that all pertinent details had been supplied'. 50 It may have been the case that the provisional liquidation of HIH intervened before the report was signed.

15.2.8 Circulation of Slee's reports

Each of Slee's reports commenced with a cover page addressing the report to the directors of HIH. The reports were sent by Slee to the director named in the introduction to each report, being the director who had commissioned the particular report. Initially the relevant director was Cassidy, then Fodera and later Wein.51 The final reports, however, were not forwarded to those persons until some months after the relevant valuation date. This was generally after the accounts for the periods ending at those balance dates had been finalised. It appears that only extracts of

Slee's reports were available prior to that time. 52

Slee said that he ' assumed that [his] reports would be considered by all the directors, senior executives and possibly also APRA'. While he addressed his reports to the directors of HIH, he had no contact with the board or directors of HIH (other than Williams, Cassidy and Fodera) throughout the period 1996 to 2000. He received no requests for information, clarification or explanation from any of the directors.53

When asked whether he considered it his responsibility to ensure that the board of HIH was aware of his methodologies and assumptions, Slee said that he found himself in a very difficult position. He was hired to do a job within given terms of reference. He did that job to the best of his ability. He presented numbers to the

executive directors of the main board, discussed those numbers with them, and eventually wrote a report. In that report he tried as far as possible to stick to the terms of reference. Slee added that during his conversations with the executive directors, but not in his reports, he drew attention to some incongruities between

management's figures and his figures. Slee felt that by raising such matters with directors of the calibre of the Williams, Fodera and Cassidy, he did draw attention to them, even if they were not included in his written reports. 54

Pearson's evidence was that PS 300 did not impose any obligation upon an actuary to follow up on, or consider the use made of, their advice. 55 It is to be remembered that PS 300 makes it clear that responsibility for establishing the level of the booked OCP was with the directors and not the actuary. 56

The failure ofHIH In surance 301

In my view, given Slee's close association with HIH, best practice required that he endeavour to ensure that the directors (as the addressees of his reports) received 57 and understood his reports. By that I mean endeavour to ensure the directors had an appreciation of both the nature of the task carried out by Slee and the significance of the various assumptions he was asked to make and did make. This might have been achieved, for example, by establishing a regular opportunity to meet with the directors.

But I do not criticise Slee for the failure of the board to receive his reports, in part because his position as a consultant meant that he was constrained by his client's instructions. He was also probably entitled to assume that Fodera, Cassidy and Williams had provided the reports to the other directors or addressees. It was not as though Slee completely ignored this issue. Although he had no contact with the board or the audit committee, he had suggested 'it might be helpful' if he attended the audit committee meetings, but that suggestion was never taken up. 58

I do not believe that Slee's professional obligations required him to take further steps. The failure of the directors, on the other hand, to ensure they had an adequate appreciation of Slee's work is something of which I am critical. This is considered later in the section of this report dealing with the directors' consideration of the OCP.

15.3 Other actuarial assistance

In addition to Slee, HIH also received actuarial assistance from a number of internal and external actuaries. In arriving at his central estimates ofHIH's global OCL, Slee (and through him, HIH) relied upon the work done by the following 59 :

• UK branch-PWC (Philip Coleman) and HIH internal (Gary Ankcorn and Adam Brunskill).

• Cotesworth-Watson Wyatt (Richard Bulmer).

• Hong Kong-Crespoint (Feike Boschma).

• US operations-Milliman & Robertson (Guy Avagliano).

• CTP (Australia)- PWC (Philip Chappell). Chappell later became an HIH employee and hence an internal actuary.

Slee's evidence was that from 1999 he maintained a working relationship with all of these actuaries and effectively peer reviewed their work. He said that he found them all to be thoroughly professional and helpful in providing him with the information he sought. He said that he had adequate discussions with these actuaries and found no reason to doubt the accuracy or veracity of their work. 60

Slee disclosed his reliance upon the work of the above actuaries in his various reports. I will consider later in this report some concerns which arose in relation to

302 Under-provisioning for claims

the propriety of Slee's dealings with Milliman & Robertson in 2000, the completeness of his valuation of the UK operations as at 30 June 1999, and the accuracy of his translation of the Cotesworth valuations provided to him by Watson Wyatt as at 30 June 2000 and 31 December 2000. But putting these specific concerns to one side, Pearson concluded from her consideration of Slee' s 30 June 1999 and 30 June 2000 reports, and Slee's description in his witness statement of his dealings with these other actuaries, that Slee appeared to have taken adequate steps to satisfy himself as to the appropriateness of the work undertaken by the actuaries upon which he relied. I have not had an opportunity to examine in detail the procedures adopted by Slee in reviewing the work of these actuaries. But

nothing has come to my attention during the course of these hearings (other than the specific concerns just referred to) which suggests that those procedures were other than adequate.

15.4 Matters affecting HIH's provisions generally

There was a number of matters considered during the course of the Commission hearings which were relevant to an assessment of the adequacy of HIH's provisions across a number of portfolios, and in some cases, across all portfolios. These included the treatment of future claims handling costs, discounting, the use of prudential margins, the allowance for claims inflation, the actuarial techniques

employed by Slee, and the accuracy and flow of provisioning information. I will consider each of these in tum.

15.4.1 Future claims handling costs

In the discussion of principles earlier in this report it was seen that both AASB 1023 61 and PS 30062 require that appropriate allowance be made for future claims handling costs (FCHC) in the estimate of a general insurer' s OCL. Because the allowance is intended to reflect the claims handling costs which the insurer

expects to incur, one would expect the allowance to be determined by reference to the insurer's actual claims handling costs. 63 This is confirmed by the references in paragraph 5.2 of AASB 1023 to 'the cost that the insurer expects to pay', and in paragraph 37 of PS 300 to the need to have regard to the ' insurer's level of expenses, organisational structure and future administrative development' .

In his 31 December 1996 report64, Slee provided for FCHC at differing rates across the various Australian portfolios. The rates varied between 2 per cent and 7 per cent, giving an approximate weighted average of just over 4 per cent. But in each of his subsequent reports, Slee provided for FCHC at the lower and fixed rate of 2 per cent of gross liabilities across HIH's entire global portfolio of liabilities. His reports

made it clear that he was instructed to provide for FCHC at this rate.

Slee's evidence was that he was first instructed to provide for FCHC at 2 per cent during a conversation with Fodera in January 1998. 65 Fodera told him that he considered the previous allowances for FCHC to be too high, and that 2 per cent of

The f ailure of HIH Insurance 303

gross liabilities would be an appropriate allowance. Fodera did not give any reason for this view, but Slee's recollection was that Brown was present at the time and told Slee that he believed he could manage the claims in the professional indemnity portfolio for less than 2 per cent. Slee did not know, or ask, whether Brown's view was backed by any analysis. 66

Slee said that upon receiving this instruction he suggested that Fodera obtain an independent report to support his view as to an appropriate allowance for FCHC. Slee said that while he did not make any recommendation as to an appropriate person to retain for this purpose, nor as to the instructions which might be given67 to that person, he and Fodera did discuss having the assessment performed on the basis that a third party would manage the claims.68 This led to HIH obtaining a report from Roy Walker. I will return to the Walker report shortly.

In his 31 December 1997 report (being the first report where a rate of 2 per cent of gross liabilities for FCHC was used) Slee included the following disclosure:

The calculations include the present value of the future cost of administering claims on the basis of a report commissioned by an independent consultant with the following terms of reference.

Ascertain a fair value of future claims handling costs as at

31 December 1997 in respect of the claims reserves of nominated portfolios of the HIH Winterthur group. The review should take into account both reported and unreported claims as at 31 December 1997 and be based on the assumption that claims management would pass to a third party claims management provider under a competitive tendering process.

This has led to the conclusion that approximately 2% of claims cost is required for future claims handling of incurred claims. It should be pointed out that, if the costs of the company are not so contained, the company will incur an expense over-run. 69

Slee ' s evidence was that his instructions in relation to the allowance for FCHC were not subsequently altered, and he included equivalent disclosures in the terms of reference of each of his subsequent reports. 70

Although the evidence does not reveal the precise allowance for FCHC in HIH's accounts for the period from 31 December 1997 until the collapse of the company, it appears to have been slightly less than 2 per cent of the gross liabilities. But regardless of the precise allowance for FCHC in HJH's accounts, as Slee's valuation was used to test the adequacy of the booked provision, it is significant that the allowance made by Slee was 2 per cent of gross liabilities.

In my view, an allowance of 2 per cent of gross liabilities was inadequate. It tended to under-state HIH's OCP as at 31 December 1997 and each subsequent balance date. An allowance of 5 per cent of gross liabilities would have been more appropriate. A 3 per cent difference is significant when it is considered that in HIH's consolidated accounts for the periods ending 30 June 1999 and 30 June 2000

it included gross outstanding claims liabilities of $3 699.5 million and

304 Under-provisioning for claims

$4 430.9 million respectively. 3 per cent of these amounts would be $111 million and $133 million respectively.

The detail of my reasons for so concluding follow , but essentially they are that the industry standard or benchmark allowance for FCHC for a general insurance company like HIH was in the order of 4 per cent to 6 per cent. The allowance of just over 4 per cent in Slee's report as at 31 December 1996 was within this range . The Walker report did not provide an appropriate basis for departing from the industry standard in assessing HIH's likely actual FCHC, and there was no other analysis of information which otherwise provided an appropriate basis for departing from the range.

The industry standard or benchmark allowance Both Pearson and Wilkinson gave evidence of an industry standard or benchmark allowance for FCHC by a general insurance company. Pearson regarded a range of 4 per cent to 6 per cent of expected future gross payments as a 'benchmark' allowance for FCHC. 71 Wilkinson referred to a usual range of 4 per cent to 8 per cent of gross claims.72 Neither Pearson nor Wilkinson was aware of any claims

handling practices or other factors which would enable a general insurer of the size ofHIH to reduce its claims handling costs below these standard ranges .

In their submissions, both Andersen and Slee cautioned against reliance upon any industry standard, pointing out that on occasions significantly lower allowances were made and were appropriate. Both referred, by way of example, to the use by Christopher Latham of an allowance of as low as 2.5 per cent for FCHC in valuing FAI's long-tail business. 73 In his closing submissions, Slee also referred to the PWC actuarial review of HIH's CTP portfolio which had included an allowance for FCHC of 2.5 per cent.

It may be accepted that on occasions a figure below the suggested benchmark allowance will be appropriate. But this does not mean that an industry standard does not exist, or is not of any significance. As Pearson explained during her cross­ examination by Slee's counsel 74, if an actuary were instructed to use an allowance that fell below the benchmark rates, then the actuary would need to do some testing, or need to be convinced that the instruction was appropriate. In my view, as the allowance for FCHC which Slee was being instructed to assume was materially

below the industry standard, it was incumbent upon him to satisfY himself of the appropriateness of the percentage figure.

The Walker report The independent consultant's report referred to by Slee in his 31 December 1997 report was a document prepared by Roy Walker of RN Walker Pty Limited, dated 22 January 1998. 75 The Walker report consisted of a three page letter, together with

a number of appendices. It was headed 'Draft for discussion only ' and commenced:

We refer to your request that we advise on the fair value of future claims handling costs (FCHC) as at 31 December 1997 in respect of the Claims

The failure of HIH Insurance 305

Reserves of nominated Portfolios of the HIH Winterthur Group as detailed in your facsimile of 20 January 1998.

It then set out a series of assumptions upon which the report was based and the author' s opinion as to the costings which would be obtained from independent run­ off managers (ROM) in a competitive tender environment.

The findings of the Walker report were summarised in the following te1ms:

You will note:

(a) We have combined certain Classes to make economic packages I.e.

Professional Indemnity and Marine (E)

Combined Liability (F)

CTI Workers Compensation and Disability (H)

We have indicated our opinion of conservative costing with a possible reduction factor for competitive tender. Clearly the Tenderer following examination of the claims may consider further savings can be made.

However, for your purposes the COMPETITIVE TENDER amounts shown onE, F, G, Hare in our opinion reasonable.

E $4 million

F $8 million

G $2.2 million

H $4.5 million

Total $18.7 million

The report was signed by Walker, as managing director ofRN Walker Pty Limited.

To the extent that it suggested claims handling costs of $18.7 million in respect of the approximately $1 billion in gross liabilities considered by its author, it might be said that the Walker report supported an allowance for FCHC of slightly less than 2 per cent of gross liabilities. But, upon closer analysis, the report did not support the adoption of an allowance of 2 per cent for the FCHC of HIH's global OCL. There are a number of reasons for this.

