AYLMER, Mr Christopher Patrick, Head, Domestic Markets Department, Reserve Bank of Australia

ELLIS, Dr Luci Beth, Head, Financial Stability Department, Reserve Bank of Australia

Committee met at 09:13

CHAIR ( Mr Alexander ): I declare open this hearing of the House of Representatives Standing Committee on Economics as part of its inquiry into homeownership. I welcome everyone here this morning. This is the second public hearing for this inquiry, and today we will hear from a number of contributors from the more than 50 submissions that have now been received by the committee. Among the organisations appearing today are the RBA, APRA and the Australian Bankers' Association. We also will be hearing today from Curtis Associates, the Law Society of New South Wales and Urban Taskforce Australia. Mr Saul Eslake will also be joining us by teleconference later this afternoon. At the hearings here in Sydney, over the next two days, we look forward to further discussing the challenges that many prospective buyers face in today's very competitive housing market and whether the current policy settings give all Australians a fair chance to own a home.

I remind you that although the committee does not require you to give evidence under oath the hearings are legal proceedings of the parliament and warrant the same respect as proceedings of the House. The giving of false or misleading evidence is a serious matter and may be regarded as a contempt of parliament. Would you like to make an opening statement before we proceed to questions?

Dr Ellis : Thank you, Chair. I would like to thank the Chair and the committee for the opportunity to discuss the bank's submission and views on this important topic. The Reserve Bank recognises the importance of housing to Australians; it provides us with shelter. Housing costs are a large part of household spending, and a home is the biggest purchase that many of us will make. It is, therefore, no surprise that housing-related issues have been the subject of several inquiries over the past decade or so.

Within that broader realm of housing, the bank recognises that homeownership is an aspiration of many Australians. Outright homeownership is widely regarded as key to avoiding poverty in old age. Before that life stage, homeownership is also regarded as a way to obtain the security of tenure that is so important to the wellbeing of many households, especially families with dependent children. Security of tenure can allow households to enjoy stable arrangements for child care, education and community engagement. It avoids the costs and disruptions involved in frequent moves, which many renters experience.

The interest of the Reserve Bank in housing-related matters goes to the heart of its mandated policy responsibilities. The housing sector is one of the most interest-sensitive parts of the economy, so a significant part of the transmission of monetary policy to the real economy comes via housing markets. Housing market developments are also highly relevant to the Reserve Bank's mandate to promote financial stability.

Housing is the most important asset class for the household sector. It provides security for the finance of many small businesses. Housing-related lending represents a large fraction of the business of the Australian banking system. History shows that housing loans have not generally been as risky as other loans but, such is the size of the sector, the risks involved are nonetheless important. Recent history from around the world shows, as well, that, although household mortgage borrowings typically do not instigate financial crises and distress, they can do so if the institutional arrangements and lending standards are configured to allow it. Australia is a long way from that situation, and we want to make sure that that remains true. More broadly, we want to promote financial stability by making sure that Australians are generally resilient to the financial shocks that might come their way. How much they pay for housing and how they finance that purchase strongly influence that resilience.

As the Reserve Bank submission to this inquiry outlined, trends in the housing market and in patterns of homeownership have shifted over recent decades. Perhaps the most obvious shift was the significant increase in housing prices relative to incomes between the late 1990s and mid-2000s. As the bank has explained on previous occasions, most of this increase was in response to the financial liberalisation and to the decline in inflation in the early 1990s. These were one-off changes, so this transition will not be repeated—and, because the rise in housing prices was largely driven by a decline in mortgage interest rates, it does not necessarily imply that purchasing a home and servicing a mortgage afterwards has become less achievable. Indeed, during this period, Australia's overall homeownership rate was broadly stable. At the same time, the amount and quality of housing that Australians actually consumed, in terms of size of home, number of bedrooms and so forth, have, if anything, increased.

Of course, beyond these broad aggregates, people's individual circumstances differ and, therefore, so do their housing experiences. Within that broadly stable overall rate of homeownership, the rate for younger households has declined somewhat over time. These are the core age groups of first home buyers. At least some of that decline occurred before the marked rise in housing prices relative to income, so it cannot all be attributed to affordability issues. Whether homeownership is affordable depends on one's definition and is open to debate. But there is no disputing that housing is expensive. It is clear that at least part of the reason for this is that demand is strong. Over the long term this can be at least partly explained by the effects of disinflation and financial liberalisation that I referred to earlier. More recently, demand has been boosted by population growth and by declines in interest rates.