First, the Walker report assumed a hypothetical run-off by independent run-off managers in a competitive tender environment. But HIH did not intend placing its business into run-off, or putting its claims handling function out to tender. Accordingly, the Walker report was not directed to the task required by AASB 1023 and PS 300-namely, an assessment of HIH's likely actual claims handing costs.

76

It would not be surprising that an independent run-off manager (presumably with a centralised claims handling function and no desire to maintain an ongoing relationship with its policyholders) could achieve cost savings that might not be

306 Under-provisioning/or claims

achievable for an insurer with its claims handling personnel spread across the world and with no intention of ceasing its ongoing operations.

Second, the Walker report only dealt with a portion ofHIH's business. In particular, it made no assessment ofHIH's likely FCHC in respect of its overseas and short tail portfolios. It thus provided no basis for an allowance for FCHC of 2 per cent of gross liabilities in relation to those portfolios.

Third, the Walker report was directed only at an appropriate allowance for FCHC as at 31 December 1997. Despite HIH' s rapidly changing and expanding business, the Walker report was not updated to take account of these changes.

Finally, the 2 per cent allowance for third party handling costs did not allow for any remaining costs of claims handling to HIH. Even if the claims handling function were put out to an independent manager it would be unlikely, given the importance of the function, that there would be no ongoing oversight by HIH.

Pearson also suggested that a further reason for rejecting the conclusions of the Walker report was that it was based upon a targeted 500 claims per claims officer in year one. In her view a target in the order of 150 to 200 would be more realistic. During cross-examination by Slee's counsel, however, it was put to Pearson that

there was some evidence to suggest that HIH's claims officers, at least in some portfolios, had been handling up to 500 claims per year. This criticism of Pearson's approach, even if valid, does nothing to undermine the other reasons for rejecting the Walker report.

Slee's adoption of a 2 per cent allowance for FCHC I have referred earlier to the circumstances in which Slee was first instructed to assume an allowance of 2 per cent of gross liabilities for FCHC. Slee said that he was satisfied that the dollar amount represented by 2 per cent of gross liabilities was an adequate allowance for FCHC. But it is not entirely clear to me the basis upon which he satisfied himself of this as at the time of the preparation of his various reports.

Slee appears to have relied primarily upon his experience (both as an actuary, and in the management of insurance companies) and estimate of the number of claims, the staff required and the speed at which claims would run-off. 77 He was not provided with any statistics or analysis which would have enabled him to assess the actual costs of claims handling. Indeed, he was informed that none were available. 78

From the face of his reports it would appear that Slee relied upon the Walker report in accepting the appropriateness of the 2 per cent allowance. I note, for example, the reference to that report in the passage from Slee' s 31 December 1997 report extracted above. In his initial statement provided to the Commission, Slee wrote that

he certainly accepted that Walker had the experience and expertise to express the views that he did, and that 'of course the costs had to be assessed on the basis of a run-off because the FCHC only applies to the costs of handling existing claims into the future' . He considered that the competitive tender basis provided 'the only real

The failure of HIH Insurance 307

way that Walker could approach the matter'. Slee continued that having reviewed the Walker report, he concluded on the basis of that report and his own assessment that 2 per cent was a fair central estimate. 79

In his subsequent statement and oral evidence, Slee tended to downplay his reliance upon the Walker report. He said that it was not the 'be all and end all of my calculations' .80 He later said that while he used the report ' to some extent' in carrying out his own calculations of FCHC, ' it was not a very good report'.81 He acknowledged that PS 300 required an assessment of the insurer's actual costs, and not some hypothetical cost. 82 He also acknowledged that his approach which assumed a run-off 'was not totally consistent' with PS 30083 , and that 2 per cent was

'certainly not' a central estimate of HIH's anticipated actual costs. 84 But he did maintain that as he did not have the actual company figures, there was no other way of considering the cost than starting with a run-off situation. 85

Slee also suggested that he relied to some extent upon the advice by Brown in 1998 that he was able to manage the claims in the professional indemnity portfolio for less than 2 per cent. 86 In his closing submissions, Slee referred to this advice and Brown' s oral evidence that not only were actual claims handling costs for the professional indemnity portfolio in the order of 2 per cent or less, but also that Peter Bacsak had calculated the actual costs for the liability portfolio (which Brown understood were about 2.2 per cent), and that Brown imagined that FCHC for the

short tail classes would be much lower than 2 per cent. 87

In my view, the assertions by Brown and Bacsak do not take the issue much further. It was not suggested that Slee was aware at the time of signing his various reports of Bacsak's views in relation to FCHC for the liability portfolio, or that he relied upon these views in assessing the reasonableness of the 2 per cent allowance. But in any event, in my view, in the absence of any analysis backing them, such assertions do

not provide an adequate basis for departing from the industry standard range of 4 per cent to 6 per cent. As mentioned, Slee's evidence was that he did not know whether Brown ' s assertions were backed by any analysis. The fact that Slee was also informed that no statistics as to actual claims handling costs were available suggests to me that no such analysis had been carried out, or, if it had been, it was not sufficiently rigorous to support the 2 per cent allowance. Certainly, Slee did not carry out, and was not aware of anyone else carrying out, any such analysis.

I also note in relation to the suggestion by Brown that the FCHC for short tail classes would be much lower than 2 per cent that, in concluding that a range of 4 per cent to 6 per cent of gross liabilities was a benchmark allowance for FCHC, Pearson referred to the detailed expense allocations in which she had been involved. Her evidence was that they tended to show FCHC of 4 per cent to 8 per cent for

short tail classes. 88

308 Under-provisioning for claims

In the course of preparing his valuation as at 31 December 1998, Slee wrote to Cassidy and Fodera by letter dated 4 February 1999, suggesting that:

Generally YOUR total net discounted reserve appears to be over discounted and under-provided for in FCHC. I am not saying at this stage that your figures are unacceptable but they are at the very lowest end of acceptability, and would not pass muster by Australian actuarial standards. One solution is to increase the FCHC by at least $5 million .. . 89

Slee then wrote to Jeffrey Simpson by letter dated 13 February 199990 in which he observed that HIH proposed booking a net discounted figure of $4 7 million for FCHC, and that he believed this number was hard to justify on the basis of the expenses that were being incurred. He wrote that he did not regard that as a central estimate of HIH's actual costs and proposed that an independent opinion be obtained. 9 1

By memo dated 16 February 199992 , Simpson informed Slee that an updated review by Walker would be requested for the 30 June 1999 audited accounts. Slee never received any such updated review. Walker confirmed that he was never requested to provide, and did not provide, any oral or written advice to HIH in relation to FCHC

beyond that set out in his initial draft report. 93

The 16 February 1999 memo from Simpson included the following, under the heading FCHC:

An exercise conducted by Roy Walker based on December 1997 reserves concluded that Australian long tail reserves could be run off for an undiscounted cost of $18. 7m. This compares to a current discounted claims reserve of $25 .2m for these portfolios. The attached schedule (F)

indicates that there is a 37% margin on these figures at a discounted level.

In his supplementary statement, Slee wrote that having considered this comparison (and the suggested 37 per cent margin in the booked FCHC over that recommended by Walker), he considered the HIH allowance for FCHC reasonable.

In my view, there is a number of reasons why Simpson's suggestion of a 3 7 per cent margin provided no support for a 2 per cent allowance for FCHC.

As Pearson explained94 , the schedule attached to Simpson's memo (and referred to in the above extract from that memo) included 'booked net FCHC' of

$46.990 million, which in tum included $25.243 million for the specific portfolios the subject of the Walker report. It was this $25.243 million which was said to provide a 37 per cent margin over Walker's figure of $18.7 million for those same portfolios. But Walker's figures related to the position for those portfolios as at

31 December 1997, and there was no analysis by Simpson of the extent to which the OCL for those portfolios had grown between December 1997 and December 1998, and whether this growth would require a higher allowance for FCHC. Further, there were portfolios beyond those covered by the Walker report for which the booked figure set out in the schedule did not include any allowance for FCHC, and in

relation to which it would be reasonable to expect some claims handling costs. In

The failure of HIH Insurance 309

other words, while in respect of the specific portfolios considered by the Walker report there may have been a 37 per cent margin (over the Walker figures , being approximately 2 per cent of the OCL of the specified portfolios as at

31 December 1997), it could not be said that a 37 per cent margin (over the 2 per cent allowance implicit in the Walker report) existed across HIH's entire business.

I also note that it is apparent from the schedule attached to the Simpson memo that in respect of the specific portfolios the 'actuary' allowance of $20.618 million was significantly less than the proposed booked figure of $25.243 million. In other words, Slee's figures for those portfolios did not include the 37 per cent margin referred to by Simpson.

But perhaps the most fundamental problem with any reliance upon the 3 7 per cent margin suggested by Simpson is that it would involve circular reasoning. The purpose of Slee's valuations was to provide a figure against which to test the adequacy of HIH's proposed booked provisions. Unless the proposed booked figure was based upon an analysis of HIH's likely actual costs, it would not justify Slee taking comfort in his 2 per cent allowance for FCHC from the fact that the proposed booked figure was in excess of this amount in respect of certain portfolios. Consider, for example, the situation if HIH had included an allowance of say 10 per cent for FCHC in its proposed booked figure. If Slee's estimate of the group OCL (based on a 2 per cent allowance) was approximately equal to this proposed booked provision, rather than suggesting the booked provision was adequate, the coincidence ofthese figures would suggest that the company's estimate of the OCL (net of FCHC) was inadequate. I do not think it is appropriate to rely upon a proposed booked figure as providing comfort in selecting an allowance for FCHC for use in arriving at an actuarial estimate, when that actuarial estimate is intended to be a test of the adequacy of the proposed booked figure.

I note that in his closing submissions Slee added that he relied not simply upon the suggestion of a 37 per cent margin, but rather on the entire exchange of correspondence between Slee and HIH in February 1999. This exchange was not limited to a consideration of the Walker report. Slee referred to Simpson's letter dated 10 February 199995 as listing a number of issues, and in particular the PWC actuarial review of the CTP portfolio which had included an allowance for FCHC of 2.5 per cent of gross liabilities. In my view, there is nothing in that correspondence which provided any basis for a 2 per cent allowance. The fact that PWC had adopted a rate of 2.5 per cent in its valuation of one particular portfolio does not deny the existence of a general industry standard allowance in excess of this amount. Nor does it provide any support for an allowance of 2 per cent in respect of other portfolios, or indeed HIH's entire global portfolio of business.

For these reasons I consider that the 2 per cent FCHC allowance that HIH instructed Slee to use was unlikely to represent a central estimate of the cost of administering and settling claims. 96 I have not been persuaded that there was sufficient analysis of HIH's actual claims handling expenses to justify the adoption of a figure that was materially low relative to industry benchmarks.

310 Under-provisioning/or claims

Andersen's consideration of FCHC In its 31 December 1997 audit, Andersen reviewed the Walker report, noting that the actuary had merely relied upon this analysis.97 The revi ew noted that the Walker report assumed a run-off manager. It recorded that Simpson had confirmed to Andersen that HIH was not actually planning to employ a run-off manager, but

rather the Walker report had only been used to justify the percentage used by the actuary to estimate FeHe. The review also recorded that no overseas or short tail portfolios had been included in the Walker report. Andersen understood, therefore, that Slee had based his 2 per cent allowance for FeHe upon a report which did not cover HIH's entire business and which had been prepared upon a basis other than

required by AASB 1023 (namely, an estimation of the particular insurer' s likely actual FeHe).

The Andersen review noted that the actual costs booked by HIH fo r the liability (Australia), professional indemnity, workers compensation, ere liability, eMG and eTP portfolios were approximately 4 per cent, and thus higher than the figure suggested by the Walker report for those portfolios. Further, they were almost $4 million greater than the figure included by Slee in respect of those same portfolios. There was no analysis, however, of the FeHe booked across the remaining portfolios of business. The review reached the following conclusions:

FCHC' s booked by HIH are considerably higher as a proportion of discounted reserves than the tender prices advised by RN Walker. This would lead to the conclusion that the actuary has unduly relied on this analysis, and thus we should consider the impact on the actuary' s estimates of increasing the FCHC to a balance more comparable to the actual cost booked by HIH (approximately 4%) ...

The difference between the total FCHC booked by HIH and that calculated by the actuary is $3.8m. ($27 .7m-$24m) The difference will be highlighted in the analysis of total claims provisions between HIH , the actuary and ourselves. Any differences will be dealt with at that level. Pass further work.

In relation to the 31 December 1998 audit, an internal Andersen memo dated 16 February 1999 recorded:

The ... percentage of 2.48% compares with 2% of reserves recorded by the actuary in the prior year. Based upon discussion with Jeff Simpson (financial controller), HIH expects FCHC in dollar terms to increase slightly with the acquisition of FAT , however it should decrease as a proportion of reserves as it will not increase by as great a percentage as reserves. If the 2.51% advocated by the actuary were to be applied against the actual reserves, FCHC of $47,650,000 would be calculated, only $660,000 than has been booked. This is not a mate rial

adjustment. 8

Andersen did not review FeHe after 31 1998. It relied upon work done in previous audits. 99

The failure of HIH Insurance 311

Andersen's work in relation to FCHC suffered from a similar problem of circularity to that which I have referred in the context of Slee's reliance upon Simpson's suggestion of a 37 per cent margin. In circumstances where neither the actuarial figure nor the proposed booked figure were based (or even purported to be based) upon any rigorous analysis of HIH's likely actual FCHC, the fact that the two figures were immaterially different, or that the booked figure exceeded the actuary figure, is essentially meaningless. The comparison says nothing of the accuracy or appropriateness of either figure. The critical issue was whether the 2 per cent allowance by Slee and the dollar sum booked by HIH for FCHC was an adequate allowance for HIH's likely actual costs of handling claims in the future.

In addition to this circularity there is a further problem with the Andersen work. In its review as at 31 December 1997, Andersen concluded that the actuarial allowance may have relied too heavily upon the Walker report and that consideration should be given to the impact of increasing the actuary's allowance to a balance more comparable to the costs booked by HIH (approximately 4 per cent). Yet in the Andersen memo dated 16 February 1999 comfort was taken from what was said to have been an immaterial difference between the allowance recommended by the actuary and that proposed to be booked by HIH. No consideration was given to the adequacy of the allowance recommended by the actuary, despite the doubt cast over its adequacy in the 31 December 1997 audit.

In its closing submissions, Andersen referred to the criticism that its analysis at 31 December 1998 was based upon HIH' s booked provision for FCHC rather than HIH's actual claims handling costs. It submitted that this criticism ignored the representation made by Simpson to Andersen in relation to the adequacy of the provision for FCHC, referred to in the above extract from the Andersen memo dated

16 February 1999. 100 Andersen further submitted that in the face of this representation it had no reason to believe that HIH's provision for FCHC was anything other than a genuine assessment of HIH's expected costs. However, there is no evidence to suggest that Andersen either requested or obtained from Simpson any analysis of HIH's expected costs. In those circumstances, the statement attributed to Simpson was assertion and in my view adds little to Andersen's work

in relation to the FCHC.

Andersen also submitted that as FCHC relates to the costs of settling liabilities into the future, it is the equivalent of a run-off scenario. For the reasons mentioned earlier- namely that the run-off scenario contemplated by the Walker report assumed a centralisation of the claims handling function and an absence of any desire to maintain an ongoing relationship with policyholders, which it would not have been appropriate for HIH to assume-! reject this submission.

Andersen also referred to Latham and his evidence that the range of FCHC allowances could be from 1 per cent to 5 per cent depending on the class of business, and that he had used an allowance for FCHC as low as 2.5 per cent for some of FAI's long-tail classes. It was submitted that without further information, cross-company comparisons were not meaningful and there was therefore no basis

312 Under-provisioning for claims

on the evidence for a finding that HIH's allowance should have been 4 per cent to 5 per cent. Again, while I have accepted the expert evidence of Pearson and Wilkinson to the effect that there exists a benchmark range of allowances for FCHC, this does not mean that in certain circumstances lower allowances are not appropriate. In my view, it simply means that the adoption of an allowance outside of this benchmark range would generally require some form of analysis of the particular insurer's likely actual costs. That is what was missing in the case ofHIH.

Finally, Andersen submitted that no consideration was given by Pearson or Wilkinson to the fact that HIH included its direct claims handling costs (such as legal and other professional fees) within its case estimates such that its allowance for FCHC was only for indirect claims settlement costs. Andersen referred by way of example to Wilkinson's evidence in relation to a discussion he had with Brown in which Brown suggested a figure of 5 per cent for HIH's direct (rather than indirect) claims handling costs. It may well be that Brown was referring to direct claims handling costs when quoting a rate of 5 per cent to Wilkinson, but then speaking of

indirect claims handling costs when referring to a 2 per cent allowance in his conversation with Slee in early 1998, and in his evidence before the Commission. This may explain the apparent inconsistency between the two percentages. At the same time, I do not accept that there is any basis for suggesting that Pearson or Wilkinson were dealing other than with indirect claims handling costs when

referring to the existence of benchmark or industry standard ranges of allowances. Pearson referred, for example, to FCHC as typically including internal claims handling department salaries, property and other operating costs, as well as a proportion of human resources, central management, finance and information

technology costs relating to claims handling. 101 She did not include in this list any direct claims handling costs such as legal or other professional fees.

In all the circumstances, I am not satisfied that Andersen took a sufficiently rigorous approach to the issue of FCHC in its various audits of HIH.

15.4.2 Discounting

General I have mentioned in the section of this report dealing with matters of general principle that the relevant professional standards in Australia (AASB 1023 and PS 300) provide for the discounting of a general insurer's OCL using a market­ determined risk-adjusted rate of return appropriate to the insurer. This seems to allow a departure from the risk-free rate of return (as represented by the yield on a Commonwealth Government security of a term matching the term of the insurer's

liabilities) where the insurer's assets could reliably be expected to give rise to higher rate of return. 1 02

The discount rates adopted by Slee, upon instructions from HIH, were consistently higher than the yields available on Commonwealth Government bonds with a three year maturity (three years being roughly representative of the mean term of HIH's

The failure ofHJH Insurance 313

OCL). 103 At the same time, Slee made no allowance for HIH's exposure to the risk arising from the mismatch in the term ofHIH's assets and liabilities. Again, this was upon instructions from HIH. In my view, this represented a relatively optimistic or aggressive approach to discounting by HIH and Slee. But, I do not criticise Slee for this. It appears that Slee was provided with information concerning the returns which HIH expected to achieve on its assets which supported the rates he was instructed to use, and was provided with certification that HIH's assets were sufficient to support the estimate of the OCL. In any event, Slee set out in clear terms in his reports the discount rates which he had been instructed to adopt, and had adopted, together with reference to the real risk associated with the mismatch of assets and liabilities, and his instruction that he make no allowance for that risk. To the extent that this approach to discounting was optimistic or aggressive, this would have been apparent to anyone reading Slee's reports. Indeed, the existence of the

mismatch risk identified by Slee is one matter which might have inclined a reader of Slee's reports to a view that a prudential margin was appropriate.

I consider that the real problem with discounting lay not so much in the rates which Slee was instructed to adopt, or in his (and therefore HIH's) adoption of the relevant interest rates without any allowance for the risk of mismatch, but more so in the fact that the relevant professional standards permitted (if not required) this approach. It is a further problem that a number of the directors, who did not read Slee's reports and the warnings they contained, was unaware of the aggression inherent in the approach taken.

Discount as at 31 December 1997 A particular concern arose in relation to the discounting of HIH's OCL as at 31 December 1997. Note 19 to HIH's consolidated accounts for the year ending 31 December 199i

04 referred to HIH' s OCP as at 31 December 1997 as discounted at a rate of 6.2 per cent over an 'average weighted term to settlement' for the outstanding claims liabilities of 2. 7 years.

In a memo dated 2 April 1998 and addressed to Cassidy 105 , Slee disclaimed any responsibility for the contents of this note on the basis both that it contained a mathematical error, and that on the figures given to him it would be difficult to justifY a discount rate of 6.2 per cent. He wrote that certainly his calculations had

not been done on that basis. Rather he had used a rate of 6 per cent, being the rate which management had agreed was the actual rate being earned on the company' s assets. In his evidence, Slee said that having written this memo, he did not subsequently discuss the matter with anyone from management or the auditors. Nor did he refer to the matter in his report (which was attached to the memo referring to the issue). His reason for not including the matter in his report was that he believed the problem lay in the peculiar manner in which HIH had derived the undiscounted figure , and that his only task was to test the net discounted provision. 106

In relation to the mathematical error, if $1849.9 million (being the reported undiscounted provision of $2377.3 million less the reported undiscounted reinsurance recoveries of $52 7.4 million) were discounted by 6.2 per cent over a

314 Under-provisioning/or claims

period of 2. 7 years, the result would be an amount of some $62 million less than the discounted provision referred to in the accounts of$1 510.5 million.

Cassidy's evidence 107 was that following receipt of the memo from Slee he spoke to Fodera. Fodera informed him that although financial services division had used a slightly different model to Slee, and hence a different undiscounted total, nevertheless the discounted or 'bottom line' totals were the same such that the difference had no effect on the reported profit. 108 On the basis of this explanation, Cassidy took the issue no further. 109

Fodera's evidence 110 was that he recalled being shown the 2 April 1998 memo by Cassidy, and hence was aware of the issue raised by Slee. 111 He said that he did not do anything at the time-that is, in April 1998-to check the accuracy of the discount. His reason for not taking any action was that while there may have been a disclosure error in the note to the accounts, nevertheless the net reserves as calculated by the company and as actuarially assessed by Slee 'tied in' .112

Andersen denied there was any error in the note. It suggested that the explanation lay in the fact that the discount used by HIH involved an amalgamation of a number of different discount calculations for the different portfolios of HIH's business. 113

I remain concerned as to the accuracy of note 19 to HIH' s consolidated accounts for the period ended 31 December 1997. Paragraph 11.4 of AASB 1023 required disclosure of the 'average' discount rate and the 'average weighted term' . In my view, even if the rate and term referred to in the note to the accounts were

amalgams, this should not have deprived them of their quality as averages. If AASB 1023 were complied with, then an application of the rate and term to the undiscounted provision in the accounts should produce the discounted provision. Both Slee and Jon Pye agreed that this calculation produced a figure some

$62 million less than the discounted provision in the accounts. 114 Even if the amalgam or average rate involved some level of estimation, I doubt whether this element of estimation could account for the entire $62 million discrepancy.

It appears to me that the HIH accounts as at 31 December 1997 involved some form of error or misstatement. There are several possible explanations, including an overstated undiscounted provision, an understated average rate, an understated weighted average term and an understated discounted provision. If the discounted provision was incorrect, then this would mean that HIH's reported profit was also

incorrect. Each of Fodera, Cassidy and Pye denied that the problem lay in the calculation of the discounted provision. Although I have some reservations in accepting the explanations proffered by them, the state of the evidence does not allow me to make any finding as to where precisely the error or misstatement lay. It

follows that I am not in a position to determine whether the effect of the error was to overstate HIH's reported profit. In those circumstances I do not propose to consider further the issue of where responsibility for the apparent error lay.

The failure of HIH Insurance 315

15.4.3 Prudential margins

The desirability and impact of the inclusion of a general prudential margin within the OCP, together with the professional standards governing the same, have been considered earlier.

In its 31 December 1997 accounts, and subsequently, HIH adopted a policy of not including a general prudential margin in its OCP. Similarly, Slee was instructed not to include any prudential margin in his valuations as at 31 December 1997 and subsequently.

Slee's instruction not to include any prudential margin In his valuation as at 31 December 1996 (which was not a valuation ofHIH' s entire global OCL ), Slee included prudential margins for the Australian liability and professional indemnity portfolios of 1 0 per cent of the net OCL. 115 In his valuation as at 30 June 1997, Slee nominally included similar prudential margins. I say

' nominally' because as Slee pointed out in a memo to Fodera dated

12 August 199i 16 , these prudential margins (totalling $24.7 million) were in fact

largely offset by 'negative' prudential margins in certain other portfolios (totalling $16.2 million). S1ee went on to say in that memo:

the reality is that considering the portfolio as an entity, the reserve quoted is the central estimate, and it would be unwise to suggest that there was any prudential margin in the aggregate.