As also noted in the Reserve Bank's submission, Australia faces a number of long-standing challenges in meeting that strong demand. The population is highly urbanised and concentrated in a few large cities. Housing prices are typically higher in larger cities. Australia's cities are also unusually low density compared with those in other developed countries. The urban fringe locations, where first home buyers have typically located, are therefore becoming further out—and potentially inconvenient for access to jobs and some services. Some of our major cities also face geographic constraints on their expansion. All of these factors tend to increase the price of well-located housing. In addition, the cost of providing new supply can be quite high because of the costs and delays involved in obtaining all the necessary approvals and in providing the necessary infrastructure to service the land.

Another area where Australia seems quite unusual is that most rental housing is owned by private individuals who are not full-time professional landlords. Investor interest in property has been especially strong in recent years, no doubt partly encouraged by low interest rates and the prospect of (concessionally taxed) capital gains. Investors typically have more equity and borrowing capacity than first-time buyers and, perhaps, also other owner occupiers. They might, therefore, be more able and more willing to pay higher prices than other types of buyers for particular properties. The result has been that the housing sales market has become unusually concentrated in investor activity, particularly in the larger cities. At the margin, this has probably priced some aspiring first home buyers from properties they could otherwise acquire. Nonetheless, while there has been much debate on this issue, from a financial stability point of view it is helpful that there has been no push to improve the relative position of first home buyers by easing lending standards. As recent experiences in other countries have shown, such a step would probably be counterproductive in the longer run. My colleague Chris Aylmer and I would be happy to answer any questions you may have on these issues.

CHAIR: Thank you for that comprehensive statement. It does lead to, it seems, from our hearings to this point, concerns regarding investors who, historically, in an ideal situation, were producing affordable rental properties. Would you agree with that as an ideal?

Dr Ellis : I think as a statement about the affordability of rents it is worth noting that rental growth has been quite low in recent times. It is also worth noting that rental yields are quite low. What we are seeing at the moment, although prices are rising quickly in the larger cities—although not elsewhere—rental growth has not followed suit. While individual circumstance differ, and many low and moderate income households do face difficulty in affording suitable rental accommodation, as a general statement I think it is fair to say that rental housing has not become unaffordable.

CHAIR: The point I was coming to is: in an ideal world, the investor is providing affordable rental properties. The concern that many of us have is: because of this situation, where, in many cases, interest rates are lower than rental returns, the previous brake on investors buying more properties was their capacity to fund their losses through negative gearing. For instance, if interest rates were at 10 per cent and rental returns were at four per cent, for $1 million worth of properties you would have to find $60,000 a year to fund that loss. In a situation where they are line ball or, sometimes, as in various parts of Australia, there is the opportunity to borrow at lower rates than what the rental property will give, is the investor purchasing properties that, otherwise, would have been able to be bought by home buyers?

Dr Ellis : It is a truism that if an investor is buying a property an owner-occupier is not. I guess, to the extent that person is not then buying their own home, they are therefore creating a market for rental and making it attractive to purchase investor properties.

The point I would make about relative yields is that most investors would be looking at that over a longer term, so the instantaneous interest rate differential—the differential between the rental yield and the current interest rate—is not necessarily the only thing that an investor would be or should be looking at. They would be thinking about what they think the average interest rate would be over the holding period. That is also the kind of calculation that lenders will be doing in assessing the serviceability of an investor loan. In particular, lenders are required to make relatively conservative assumptions about the rental yield because, although the rental yield when the property is rented might be one number, for any individual investment you have to allow for the fact that you might have a vacancy, and that makes a very big difference to the actual return. You also need to factor in the costs that are involved in running the property—the maintenance, the property management fees and so forth.

CHAIR: But that aside, you would agree that, as you have said, the investor has the advantage at an auction over a homebuyer, and often it might be the case the investor buys that property—it could even be the case that the losing bidder, the homebuyers, might end up renting that very property from the investor. Would you agree that is a likely scenario?