Slee was first instructed to exclude any prudential margin during January 1998, while preparing his valuation as at 31 December 1997. The instruction was initially given orally by Fodera117, but then confirmed in a letter from HIH (signed by Ross Eade) dated 20 January 1998. 118 Fodera informed Slee that the reason for not including a prudential margin was that the decision to do so was the prerogative of the directors.11 9 Slee said that he accepted that the inclusion of a prudential margin was the prerogative of the directors, but nevertheless continued to warn the directors in his reports and other correspondence of the need for such a prudential margin, and the consequences that could follow from the failure to include such a margin. 120

In each of his reports including and following the 31 December 1997 report, Slee made it clear that no allowance had been made for a prudential margin. The reports followed a relatively consistent format, with each referring to the instruction given to Slee not to include any prudential margin and at the same time warning of the risks inherent in following that course. Although the reports followed a relatively consistent format, the warnings in relation to the risks inherent in his valuations, and hence the risks associated with the absence of a prudential margin, became more strident in the 30 June 2000 and 31 December 2000 reports.

In Pearson' s opinion, there was no professional obligation preventing Slee accepting the instruction not to include any prudential margin within his valuations from 31 December 1997. 12 1

Having examined Slee' s disclosures in relation to uncertainty and prudential margins in his 30 June 1999 and 30 June 2000 reports, Pearson concluded 122 :

316 Un der-provisioning for claims

Regarding the discussion on uncertainty, my own interpretation of the professional standard would be that it required a greater level of quantification of certain aspects of uncertainty in the liabilities than provided by Mr. Slee. However, I believe that differences in

interpretations of this aspect of the standard are possible and conclude that Mr. Slee's discourses on uncertainty probably satisfied professional obligations.

Regarding Mr Slee's warnings on the exclusion of a prudential margin, these are clearly stronger at June 2000 than June 1999. In fact, Mr Slee changes his advice from agreeing that not holding a prudential margin is 'acceptable' to recommending the inclusion of a prudential margin. This change in recommendation is consistent with the concerns expressed by Mr Slee in the 30 June 2000 report about the company's performance and

the impact of the Swiss Re contract.

I make no criticism of Slee for accepting the instruction to prepare his valuations on a central estimate basis, and therefore without any prudential margin.

HIH's decision not to include any prudential margin

Regardless of the instruction to the actuary, a question arises whether it was appropriate for the company not to incorporate a prudential margin within its OCP.

The ultimate decision as to whether a prudential margin should be so incorporated was one for the directors. Because the function of a prudential margin is to protect the company against the risks inherent in the estimate of its OCL, any consideration of whether it was appropriate to incorporate a prudential margin ought to have

involved consideration of the nature and extent of those risks. A natural starting point for an understanding of those risks would, of course, have been the reports prepared by Slee. In my view, even a relatively cursory review of those reports would have suggested a relatively high degree of uncertainty and a need for caution.

Slee provided valuations which he was prepared to put forward as central estimates as at the various balance dates. At the same time he made it clear in each of his reports that his valuations were heavily underpinned by the instructions he had been given, and hence assumptions he had made. Those assumptions included, for example, an allowance of 2 per cent of gross liabilities for FCHC, the full

recoverability of reinsurance, and the absence of any allowance for the mismatch of assets and liabilities. He also placed reliance upon the work of other actuaries and case estimates put forward by management with respect to a number of significant portfolios of business. Slee warned in his reports of the need for, and desirability of, a prudential margin in the face of risks such as these. In my view, in the absence of a more detailed appreciation of the risks inherent in the various assumptions made

by Slee and his valuations more generally, best practice suggests that the directors ofHIH should have resolved to include a prudential margin within their OCP.

This, of course, assumes that AASB 1023 permits the inclusion of a prudential margin within an insurer's OCP. I am not entirely sure that it does. But even if it does not, it is apparent from Pearson's evidence of the market practice that the industry treats AASB 1023 as permitting the use of a prudential margin.

The failure of HIH Insurance 317

I have considered the accounting treatment of a prudential margin elsewhere in this report. For present purposes, I am not so much concerned with the precise manner in which the risk inherent in the valuation ofHIH's OCL was addressed (for example, through a prudential margin, solvency margin or capital surplus) as with ensuring that it was addressed.

It seems to me that it would not have been possible for the directors to address adequately this risk if they were not aware of its nature and extent. I will return to this issue in my consideration of the directors' role in setting the OCP towards the end of this Chapter.

Market practice Pearson 123 wrote in her statement that her understanding of industry practice in relation to prudential margins was based on a survey conducted by her firm, Trowbridge Consulting, in 1998. The ten insurance companies which were the subject of the survey between them covered two-thirds of the market by premium volume. Pearson summarised the results of the survey in the following terms:

• All companies surveyed held prudential margins in their outstanding claims provisions for balance sheet purposes in respect of long-tail classes, with 64 per cent of companies holding a prudential margin in respect of short tail classes.

• Over 50 per cent of companies surveyed held prudential margins that were between 10 per cent and 20 per cent of the central estimate for long-tail classes.

• Over 50 per cent of companies held prudential margins that were 5 per cent or less than 5 per cent of the central estimate for short tail classes.

• Most companies used a probability of sufficiency of provisions approach in setting prudential margins, although a significant minority used a more qualitative assessment.

Reinsurance as a justification for exclusion of a prudential margin At various stages in the evidence before the Commission reference was made to the existence of HIH's reinsurance programme as a justification for the policy not to include any prudential margin in HIH's OCP.

Pearson 124 acknowledged that existence of aggregate excess-of-loss reinsurance was relevant to the determination of the size or need for a prudential margin. She explained that depending upon the exact nature of the reinsurance, the probability of the sufficiency of the actuarial central estimate of the OCL may be altered. However, Pearson added an important qualification. If such an approach were to be taken it would be necessary to evaluate whether there remained sufficient appropriate aggregate excess-of-loss reinsurance to maintain the desired level of sufficiency of the provision. She said that in the case of the HIH reinsurance arrangements, the existence of the cover at any point in time was dependent upon the continued payment of premiums. To the extent that the cover was to be funded

318 Under-provisioning/or claims

through future premiums, Pearson did not regard it as a substitute for a prudential margm.

Generation and release of specific margins Aithough the general policy adopted by HIH in its accounts for the period ended 31 December 1997 and subsequently was not to include a general or group prudential margin, there were various references in documents to portfolio specific prudential margins.

I am aware of four examples of what I have referred to as portfolio specific prudential margins. Two of these related to Australian portfolios. The other two related to the US operations. The latter two were greater in size and are considered briefly in the section of this chapter dealing with provisioning for the US operations.

Each of the portfolio specific prudential margins was introduced or generated through an accounting technique referred to in the evidence as acquisition accounting. This involved the creation of a prudential margin upon the acquisition by HIH of various entities, either through taking advantage of the differing accounting requirements in the United States and Australia as to the discounting of the OCL, or through an increase in the goodwill on acquisition. In each example,

however, the portfolio specific prudential margins were utilised or released within a few years of the relevant acquisition against pre-acquisition deteriorations in the OCL of the acquired business. In this way, the bottom line or reported financial performance of HIH was protected against these deteriorations.

It would appear that the accounting standards permit this form of accounting, so long as the prudential margins thus created are released against pre-acquisition deteriorations in the relevant portfolios of business, and not against more general deteriorations in the company's OCL. In the case of HIH, Andersen satisfied itself during the course of its various audits that the releases made by HIH did take place against genuine pre-acquisition deteriorations. I do not suggest otherwise.

From my point of view, the significance of the above is that the existence from time to time of various portfolio specific prudential margins was not a substitute for the incorporation of a general prudential margin within HIH's OCP. The portfolio specific prudential margins did not protect HIH's OCP against fluctuations in its

general OCL. They only protected HIH against the specific risk associated with pre­ acquisition deterioration in particular portfolios.

In any event, to the extent that those portfolio specific prudential margins existed, they were short lived and were exhausted within a few years after the time at which they were raised. In this way the existence of those margins tended to disguise the deteriorations in acquired portfolios of business which in fact occurred. By this I mean that without knowledge of the existence and release of these portfolio

specific prudential margins, an observer might have been misled by the bottom line performance of the acquired businesses.

Th e failure of HJH Insurance 319

15.4.4 Claims inflation

Both AASB 1023 and PS 300 require that the estimate of an insurer's OCL include an appropriate allowance for claims inflation. This includes both economic (that is, price or wage) inflation and superimposed inflation.

Slee's approach to superimposed inflation in respect of a number of portfolios was criticised by both Wilkinson (through the phase II report) and Pearson. In her witness statement 125

Pearson considered Slee's allowance for superimposed inflation in respect of a number of his individual portfolio specific valuations as at 30 June 1999 and 30 June 2000.

In his reports, Slee did not always distinguish between economic inflation on the one hand, and superimposed inflation on the other. This made it difficult to determine from the face of those reports the extent to which allowance was made for superimposed inflation in respect of the individual portfolios.

Pearson commenced her analysis with Table 15.2 (and footnotes), summarising the claims inflation assumptions adopted by Slee in his 30 June 1999 and 30 June 2000 reports, as well as those adopted in the phase II report.

Table 15.2 Slee's claims inflation assumptions, 30 June 1999 and 2000

30 June 1999 30 June 2000 KPMG

Claims Claims phase II

inflation Inflation report

Portfolio $million (%) $million (%) (%)

C&G Liability 126 239.7 4.0 242.8 4.00 9.0

FAI Corporate Liability 127 51.7 4.0 51.2 4.00 9.0

C&G PI (includes Law Society)12B 210.8 3.5 233.3 3.55 8.0

FAI Professional liability (includes MDU) 121 .7 4.0 141.6 4.00 8.0

Australian Workers Compensation 212.9 8.0 162.2 7.50 10.0

HIH America (excluding Great States) 126.6 129 nfa130 244.4 6.00 131 n/a

Having compared the Slee and the phase II report allowances for superimposed inflation, Pearson wrote 132 :

• Taking A WE (average weekly earnings, being a measure of wage inflation) to be around 4 per cent per annum, the phase II report included an allowance in the order of 4 per cent per annum for superimposed inflation in the valuations of the liability portfolios. Slee did not include any allowance for superimposed inflation for these portfolios.

• Taking CPI (the Consumer Price Index, being a measure of price inflation) to be around 3.5 per cent per annum, the phase II report included an allowance in the order of 4.5 per cent per annum for superimposed inflation in the valuations of the professional indemnity portfolios. Slee did not include any allowance for superimposed inflation for these portfolios.

320 Under-pro visioning for claims

• Taking A WE to be 4 per cent per annum, the phase II report included an allowance for superimposed inflation in the order of 6 per cent per annum for the Australian workers compensation portfolios compared with an allowance of 3.5 per cent in Slee' s 30 June 2000 valuation.

During her cross-examination by counsel for Slee 133 , it was put to Pearson that CPI (or price inflation) during 1999 and 2000 was running at less than 3.5 per cent; that an allowance of closer to 2 per cent would have sufficed; and that in assuming price inflation of 3.5 per cent, both the phase II report and her analysis, overstated the position.

It is likely that an allowance of a percentage point or so less than 3.5 per cent would have sufficed for economic inflation during 1999 and 2000. But given the concern was with future rates of inflation (after March 2001 ), I am not sure that criticism takes the matter far at all. I do not know whether the same criticism could be made of Pearson's assumed allowance for A WE (or wage inflation). But in any event, the criticisms made by Wilkinson and Pearson went beyond a difference of a percentage

point or so in the level of economic inflation. The phase II report suggested an extra 4 or 5 percentage point allowance for claims inflation should have been made. Pearson' s analyses of various individual portfolios are outlined later in this report, but generally speaking her criticisms were not referable to any difference in her view as to an appropriate allowance for economic inflation.

I also note that the criticism made by Slee's counsel is not entirely consistent with the approach taken by Slee. As mentioned, it was not always possible to discern from the face of Slee's reports the extent to which any allowance he made for claims inflation was attributable to economic inflation. However, in describing his approach to claims inflation in his witness statement134 , Slee said that in 2000 he considered that the central estimate for money losing its purchasing power over the foreseeable future (that is, economic inflation) was 2.5 per cent per annum. At the

same time, he also wrote that his view during 2000 was that there was no evidence of superimposed inflation in either the professional indemnity or liability portfolios.

Pearson considered in some detail Slee's valuations for a number of portfolios including C&G liability, FAI corporate liability, C&G PI, and Californian workers compensation. Her analysis of each of these portfolios is considered in the later sections of this chapter dealing with those individual portfolios. In summary Pearson's opinion was that Slee's assumed rates of superimposed inflation across these portfolios as at 30 June 1999 and 30 June 2000 were either optimistic (in the sense of being below the lowest reasonable level) or at the lowest reasonable level.

In that way, Slee's approach to superimposed inflation as at 30 June 1999 and 30 June 2000 tended to understate the central estimate of HIH' s OCL. For the reasons set out during the course of the portfolio specific sections of this chapter, I have preferred the opinions of Pearson over the explanations given by Slee (and in relation to some portfolios, by Cameron). It follows that I think Slee ' s approach to

superimposed inflation was optimistic in a number of respects, and did tend to understate the central estimate of HIH' s OCL as at 30 June 1999 and 30 June 2000.

Th e failure ofHIH Insurance 321

15.4.5 Actuarial techniques

In his witness statement135 , Slee explained that in carrying out his valuations he relied primarily on payments per claim finalised (PPCF) and payments per open claim (PPOC) models. He explained that these models were similar. He preferred them to a number of alternative models because they took into account not only the payments made at various durations but also the speed of finalisation of claims. Slee estimated that he relied upon PPCF or PPOC models in respect of 68 per cent of HIH's net OCL in 2000. Where he was not able to obtain the data necessary to carry out a valuation using a PPCF or PPOC model, he used either an expected loss ratio model or relied upon management or case estimates. Slee estimated that in 2000 he relied upon the former for 25 per cent ofHIH's net OCL (consisting primarily of the UK business, where this was the customary method of analysis). He estimated that he relied upon the latter for about 9 per cent ofHIH's OCL.

Both Wilkinson and Pearson criticised the actuarial techniques adopted by Slee in a number of respects. In particular they criticised:

• reliance upon payment-based valuation methods, with virtually no reference to exposure data (including policy and premium information) and profitability

• paucity of analysis of claims development, or the historical adequacy of claims, management opinions and case estimates

• reliance upon management opinions and case estimates

• reliance upon inadequate and out of date data.

I propose to address the first three of these criticisms individually, before making a general comment in relation to all three. I will consider Slee's reliance upon inadequate and out of date data in the following section of this chapter, which deals with the accuracy and flow of provisioning information.

Exposure data PS 300 provides that the selection of the most appropriate valuation model to estimate the liabilities is the responsibility of the actuary. The actuary may investigate more than one model before arriving at an estimate. The model or models should take into account the available data, the nature of the portfolio and the results of the analysis of experience. PS 300 also provides that it is the actuary ' s responsibility to ensure that the data utilised is appropriate and sufficient for the valuation. The actuary should, where possible, take steps to verify the overall consistency of the valuation data with the insurer' s financial records.136

Both Wilkinson and Pearson explained the importance and usefulness of exposure data (and in particular premium information). It acts as a check on the claims data and the valuations arrived at by the actuaries. Wilkinson said that while it was not unusual in Australia for only claims information to be given to the actuary, best practice required that such information be made available and be taken into account in any reserving analysis. 137 Pearson said that normal practice would be to obtain

322 Under-provisioning for claims

information on exposures as part of the valuation process, and that the actuary would need to consider whether the absence of such information prevented him or her forming a view as to the central estimate of the OCL. The actuary would also need to consider whether without this information the data supplied could be considered appropriate and sufficient for the valuation. 138

Slee's response to these criticisms was two-fold.139 First, he said he was not provided with premium data by HIH. Second, while he acknowledged that exposure data might give some indication of the level of exposure or the amount of business being written, he regarded it as a fairly unreliable indication. Generally he did not rely on premium information, or other forms of exposure data, when payment information was available.

Claims-development data PS 300 provides that an actuarial report should deal with 140 :

• the analysis of experience

• any changes in the method and key assumptions since the last similar report

• comparisons of actual experience with that expected under the assumptions in the last similar report.

Again, both Wilkinson and Pearson emphasised the importance of information concerning the historical adequacy of case estimates and claims. The latter is referred to as claims development information. They criticised the relative paucity of this information, and analysis of it, in Slee's reports. It is only through a consideration of this information and analysis, they explained, that the historical

adequacy both of case estimates and previous actuarial valuations can be assessed. Such an assessment is useful in determining the appropriate approach to current and future valuations, and in particular in determining whether further provision might need to be made by way of IBNER. Wilkinson suggested that given the significant deterioration which had occurred in the claims estimates for most components of the C&G, CIC and F AI business, it was unlikely that the provisions arrived at by the methodologies used by Slee would be at a 50 per cent confidence level. In his view

it would have been appropriate to include a margin for claims development. 141

In response to these criticisms, Slee queried the usefulness of any consideration of this historical information in circumstances where he did not rely upon those case estimates in valuing the majority of HIH's business, and where (particularly following the acquisition of F AI) he was dealing with a rapidly changing

business. 142

Reliance upon management case estimates Pearson and Wilkinson criticised Slee's reliance upon case estimates and management opinions in relation to a number of portfolios, and in particular, large PI claims (greater than $1 million), short tail, the run-off treaty account, Charman and film finance.

Th e failure of HIH Insurance 323

I will deal separately with Slee's reliance upon case estimates in respect of large PI claims in the latter section of this chapter dealing with the PI portfolio.

In relation to short tail, both Wilkinson and Pearson indicated their reluctance to rely upon case estimates 143 , but accepted that in certain circumstances it would be appropriate for an actuary to accept management' s case estimates. Examples include where the actuary had verified the appropriateness of the company's systems and procedures for dealing with and assessing claims reported 144, or where the actuary

had performed an analysis of the historical adequacy of the short tail case estimates and thereby determined them to be reliable. 145 In his reports as at 30 June 1999 and 30 June 2000, Slee referred to his reliance upon case estimates for the short tail business. There is nothing to suggest that Slee conducted any analysis of the hi storical adequacy of the case estimates for the short tail business. But he did seek 146 and obtain 147 a certification from Andersen, albeit in slightly unclear terms, as to the figure of $172 million which it was proposed be booked for the short tail business as at 30 June 1999. And one of the reasons Slee gave for not signing his 30 June 2000 report was the absence of any equivalent certification from the auditors for the period ended 30 June 2000.

In relation to the run-off treaty account, Pearson considered Slee's reliance (again apparent from the face of his reports) upon PWC's valuation prepared as at 31 December 1998 for certain F AI overseas business (including the MIPI, ALAS and MDU business) as at 30 June 1999 and his reliance upon underwriter' s estimates as at 30 June 2000. For 30 June 1999 Pearson said that Slee's approach appeared reasonable given that these portfolios were in run-off and had been valued

by PWC only six months earlier. On the other hand, Pearson thought that by 30 June 2000 it was no longer appropriate for Slee to rely on underwriters' estimates in arriving at a central estimate in accordance with PS 300. 148 Slee cited the existence of stop-loss reinsurance protection as a reason for a limited exposure, and hence the appropriateness of his reliance upon underwriters ' estimates. But Pearson opined this was an inadequate justification in the absence of any examination of the remaining availability of cover under those reinsurance contracts. 149 I will return to this matter in my later considerations of the F AI overseas business (referred to as the FAI problem claims) and the 'FAI Other less stop loss' item in Slee ' s 30 June 1999 valuation.

In relation to Charman and film finance, Slee again relied upon case estimates, and again made that reliance clear in his reports. Both of these risks were subject to significant legal disputes and Slee' s understanding was that the management estimates upon which he relied represented the company's legal advice. In relation to the latter, there was an additional difficulty from Slee's point of view in that he understood the business to be more in the nature of bank guarantees than insurance such that there were no actuarial statistics available to assist him in arriving at a valuation. 150

Here, Wilkinson and Pearson again criticised Slee's approach. Pearson said that in the case of Charman, Slee ought to have obtained some understanding of the range

324 Under-provisioning/ or claims

of possible outcomes (and disclosed the same in his reports) before determining whether it was appropriate to incorporate the case estimates as part of his central estimate. 151

Wilkinson criticised Slee's reliance upon management estimates of the outcome of legal disputes generally. He said that he would not have accepted management's estimates of the likely outcome of these disputes in isolation. Rather, he would have wanted some confirmation that the company had reasonably assessed the likely outcome. He said that he would usually speak to the company and the lawyers about the matter before forming a view. 152 In broad terms he believed that HIH management tended to take an optimistic view of the way legal disputes would

settle, citing film finance as an example. 153

Summary

In my view, it follows from the above that Slee's approach to the valuations was not without difficulty. It is likely that Slee's reports would have been improved, and of more use to their readers, had they adopted the techniques advocated by Pearson and Wilkinson. It may well be that the adoption of those techniques would have resulted

in changed and more accurate valuations. But it may also be that these are matters of professional judgment on which views will differ. The impression I have from the evidence is that to some extent at least the problems which Slee experienced in preparing his reports were attributable to difficulties which he experienced in relation to the accuracy and flow to him of provisioning information and data. I now tum to that matter.

15.4.6 Accuracy and flow of provisioning information

Assessment of a general insurance company's OCL is, even in ideal circumstances, an inexact science. And it is a truism that the accuracy of the assessment of the OCL depends upon the adequacy and accuracy of the data upon which it is based. In my view, inadequate controls and procedures existed within HIH to ensure that accurate and complete information and data was available and communicated to the appropriate people (and in particular, to Slee). These inadequacies contributed

significantly to the provisioning problems with which the officers of HIH were beset.

During the course of the evidence, I have encountered a number of examples of these inadequacies. They can be described generally as falling into three categories. First are the inadequacies in the information available to management. An example is the F AI reinsurance problems which I will come to in the next section of this chapter. Second, inadequacies in the flow of information and data to Slee; examples

include reliance upon out of date data in relation to the Asia operations, the provision of incomplete data in relation to the builders warranty business, the unknown liability binder exposure and the PWC reports in relation to the F AI liability and PI portfolios as at 30 June 1999. Third, inadequacies in the flow of

information to the board; examples include Slee's reports and warnings, the

The failure of Hi H Insurance 325

unhooked exposures in the liability portfolio, the PI reserve adjustments and the views expressed by Milliman & Robertson in relation to the US operations in March 2000.

I have dealt with matters in the third category elsewhere in this chapter. In relation to the former two categories, I heard a great deal of evidence as to the circumstances surrounding information and data inadequacies. That evidence was described in detail in the submissions of counsel assisting. 154 I will endeavour to summarise the most pertinent examples in a moment. I have come to the view that it would not be a useful exercise for me to attempt to attribute responsibility for each of these inadequacies. In any event, in respect of many of the instances cited the evidence was not sufficiently clear to permit me to proceed confidently to a conclusion. In my view, the primary significance of these inadequacies lies in their very existence, the fact that they recurred and were not remedied, and the fact that they tended to

undermine the accuracy of the assessments of both HIH management and Slee of HIH's OCL. In that way the inadequacies contributed significantly to the under­ provisioning within HIH. Ultimately under-provisioning was the major cause of HIH's collapse.

I turn now to summarise some of the inadequacies in the accuracy and flow of provisioning information and data.

F AI reinsurance problems

Following the acquisition of FAI, the risk management and reinsurance division of HIH conducted a review ofFAl's reinsurance procedures and practices. The review commenced in January 1999 when Peter Thompson instructed Efty Dimos to visit the F AI reinsurance team in order to obtain an understanding of their practices and procedures. During the review, which continued throughout 1999 and 2000, Dimos identified a number of issues and problems in relation to the FAI reinsurance programme. She documented the progress of the review of these issues and problems in a series of 'F AI Reinsurance Issues' quarterly reports (the F AI RI reports). These reports were addressed to Thompson but were forwarded to various people including Fodera. At least in the case of the June 1999 and June 2000 FAI RI reports, they were also forwarded to Andersen.

The main issues the subject of those quarterly reports (other than those relating to the financial reinsurance arrangements with GCR and National Indemnity dealt with elsewhere in this report) were:

• incorrect reinsurance recoveries generated by FAI's insurance system (AEGIS)

• the systems fix

• disputed reinsurance recoveries

• FAI's commutation of the 'FAI Re' arrangement with GCR.

Each of the problems was outlined in a statement provided by Dimos 155 , and the summary of each which follows draws heavily upon that statement.