Dr Ellis : It is not necessarily the case that the prospective buyer ends up renting that particular house, but if you look at the system, to the extent that there are more investors in the market and the same number of households overall, there will certainly be more renters. One of the things that we went through in our submission was that the increase in the housing stock that seems to be owned by private landlords is basically making up the difference of the decline in public rental—so public sector housing—and the overall rate of home ownership has been broadly stable.

CHAIR: We have learnt that there are two areas of investors: (1) liberated from the need to fund negative gearing in this time of neutral gearing or positive gearing, there are those who are buying very large numbers of properties; and (2) people who cannot afford to buy homes—in particular, schoolteachers, police and nurses—are actually involved in the purchase of investment properties because there is this fine line, neutral or possibly even positive gearing. This is a new group of people investing in real estate. Are you aware of that?

Dr Ellis : We have certainly heard reports that that is true, but we do not see that in the overall data. As we reported in our submission, the tax numbers do suggest that people who declare rental income are disproportionately people who are in higher income brackets. I have no doubt there are people from right across the income distribution brackets who are buying investor property. Whether it is the tax benefit that you refer to that is driving that, or whether it is simply that some people do not want to live in the places that they can afford—there are certainly situations where people may not be able to buy in a prestige suburb although they can afford to rent there, but they can afford to buy a property, say an apartment, in an outer area and they would rather operate in that way. They would rather own a property in an area where they choose not to live, and then rent elsewhere. That is certainly a choice that some people make.

CHAIR: A person on a modest income who cannot afford to buy their own home because they would have to fund their mortgage payments, the interest component, with after-tax dollars may find themselves in this time, when interest rates are so low and rents may cover that cost, able to buy a property at this moment, whereas historically they could not?

Dr Ellis : It is possible. We do not have—

CHAIR: Because they do not have the capacity to fund a loss.

Dr Ellis : That is true. I think people on modest incomes would find it more difficult to fund the loss. Certainly negative gearing does make the use of leverage a little bit more comfortable, because then you know that if you do have the property vacant, or for some other reason you end up making a loss in a particular year, you are only wearing the post-tax loss rather than the pre-tax loss.

CHAIR: And for high-income earners that is a greater deduction than for low-income earners.

Dr Ellis : That is true. Of course, remembering also that you might be making a loss on your full marginal tax rate on the cash flow, but somewhere down the track you are gaining capital gains, which are concessionally taxed. So that is a difference.

CHAIR: You would agree, as you said in your opening statement, that when interest rates are lower it creates an upward pressure on housing prices, driven both by the capacity of an investor and by a person buying their own home. Because the interest rate is lower, they are looking at a monthly payment and therefore they can pay more.

Dr Ellis : It is true that when interest rates are low it encourages people to borrow and to buy assets, particularly the assets that you can leverage, of which property is one. I would just add one nuance to that, and that is the way that lenders actually work out how much they will lend to someone. Lower interest rates now might make people more willing to borrow, but under the current arrangements for determining serviceability in lenders' tests there is usually a floor interest rate that they are now required to use when they are assessing how much they will lend to a particular person. For many lenders that floor is binding. So lower interest rates may not necessarily allow a particular borrower to borrow more, but it may encourage them do so.

CHAIR: What we have seen recently in regard to what you have just said is that, as the Reserve Bank sets an interest rate to contain inflation within a certain range, it has collateral effect: when interest rates go up, prices might go down; and when interest rates go down, prices go up.

Dr Ellis : As my colleague Chris Kent pointed out in a speech recently, the housing market is one of the primary channels of the transmission of monetary policy. It is something that we are very aware of, with both housing construction and the desire and ability to purchase existing property. They are both very interest sensitive. That is how monetary policy works.

CHAIR: We know that APRA has recently required banks to lend only 80 per cent to investors.

Dr Ellis : APRA has not specified particular loan-to-valuation ratios for investor lending. Those have dropped out of banks' response to what APRA has actually required. You will be talking to Wayne Byres next, so he will be able to discuss this in more detail. In broad outline, the announcements that APRA made in December, which were part of a more general ratcheting up of their supervisory activities in this area over a period of time, had three main prongs. One was that they did not want banks to be growing their high-risk mortgage lending particularly strongly or to have very high fractions of high-risk lending, and they defined high-risk lending across a number of dimensions of lending standards—loan-to-valuation ratios being one; serviceability loans to nonresidents and so forth; and low doc lending, which is very low nowadays.