326 Under-provisioning for claims

Incorrect reinsurance recoveries generated by AEGIS This problem was discovered around the time ofthe June 1999 FAI RI report (dated 10 July 1999). 156

The incorrect recoveries arose because AEGIS generated reinsurance recoveries 15 7 :

• based on incorrect retentions on a number of claims

• on a reinsurance contract which had been exhausted in December 1996

• against claims from various professional indemnity schemes which had been aggregated and processed as one claim (and hence having one retention) rather than as multiple claims with multiple retentions and

• against classes of business not protected by reinsurance.

Dimos said that although this problem was undoubtedly in existence at F AI pre­ acquisition, she had not been able to ascertain the extent to which it had been identified and dealt with by people within F AI.

In the June 1999 F AI RI report, the extent of the problem was estimated to be $15 .7 million (consisting of incorrect recoveries of $19.7 million less a provision taken up by F AI prior to its acquisition by HIH of $4.0 million). 158 But in an endorsement to the July 1999 F AI RI report (dated 11 August 1999) 159 the extent of the incorrect recoveries was estimated at $35 .1 million (consisting of $19 million in

respect of paid claims and $16.1 million in respect of outstanding claims). Dimos' evidence was that from that time, the RMR division continued to monitor the problem from quarter to quarter, summarising the extent of the incorrect recoveries and recommending appropriate write-offs in each of the subsequent quarterly F AI RI reports. 160

Both the size of the problem and HIH's understanding of it grew over time. By the time of the March 2000 FAI RI report 16 1, the incorrect recoveries were estimated at $61.8 million (consisting of $29.8 million in respect of paid claims and $32 million in respect of outstanding claims). The report explained that $29.8 million in respect of paid claims had been ' backed out' but remained in the general ledger. The

$32 million figure in respect of outstanding claims was still to be addressed. Dimos explained that the total extent of the problem was $61 .8 million. She and Thompson had difficulty in having the balances written off as Fodera and the financial services division (FSD) wanted to be certain they were bona fide write-offs. Dimos said that Fodera and FSD initially believed the balances might be recoverable under the GCR and National Indemnity reinsurance arrangements, but they were eventually written off in approximately June 2000. Dimos explained that thereafter she and the RMR

division monitored the problem on a monthly basis and wrote off the subsequent incorrect recoveries with FSD.

The systems fix This issue stemmed from a ' fix' of the AEGIS system attempted by FAI in May 1998. It resulted in an increase in reinsurance paid of $4 .9 million which could not be substantiated. An attempt by the HIH IT department to correct 'the fix'

Th e failure ofH!H Insurance 327

in June 1999 resulted in a further unsubstantiated $10.9 million in reinsurance paid, and the need for a total write-off of $15.8 million. 162

The problem was first documented by the RMR division in the September 1999 F AI RI report 163 , with that report noting the need for a $15 million write-off. It is not clear to me when this amount was written off but evidence of Dimos suggest that it may not have been written off until during 2000. 164

Disputed reinsurance recoveries The disputed F AI reinsurance recoveries had been in question since at least 1997. Dimos became aware of the problem in mid-1999 when she requested from the relevant brokers a ' status position' on a large unpaid balance due from various London market reinsurers. She was advised that the lead insurer (then Imperial Re, now ALEA) had conducted an inspection ofFAl's PI and CTP reserves in 1994 and

1995 ; that they had discovered various unsatisfactory reserving practices within F AI; and that this resulted in the PI recoveries being disputed by the reinsurers.165

Dimos liaised with the various reinsurers throughout 1999 and 2000 in an effort to resolve the issue of the disputed recoveries. 166 In the June 1999 FAI RI report 16 7, she referred to $12.1 million in disputed recoveries, commenting that 'it is too early to comment on the collectability of these claims recoveries at this stage'. 168

Discussions then took place over the last half of 1999 and into 2000. As at December 1999 it was estimated that the likely loss on the settlement of these disputes would be $2 million, although the evidence does not permit a finding as to the precise amount of the actual loss upon resolution of the last of the disputed recoveries during 2000.

F AI 's commutation of the F AI Re arrangement From 1995 to 1997 FAI reinsured with the GCR bottom layers (up to $5 million) of its whole-of-account reinsurance programme for all classes of business. F AI commuted this arrangement in March 1998. GCR repaid FAI the balance of the pool of premiums that had been built up ($28 million), thereby discharging their liability for claims arising under the F AI Re arrangement. F AI took up $1 7 million of the $28 million as a profit on the commutation, and then set aside the remaining $11 million as a provision for future claims recoveries against the F AI Re arrangement. 169 Dimos explained that the $11 million was 'in no way enough' to cover the claims and hence HIH was left with a loss arising from the

commutation. 170 The RMR division became aware of this problem

in July or August 1999. 171 The issue was referred to briefly in the June 1999 F AI RI report 172 , although not quantified.

The problem remained unquantified in the September 1999, December 1999 and March 2000 FAI RI reports. By the time of the June 2000 FAI RI report 173 , there were recognised claims of $12.6 million against the $11 million which had been set aside. The report stated that the minimum loss of $1.6 million should be taken up as at 30 June 2000. The author added that ' we remain of the view that a large IBNR provision is needed which will obviously worsen the loss ' .

328 Under-provisioning f or claims

Dimos added that in December 2000 she visited the Hong Kong office. She found that they had continued to book reinsurance recoveries to the F AI Re arrangement to the value of $17 million. She said that it may have been that the Hong Kong office was unaware of the commutation of that arrangement.

Treatment of the F AI reinsurance problems in HIH 's accounts It is apparent from this summary of four of the problems ansmg from FAI's reinsurance programme that HIH inherited significant problems as a result of FAI's troubled AEGIS system and inadequate reconciliation and recording procedures. Nevertheless, it also appears that HIH was slow to recognise the financial impact of

the various problems that had been discovered and investigated by the RMR division.

By the time the June 1999 accounts were finalised a reader of the F AI RI reports would have been aware that:

• the AEGIS system had generated incorrect reinsurance recoveries, estimated as at 11 August 1999 to total $3 5.1 million

• there was a dispute dating back some three years in relation to approximately $12 million in reinsurance recoveries and

• there was a potential loss on the commutation ofFAl Re with GCR.

However, it appears that the only provision or allowance for these issues in HIH's accounts was an amount of $10 million included within the F AI exposures of $125 million referred to in the HIH presentation to the 25 August 1999 audit committee meeting as 'protected' by the GCR and Hannover reinsurance contracts. 174 I find later in this chapter that this amount of $125 million was not in fact booked. But even if it had been, in my view, an allowance of $10 million on account of the F AI reinsurance problems as at 30 June 1999 was insufficient. Even on the state of knowledge within HIH at the time it was unlikely that $10 million

was sufficient. With the benefit of hindsight, and accepting that it is impossible to be precise, I think a number of allowances ought to have been made as at 30June 1999.

First, between say $30 million and $40 million for the incorrect reinsurance recoveries (being the total incorrect recoveries of $61.8 million referred to by Dimos, Jess the $11.2 million provision referred to in the F AI RI reports and the amount of any other provision already included in respect of the incorrect recoveries which existed-the evidence does not allow me to be precise about this figure).

Second, approximately $2 million in respect of the disputed recoveries, being the approximate amount of the losses made on the settlement of these disputes.

Third, in relation to F AI Re, it appears that the extent of this problem was at least $17 million (being the recoveries booked against F AI Re by the Hong Kong office-although I am not sure when it was that these recoveries were generated and booked) plus $1.6 million (being the loss which had been recognised by June 2000)

The failure of HIH Insurance 329

'"

and some allowance for the IBNR which was still to be quantified as at 30 June 2000.

And finally, in relation to the systems fix problem, although there is no evidence that the RMR division or FSD had knowledge of the extent of the problem at the time of finalisation of the 30 June 1999 accounts, with the benefit of hindsight, the $15 million amount identified in the F AI R1 reports ought to have been written off.

Taken together, it seems reasonable to suggest that an amount of between $67 million and $77 million should have been taken up as at 30 June 1999 (assuming the benefit of hindsight) to reflect the F AI reinsurance problems ultimately uncovered by HIH, rather than the $10 million allowance which was made (but not in fact booked) as at 30 June 1999. 175

In relation to the accounts of HIH for the year ending 30 June 2000, it seems to me that most of the issues had been resolved or reflected in the accounts by that time. The exception to this was the $17 million in claims against the F AI Re arrangement later discovered in the Hong Kong office, together with the identified losses ($1.6 million) and an unquantified amount of IBNR for future development of the claims against F AI Re. With the benefit of hindsight it is apparent that this amount (of at least $18 .6 million) ought to have been reflected in HIH's consolidated accounts as at 30 June 2000.

Unknown liability exposures

In his evidence, Cameron 176 referred to a difficulty in obtaining data in relation to an ongoing exposure to certain binder claims in the C&G liability portfolio.

Cameron said that he had understood that the exposures to these claims had ceased by the end of 1998 . He referred to a passage in his draft C&G liability valuation report dated 19 February 2001 in which he indicated his understanding that any exposure to binders, schemes or bulk payment files had ceased during accident year

1997- 98. 177 Later in that report he queried whether there were any policies open after that period, noting that if they were not disclosed then the company ran a serious risk of under-stating the OCL by a significant amount. Cameron added that it had been clear from his valuations as at December 1999 and June 2000 that he had assumed the bulk payment claims stopped in 1997- 98 .

178

Cameron recalled the deadline for completing the valuations (so that HIH could release its results to the Australian Stock Exchange) fast approaching. The draft had been sent to Eade for him to dispatch to the various senior staff for their review. It was only at that late stage that the omission of exposures post dating 1998 was noticed by senior management. Cameron explained that when it was brought to their attention, he and Slee realised that all their valuations since 1998 had been short by the amount of the bulk payment claims that had not been recorded (an amount of $2 .9 million). Management had apparently not previously noticed this omission. An amount of $2.9 million was added to the valuation at that stage. 179

330 Un der-provis ioning for claims

Cameron said that at the time they learned of this exposure he and Slee had only recently learned from one of the underwriters that HIH C&G had some years earlier underwritten $3 million in aggregate stop-loss reinsurance on local government covers and that the entire $3 million was going to be incurred. There had also been an earlier failure by management to advise them of a quota share treaty in the valuation as at 30 June 2000, which again had been overlooked by management

when reviewing the draft report. 180

Cameron's evidence was that discovery of the binder exposure was 'the last straw ' in a series of revelations of unusual exposures and reinsurance lurking in what had for many years been represented to him and Slee as a homogenous book of risks. Cameron said it was this latest revelation that resulted in Slee meeting senior management and suggesting that they had been 'economical with the truth'. 181

The circumstances surrounding the apparent late notification of these exposures to Slee and Cameron were not the subject of sufficient evidence to enable me to attribute responsibility to a particular individual or individuals. Nevertheless, they provide a further useful illustration of some of the inadequacies in the data

maintained by HIH, and in the process of communication of that data to Slee and Cameron. In this instance, the inadequacies contributed to the difficulties which Slee and Cameron faced in attempting to assess accurately the OCL of HIH's liability portfolio.

Builders warranty Both FAI and HIH wrote builders warranty business. In relation to HIH's builders warranty business, the data for years of account 1995 to 1998, was kept and administered as part of the PI portfolio on the LAIRS computer system. But the data for builders warranty renewals from 1999 onwards was kept in a separate builders warranty portfolio on the LAIRS system. In his 31 December 2000 report, Slee

made reference to the separate renewal policies, noting that the business had been valued by 'the BW division'. 182

It appears that Slee was not supplied with complete data for either the FAI builders warranty business or the HIH renewal business.

In relation to the F AI builders warranty business, Eade sent an email to Trahair dated 10 December 1999.183 The email was headed 'FAI November 1999 Liability and PI data' and included the following :

Professional Indemnity: BLD; BWB; BWI; FIN; FST; INP; LB ; LEG; PLU; PMD; PTR; SS.

I am not sure why these sub classes are being excluded from the actuarial data. I note however that some $11M in premium was apparently written in the 1998/99 period for these risks, and I am assuming that there is also some exposure to outstanding claims on these as well.

By further email dated 10 December 1999, Howard forwarded this email to Fodera and Cassidy. 184 He wrote:

The failure of HIH Insurance 331

'!11'

Gentlemen,

This memo has caused Mr Martin to be unduly upset.

It would appear that Rob has fully discussed this issue with Ross, but Ross has taken it upon himself to give a slant that is not appropriate at this point in time-given that this is in progress to a long term solution.

This scenario/style of communication to all parties is not appropriate, and on my return will be addressed in person.