The second prong was around serviceability and the loan size, and so they said: we want you to work out the loan size that you will offer to a borrower based on at least a 200 basis point, two percentage point, buffer over the top of the current interest rate and with a floor on that of seven per cent—so, a minimum of seven per cent as the actual rate they use to calculate the repayment to work out how much they will lend. The third prong was a benchmark that lending growth to property investors should not grow faster than 10 per cent over the course of a year.

There were three separate prongs to that policy, to those measures, and none of them were specifically about having a limit on loan-to-valuation ratios. But what has happened is that some lenders have taken particularly the third of those prongs, to slow down the growth in investor lending, and they have tried to get at that in a number of ways. You have seen changes in pricing relative to interest rates offered to owner occupiers and you have also seen tightening in non-price lending standards, including the loan-to-valuation ratio.

CHAIR: So this is a prudential measure to protect the banks from being in a precarious situation, but it has the side effect of also protecting investors from overborrowing?

Dr Ellis : Yes. We would regard the measures that APRA has announced as being consistent with promoting financial stability from the perspective of the financial stability of both the household sector and the financial system.

CHAIR: So the qualification of homebuyers being able to service a seven per cent mortgage is the other measure?

Dr Ellis : Yes.

CHAIR: People being people, a homebuyer might demonstrate that they have got the capacity to pay seven per cent. Both partners might be working, and then their circumstances might change. One partner might choose to have a baby, and the other partner agrees to have that baby and stops working. Or, two or three years after having this four per cent mortgage, they have money left over, they buy a new car and their circumstances change. Would you agree that it is not enforceable that they will retain this capacity to pay seven per cent?

Dr Ellis : It is true that individual circumstances differ. The fact that individual circumstances can change is one of the things that lenders take into account as part of any good practice in credit risk management. It is important for them to have a realistic view of the expenses of that household. That is one of the reasons why there is that buffer of seven per cent—or 200 basis points higher, whichever is the greater. It is important to ensure that there is a buffer in case circumstances change. It is not just about ensuring that there is a buffer in case interest rates rise; it is also about ensuring that there is a buffer in case income should fall. That also goes to what income they count. For example, among the practices that APRA has been paying attention to—and I am sure Wayne will be able to go through this in more detail—is credit risk management. Doing a loan approval is not just, 'Hello. You can fog a mirror; here is your loan.' That is how it was in the US prior to the crisis, but in Australia there are strong credit risk management practices. We have discovered that they could be better in some instances, but they are good enough. This is really about making sure that there is enough of a buffer in case individual circumstances change, and that is what credit risk management is all about.

CHAIR: A creative investor who arranges an 80 per cent loan with a bank may also be able to go somewhere else and borrow the other 20 per cent, and, therefore, manipulate the system?

Dr Ellis : That is true, but when a bank is actually establishing the security they are bringing, the bank wants to know the source of that. If someone is borrowing that and purporting that they have not borrowed it, they are not telling the truth—and that could have repercussions for them.

CHAIR: Although, you could put that 20 per cent down and then, after the fact, borrow that 20 per cent to effect 100 per cent negative gearing on that property.

Dr Ellis : Finance is fungible.

CHAIR: I guess what I am getting at is that it is not absolutely enforceable.

Dr Ellis : I do not want to get too much into the detail of banks' credit risk management practices and how they deal with fraud. I believe you are speaking tomorrow with lenders' mortgage insurers, and I think they will be able to give you some quite good information about, essentially, how they work with lenders to ensure those sorts of practices do not happen.

CHAIR: The charter of the RBA outlines a responsibility to keep the currency stable and maintain economic prosperity and wellbeing for the people of Australia. Many Australians express their wealth, their prosperity, with either investment properties or their own home. When the RBA adjusts interest rates to achieve a range of inflation, it has a collateral effect on the upward or downward pressure on homes. Would the RBA ever consider a course of action that might achieve a similar result that APRA appears to have achieved, if the evidence is correct, in regard to limiting borrowing of 80 per cent? Is there an area that you might explore there?