The ora] . evidence explaining the circumstances which caused Martin to become 'unduly upset' was inconsistent. I do not propose to resolve that conflict. Rather, it is sufficient to note that it is apparent from this email that Slee was not provided with certain sub-classes ofF AI data (including some builders warranty data) in late

1999.

Further concerns arose in relation to the completeness of the builders warranty data provided to Slee around the time of the finalisation of his 31 December 2000 report. The concerns this time involved the discovery of the separate HIH data in relation to the 1999 renewal business. Cameron 185 and Slee 186 each gave evidence about this matter, as did Cassidy. 187 Wilkinson adverted to the issue in the phase II report. 188 There was a suggestion in the evidence that the 1999 renewal business had been included within the short tail portfolio and would therefore have been provided for in management's estimates of the OCL for that portfolio. But Wilkinson's conclusion was that no provision had been made for it. Wilkinson valued this builders warranty business at $50 million.189

In his December 2002 statement 190, Wilkinson noted that his estimate of the gross reserve required for the builders warranty business had grown from $70 million to $300 million. But he noted that as a result of reinsurance cover which had been identified in relation to the F AI component of this business, the overall reserve increase in net terms was in the order of$125 million.

This is a further illustration of the inadequate controls and procedures in place within HIH to ensure that accurate and complete data was available, and communicated to Slee.

PWC valuations as at 30 June 1999 Fodera commissioned PWC to value the F AI PI and F AI corporate liability portfolios as at 30 June 1999. By letters dated 12 August 1999, Chappell provided those valuations. He valued F AI corporate liability at between $63 million and

$68 million. 191 This was approximately $12 million to $17 million higher than Slee's estimate for that portfolio.

In relation to F AI PI, the position was a little more complicated. Chappell valued FAI PI at $141.064 million. 192 Slee's figure for this business was $121.7 million. But the PWC figure included an amount of $17.4 million for MDU. If this amount were excluded, then it would leave a PWC valuation of $123 .7 million. Slee

332 Under-provisioning for claims

included an amount of $7.67 million for MDU (net of stop-loss). If this amount were excluded from Slee' s valuation, his figure would be $114 million.

On its face, this suggests Slee' s valuation was some $9.7 million less than that of PWC. But Slee's figure assumed reinsurance recoveries of $27.977 million. It appears this amount was overstated by $10 million. Support for this is found in an email from Paul Bateman to Howard (with copies to Fodera, Brown, Eade and Peter Duesbury) dated 16 August 1999. In that email, Bateman wrote:

[ have reviewed the draft reinsurance recovery schedule that was supplied by RMR today in relation to the June reserves and unfd rtunately, have to report that the amount of $27,977,000 taken as recoveries on case estimates greater than $1M by the actuary appears to be overstated.

This would mean that the total valuation by the actuary could potentially increase by $12 ,154,000 pre discount and $10M post discount.

The above interpretation may be subjective given the uncertainty regarding the F AI reinsurance arrangements and how the reinsurance may apply to the estimated IBNR claims in excess of $1M.

Please advise if you wish me to take the matter further or to notify the actuary. 193

On this basis, Slee's figure appears to have been under-estimated by $10 million. If $10 million were added to the Slee figure, then both the PWC and Slee valuations would be approximately $124 million. This amount was approximately $7 million more than the booked provision of $117 million. 194

The existence of the two PWC reports as at 30 June 1999, and the error in the reinsurance recoveries, were matters clearly material to the consideration of the provisions for the F AI PI and F AI Corporate PL portfolios as at 30 June 1999. The evidence suggests that neither Slee 195 nor Andersen 196 was made aware of these

matters.

Peter Duesbury gave evidence in relation to this matter. He said that he was aware of the PWC reviews. Simpson informed him of them and told him that Fodera wanted to review what Slee had done. When asked whether Slee or Cameron were informed that PWC had been engaged to carry out these reviews, Duesbury said he was asked not to do so by Simpson. Simpson told him that the reviews were being done on a private and confidential basis and that he would have to keep it that way. 197 Duesbury did not pass on the results of the PWC review to either Slee or Cameron. He regarded it as a matter for Simpson, Fodera and Howard, even though

he considered that the results of PWC's work would have been important information for Slee and Cameron to have had. 198 Duesbury also confirmed that he did not provide Andersen with a copy of the PWC reviews. 199

The failure of HJH insurance 333

The evidence establishes that each of Fodera, Simpson, Howard and Duesbury knew of the existence of the PWC reports, and that Fodera, Howard and Duesbury also knew of the error in the reinsurance recoveries which Slee had assumed. The evidence also establishes that none of these matters was communicated to Slee, and that Andersen was not informed of the PWC reports. This was obviously an unsatisfactory breakdown in the communication of material provisioning information.

Out of date data for the Asian operations In his 31 December 2000 report, Slee included an amount of $74 million for the Hong Kong office, and a further $8 million for the other Asian operations. He wrote200 :

Another area of concern is the Asia operation where no data has been provided since the June 1999 valuation, and reliance on booked value is the only course open to me . That this is unsatisfactory states the obvious.

Later in that report Slee referred to his use of underwriters' estimates for the Hong Kong and Asian operations?01

In her statement, Pearson202 referred to Slee's reliance in his 31 December 2000 report upon data not updated since 30 June 1999. She wrote:

In my opinion it was not appropriate for Mr Slee to adopt the booked value for the Asian operations in these circumstances, despite the statement made in the reliances. The lack of data from the Asian operations over the eighteen month period should have caused Mr Slee to be extremely concerned about the quality of the operation and the ability of anyone (including the company) to make an assessment of an appropriate outstanding claims provision.

Again, I do not propose to attribute responsibility for this obvious inadequacy in the data available to Slee. But again it shows the difficulties he and others faced in accurately assessing HIH' s OCL.

15.5 Liability

15.5.1 Introduction

My consideration of the liability portfolio has revealed a constant and significant deterioration in claims experience in the years leading up to the collapse of HIH. It has also revealed actuarial analyses infected by optimism, and a failure by senior management (and the auditors) to bring to the attention of the audit committee and main board information which was material to their consideration of the adequacy of the provision for that portfolio.

As I mentioned in the overview of HIH's insurance business, in the period 1998 to 2001 HIH wrote and managed liability business through each of C&G, FAI, CIC

334 Under-provisioning for claims

and CMG. During that period, HIH booked a provision for the liability portfolio which ranged between approximately $350 million and $450 million. The C&G component of that provision ranged between approximately $220 million and $240 million, or over 50 per cent of the total liability portfolio. The F AI liability

business accounted for approximately $100 million of the liability provision. The F AI component was divided between F AI corporate liability and F AI General Liability. The CIC component of the liability portfolio represented approximately $80 million to $90 million of the portfolio provision. The CMG component was

much smaller. The liability portfolio also included a line of business referred to as US Liability. This involved business written from Australia but covering certain liability risks in the United States. It was managed by the RMR division and is dealt with in the later section of this chapter covering the run-offtreaty account.

The phase II report suggested a provision for liability as at 15 March 2001 of $1001 million (undiscounted) (plus prudential margin of $250 million), or $864 million (discounted).203 When compared with HIH' s provision for liability as at 31 December 2000 of $3 7 5 million ( discounted)204 , the phase II report suggested (ignoring any discrepancy arising from the differing dates of valuation) the liability

provision was inadequate by some $489 million on a discounted basis. On any view, this is a significant discrepancy.

A comparison of the provisions booked by HIH and the actuarial estimates during the period 31 December 1997 to 31 December 2000 shows that a material discrepancy existed throughout. The discrepancy was in the order of $30 million to $40 million with the actuarial estimate being consistently higher than the booked provision. By 31 December 2000 the discrepancy was significantly larger and in the order of $130 million. The bulk of the discrepancy throughout the period was attributable to the C&G component of the portfolio. The existence and order of the

total discrepancy (at least until 30 June 2000) was apparent from the 'Adequacy of Provisions' slides included in Andersen's six monthly presentations to the audit committee.

In due course, I will embark upon a detailed consideration some of the actuarial work in relation to the C&G and F AI corporate liability components of the liability portfolio. For the moment it is sufficient to say that I have formed the view that the work was infected with optimism and significantly understated the OCL of HIH' s

liability portfolio. I will commence my analysis by looking at an issue concerning certain unhooked exposures in the liability portfolio.

15.5.2 Unbooked exposures

The management of the liability portfolio was undertaken through HIH (Liability) Pty Limited. During the period 1998 to 2001 there were several directors of HIH (Liability). They included not only management personnel such as Brown, Rod Aistrope, Eade and Bacsak, but also Cassidy (from 18 August 1986 to

15 March 2001) and Fodera (from 31 January 1996 to 15 March 2001).

Th e failure of HJH insurance 335

' r

The board of HIH (Liability) met about once each quarter to consider the results of the previous quarter. The board papers for the meetings included reports from the divisional management in relation to the division' s performance for the period. As members of the board of HIH (Liability), both Cassidy and Fodera received and considered the board papers. They also attended and participated in most of the meetings of the board.

Up to 1996 the method used by HIH management to arrive at its estimate of the liability OCL was to add a 30 per cent loading on case estimates to obtain an IBNR factor. The total would then be tested against external actuarial advice. But during 1996 the company adopted what was described as an Ultimate Loss Ratio (ULR) exposure model. This model relied upon an estimate of the exposure to outstanding claims according to a ULR for a given year of account. 205 The ULR for each year of account was selected by reference to recent years' experience. This ULR would then be used to determine the amount to be taken up in the general ledger as the company's provision for the liability portfolio. As a particular year of account began to develop, an actual loss ratio could then be calculated. The difference between the booked amount (calculated by reference to the ULR) and the amount calculated by reference to the actual loss ratio at any given time was described as the provision for IBNR. It follows from this approach that once the actual loss ratio reached the ULR, then unless the ULR was adjusted upwards, the IBNR for that year of account would be exhausted. I note that as the bulk of the liability portfolio was claims incurred business, it would be usual to see significant IBNR for a number of years following the conclusion of each year of account. The position might be contrasted with the PI portfolio where the bulk of the business was written on a claims made

basis. In those circumstances, there would usually be a smaller provision by way of IBNR.

During 1998 and 1999 senior management within the liability division began to calibrate more accurately the likely ULR for a particular year of account. A new ULR exposure model was developed. This resulted in a significant increase in the ULRs and consequently in the booked provision for the year ending

31 December 1998.

From May 1999 the divisional modelling identified and continued to track an exposure which was ultimately described as an 'unhooked exposure ' and estimated to be $65 million by year end 31 December 1999. In response, the company maintained the December 1998 ULRs for each year of account in the company's general ledger unless and until the actual incurred loss ratio for that year of account exceeded the ULR. When that occurred the ULR would then be adjusted upward to coincide with the actual loss ratio. The effect was that for each year of account, as the actual loss ratio moved closer to the amount calculated by reference to the 31 December 1998 ULRs, the IBNR provision was correspondingly reduced and in many cases exhausted. Once the IBNR was exhausted, new reported claims and any development in existing claims were reflected directly in the profit-and-loss statement.

336 Under-provisioning for claims

In the submissions of counsel assisting the relevant extracts from the HIH (Liability) board papers were set out at some length. 206 I do not propose reproducing that material. The following is a summary of what I regard as some of the more pertinent passages from those board papers.

In the minutes of the HIH (Liability) board meeting of 18 May 1999 reference was made to a detailed review ofULRs compared with those as at 31 December 1998. It was noted that in relation to the 1996 to 1998 years there was an adverse claims trend with an 'overall $23 million worst case exposure' .207 In the underwriting performance overview for the quarter ended 31 March 1999, reference was made to a 'worst case exposure' as at 31 March 1999 of $19.276 million, followed by this comment:

Whilst the actuarial assessment is expected to support this potential worst case scenario, ne ither model fully takes into account the impact of the aggressive claims settlement strategy currently being followed by our claims personnel and is therefore considered to be somewhat unrealistic on that basis. Claim number and average cost of claims trends will continue to be monitored over the next few months to confirm that our current strategy is achieving expectations?08

The minutes of the HIH (Liability) board meeting on 10 August 1999 noted the adoption of a policy of 'maintaining the December 1998 assumptions and ULRs wherever possible at this stage' .Z 09 The underwriting performance overview for the preceding quarter recorded that the division was maintaining the December 1998 assumptions when calculating ULRs 'only amending them if actual incurred claims exceeded the ULR' . 210 The minutes also noted that the watching brief established as at March 1999 was being maintained211 and that the overall worst case exposure was then considered to be in the vicinity of $27 million. 212

In my view, those present at the divisional board meeting on 10 August 1999 (which was prior to the finalisation of HIH's consolidated accounts for the period ended 30 June 1999), or those reading the board papers for that meeting, would have appreciated there were already some serious concerns about the adequacy of the provision for the liability portfolio.

By the time of the next HIH (Liability) board meeting on 17 November 1999 the ULR exposure model was suggesting an even greater exposure. There were now no IBNRs for the 1996 and earlier accident years. The minutes repeated the policy of maintaining the December 1998 ULRs wherever possible, but booking actual

figures where the actual loss ratios exceeded the ULRs. The minutes then noted a $45 million exposure, which was explained in the following terms:

Of the $45M exposure some will need to be booked because of the lack of IBNR. We anticipate a loss as at 30 September of ($3.7 M). For 30 June we may be forced to book a $20M loss . Dominic Fodera confirmed that we would expect in the next nine months a I 99 7 YOA IBNR in the range of $15-20M. Greg Brown confirmed that the settlement

strategy could bring this down to approximately $15M at best. Dominic Fodera stated that it is essential that an IBNR model be developed . Peter

The failure ofHIH Insurance 337

Bacsak suggested that it could be possible that we over estimate our IBNR claims as we do expect them to cost us more due to inflationary factors . 213

The minutes of the HIH (Liability) board meeting on 15 February 2000 noted that the 'unhooked exposure' had increased to $65 million as at December 1999, being an increase from the $45 million as at June 1999.214 The board papers noted that an additional $12 million would need to be booked at 30 June 2000 and $14 million by 31 December 2000, leading to losses over the ensuing years.

215

The minutes of the HIH (Liability) board meeting on 16 May 2000 referred to the $65 million unhooked exposure identified as at 31 December 1999, noting an 'agreed strategy' for booking that exposure ' as and when those losses

crystallised' .216 The underwriting performance overview for the period ended 31 March 2000 noted that it had been agreed that the $65 million exposure to the 1998 and prior accident years would continue to be booked as claims were reported and the existing IBNR reserves were exhausted. It also noted that the exposure was

'expected to crystallise as follows': $12 million of the half-year ending 30 June 2000, $24 million for the year ending 30 June 2001, $24 million for the year ending 30 June 2002 and $5 million for the year ending 30 June 2003. 217

Although board papers were still prepared, there were no further meetings of the HIH (Liability) board after the May 2000 meeting. Cassidy explained that this was because of the Allianz transaction. 218

The underwriting performance overview for the period ended 30 June 2000 noted that since 31 December 1999 there had been a reserve strengthening of $15 million, but also a lift in the projected unhooked exposure by $8 million, taking the unhooked exposure at 30 June 2000 to $58 million. 219 The underwriting performance overview for the period ended 30 September 2000 did not put a number on the unbooked exposure. Instead, it simply noted that there had been a further deterioration in the OCL, which had been partially offset by a $3 million release in IBNR in the council portfolio.220 Finally, the underwriting performance overview for the period ended 31 December 2000 refetTed to the $58 million unhooked exposure identified as at 30 June 2000 before commenting that some further exposure had been identified, taking the total to $62.9 million. This was said to be disappointing given that since 30 June 2000 the reserves had been strengthened by some $12.2 million (or $7.6 million net ofiBNR releases).221

It was put to me by counsel assisting that knowledge of the existence and approximate size of the unhooked exposure referred to in the board papers of HIH (Liability) was at least material to any sensible consideration of the adequacy of HIH' s provisions. It was also put that Cassidy, Fodera and Williams were aware of the existence and approximate order of the unbooked exposure, yet they did not inform, or adequately inform, their fellow directors of the exposures. Nor did they take adequate steps to ensure that the unhooked exposures were reflected in HIH's provisions as at 31 December 1999 and 30 June 2000. These are serious allegations, which I must consider carefully.

338 Under-provisioning/or claims

Cassidy, Fodera and Williams's knowledge of the unhooked exposures I am satisfied that Cassidy and Fodera were aware of the unhooked exposures from their attendance at the relevant HIH (Liability) board meetings and their perusal of the board papers for those meetings. They acknowledged as much in their oral evidence. 222

Williams did not attend the HIH (Liability) board meetings. But, in his evidence he said that he was sent the papers of the divisional board meetings on a regular basis. It was his practice to read them. He said that while they were quiet voluminous, he would go through them and review the executive summaries and some of the salient

figures. 223 He acknowledged that in relation to the papers for the board meeting dated 15 February 2000, he had noted the paragraph commencing 'the lack of IBNRs'. 224

I note that this reference to the lack of IBNRs was followed by a reference to an unhooked exposure of $65 million as at 31 December 1999. Williams said that this was of concern to him and that he understood that it meant that they lacked additional provisions to cover IBNRs. He understood this to mean that on a divisional level claims were only being booked as and when they occurred, although he believed that at a group level the provision was taken up. 225 He accepted that this paragraph from the divisional board meeting papers dated

15 February 2000 suggested an unhooked exposure of $65 million as at 31 December 1999. 226

Cassidy and Fodera knew that the accounting standards required that HIH book a provision which was at least equal to its best estimate of its exposure to all claims, including those which had been incurred but not yet reported.227 Although it does not appear to have put directly to Williams, he ought to have known this.

In my view it follows that each of Cassidy, Fodera and Williams knew that on the basis of the ULR exposure model being used by management, HIH was significantly under-provisioned in the liability portfolio by the time offinalisation ofthe accounts for both the period ended 31 December 1999 (by some $65 million) and the period

ended 30 June 2000 (by some $58 million).

Response to the unhooked exposures The issue which then arises for consideration is the adequacy of any action taken by Cassidy, Fodera and Williams either to ensure that the unhooked exposures were adequately reflected in the accounts or to ensure that the audit committee or main

board were made aware of the unhooked exposures which had been identified.

The financial report for the half-year ending 31 December 1999 included the following note in relation to the liability division:

The 1995 and 1996 year of account General Liability portfolio deteriorated to such an extent that all IBNRs for the 1996 and prior year of accounts have been utilised. 228

Fodera was asked whether the adequacy of the liability provision was discussed at the 22 February 2000 meeting of the main board of HIH. He said that the topic was

Th e failure ofHIH Insurance 339

discussed during the audit committee meeting on the same day. However, he was not able to recall any discussion of the above note in the financial report at either the audit committee meeting or board meeting. Moreover, he did not recall any member of the board asking for an explanation of that item, or querying the significance of the IBNRs having been utilised. 229

I think that the reference to the utilisation of IBNRs for the 1996 and prior years of account reflected only part of the picture known to Cassidy, Fodera and Williams. The note said nothing about the potential impact of this matter upon the adequacy of HIH's provision for the liability portfolio. It ought to have made clear that in accordance with the ULR exposure model being used by the liability division management, an unhooked exposure of $65 million had been identified. I consider that in omitting this additional information the financial report was misleading. I also consider that Cassidy, Fodera and Williams either knew or ought to have known that to be so.

The disclosure in the financial report for the period ended 30 June 2000230 was even less informative. There was no mention of the exhaustion of IBNRs or of an unbooked exposure. There was an oblique reference to the problem in the statement that 'the major factor contributing to this variance (between actual and forecast loss ratio deterioration over the previous six months] is the impact that new IBNR claims continue to have on the overall divisional loss ratio'. But this provided little more than a hint of the true nature and extent of the problem or concerns of which Cassidy, Fodera and Williams were aware.

In my view, neither the financial report for the period ended 31 December 1999, nor the financial report for the period ended 30 June 2000, made adequate disclosure of the nature and extent of the unhooked exposure.

The responses (by way of evidence and submission) on behalf of Cassidy, Fodera and Williams to the suggestion they failed to make adequate disclosure to their fellow board members of the unbooked exposures can be summarised in four broad propositions. First, the matter of the unhooked exposures in the liability division was adequately disclosed to the board members during the May 2000 budget briefing sessions. Second, they had reservations concerning the accuracy or reliability of the ULR exposure model. Third, even if the unbooked exposures meant that the provision for the liability portfolio was less than the divisional management view as to the same, nevertheless HIH booked a provision which agreed at a group

level with the estimate of the actuary such that the provision was adequate at a group level. Finally, that maintaining the ULRs at the level assumed as at 31 December 1998 for divisional purposes was a legitimate management tool.

In relation to the first of these responses, each of Cassidy, Fodera and Williams asserted that their fellow board members were informed of the unhooked exposures at the May 2000 budget briefing sessions.Z 31 They referred to the evidence of Brown to the effect that he was asked a question by Justin Gardener about the $65 million at the budget briefing sessions. Brown's evidence232 was that he said they had developed a model which was suggesting a potential shortfall of $65 million and

340 Under-provisioning for claims

that management had decided to amortise it over the coming three years. It was submitted that counsel assisting did not question any of the directors, including Gardener, about their knowledge of the liability unhooked exposure, and that to conclude they were not aware of this exposure in those circumstances would involve a denial of procedural fairness.

The first point to note here is that even if adequate disclosure was made through the May 2000 budget briefing sessions, this was well after the time at which the directors were required to satisfy themselves as to the adequacy of the group provision as at 31 December 1999. In other words, the asserted disclosure came too

late.

Fodera was asked, in his oral evidence, about the nature of the disclosure said to have been made at the May 2000 budget briefing sessions. He said the directors were told that there was an unhooked exposure and that 'we were breaching the gap between what management had booked and what the actuary had projected in a staged format over a two, three year period' .233 Fodera said that Williams spoke to this issue, explaining to the other directors that in the liability division 'there was a variance that we were booking over two, three years' . 23 4

I have some reservations in accepting that the unhooked exposures were adequately disclosed through the May 2000 budget briefing sessions. I say this for a number of reasons. First, Fodera gave evidence that the main board was not told that the liability division had identified an unhooked exposure of $65 million. He said the reason for this was that management were still coming to grips with the ULR exposure model. 235 There is also Pye's evidence that nobody told the audit committee during the meeting on 22 February 2000 that management of the liability division felt that the division was under-provisioned by $65 million. Pye said

discussions would have been restricted to the overall level of provisioning and would not have descended to the line by line provisions.236 Second, even if the unhooked exposures were raised at the May 2000 budget sessions, the information should also have been disclosed in the financial reports for the periods ended 31 December 1999 and 30 June 2000. It should have been raised during the consideration by the main board and audit committee of the adequacy of the

provisions as at 31 December 1999 and 30 June 2000. Third, there is no suggestion in the evidence of any disclosure or discussions of the unhooked exposure other than during the May 2000 budget briefing sessions. Although, it is fair to say that the directors (other than Williams, Cassidy and Fodera) were not directly questioned as to their knowledge of the unhooked exposures, at the same time, a number of the directors denied having been alerted to concerns as to the adequacy of HIH's provision at any stage.

In relation to the second submission put by Cassidy and Fodera, even if it be the case that they had reservations about the accuracy or reliability of the ULR exposure model, the information generated by use of the model should have been made available to the board. I did not understand Cassidy and Fodera to deny the existence of the unhooked exposure. Indeed Fodera and Williams both said they

The failure of HIH Insurance 341

informed the May 2000 budget session of the variance between management's figures and what the actuary had projected. It was for the board to decide whether there were additional exposures and, if so, how they should be dealt with. It would have been material for the board to be informed about results of the ULR exposure model and, if management felt it appropriate, deficiencies seen in it. The board could then decide.

In any event it is difficult to reconcile the reservations asserted by Cassidy and Fodera with the material in the HIH (Liability) board papers or minutes. In the first half of 1999 the exposures were referred to as being in the nature of 'worst case' or 'potential' exposures. But this was not so from the second half of 1999. The board papers and minutes from that time referred to 'unhooked exposures' and suggested an increasing confidence in the figures produced by the new model. The acknowledgment in the board papers and minutes by May 2000 that it would be necessary to book the full amount (albeit over a three year period) is also difficult to reconcile with the asserted reservations. I further note that the deficiencies apprehended by Cassidy and Fodera do not appear to have been shared by Williams or Andersen. The ongoing monitoring of the exposures as summarised in the board papers and minutes tended to confirm (and indeed suggest a worsening of) the previous estimates of the exposures. There is