Dr Ellis : The first point I would make is that the charter in the Reserve Bank Act is further elaborated in the agreement with government in the Statement on the Conduct of Monetary Policy, which refers both to our monetary policy inflation target mandate and also to our responsibility to promote financial stability. We would regard the responsibility to promote financial stability, which has been seen by past governments as a longstanding responsibility, as something that stems from that notion of the prosperity of the Australian people. Certainly, we would regard the financial resilience of the Australian household sector as being something that we should seek to promote within the bounds of our powers. APRA also have a financial stability mandate; in fact their financial stability mandate is legislated. APRA are the ones with the prudential powers. So we would regard what APRA has been doing as fully consistent with both institutions' mandates and with the cooperative way in which the Council of Financial Regulators operates. They are helping us to achieve what we need to achieve as well.

Mr HUSIC: I thank the RBA for its appearance and submission today. I want to start on page 23 of your submission. The RBA does focus on the impact of taxation, which has been the subject of opening questions today, but has also obviously been a matter of quite energetic discussion in the public space. In your submission, the RBAs says:

The Bank believes that there is a case for reviewing negative gearing.

You say that it should not occur in isolation, but needs to occur in tandem with a review of the CGT arrangements that were put in place back in 1999. I am just wondering if you are able to walk the committee through some of the bank's thinking on why the bank believes that this is an appropriate time to look at negative gearing?

Dr Ellis : I would not hang our statement on a particular timing. This is the same position that the Reserve Bank expressed in 2003 in the context of the Productivity Commission's inquiry into first home ownership. That is the first point I would make.

Mr HUSIC: So the bank has had a longstanding view about this issue?

Dr Ellis : That is correct. The bank also recognises that we are not tax experts, so we have not made any specific recommendations about what the review would ideally come up with. We have made an observation that the combination of negative gearing and concessional taxation of capital gains creates an incentive for people to invest in assets that produce capital gains versus assets that do not. Even if negative gearing is not currently required given the current combination of interest rates, the fact that it is available should something goes wrong, should your rental yield not be what you expected and so forth, makes people more comfortable about taking that leverage. So in terms of our financial stability mandate, we think that it is within our mandate to make observations about where in the institutional framework, including the tax system, there might be incentives to engage in more leverage, because it is the leverage piece that is so important for financial stability, both of the financial sector and of the household sector.

Mr HUSIC: In your submission—and I have noticed this elsewhere—there seems to be a divergent reliance on particular groupings of statistics. It appears that the RBA takes a lot of lead from the HILDA stats. The ATO takes a different pathway, and it seems to divide on an issue about gross versus net earnings or after-tax earnings. I am wondering if you can explain why the RBA thinks that after-tax disposable income is a better measure than gross income.

Dr Ellis : For what metric?

Mr HUSIC: In terms of, particularly, looking at who is likely to invest in property.

Dr Ellis : I do not think we have a strong view on which metric is better. I think the way to characterise how we presented what we presented both in the submission and in our September Financial stability review is that we will take whatever data we can get, noting the relative data quality, and we look at all sources of data and try and triangulate against them and work out a story from that. I do not think we have a strong view that one or the other is more meaningful. I think it is helpful to know that the Tax statistics refer to taxable income, and there are a number of people who have very low taxable income but high actual income because they are retired, they have a lot of non-taxable income, they can draw down on super funds and so forth. So you have to look at all sources of data and triangulate against them. We do not have a position on which measure is better.

Mr HUSIC: I am not trying to pitch the RBA against the ATO in this space, but I have noted in your submission, particularly at page 22, that there are a number of observations that you rely on using HILDA data—for instance:

While the incidence of property investment increases with the level of income, the Household, Income and Labour Dynamics in Australia (HILDA) Survey also suggests that most investor households are in the top two income quintiles. These households hold nearly 80 per cent of all investor housing debt …

A lot of store is placed on the reliability of these statistics. I am interested to understand better why the RBA would rely on HILDA more.

Dr Ellis : I would not regard it as relying on HILDA more. With the specific metric you are referring to, the HILDA data allows us to make a statement about how much debt they have. It gives you the balance sheet information, whereas the Tax data only gives you what the deductions are and whether they have deducted interest payments. The Tax data just does not allow us to make a statement about who, in what income bracket, has how much debt. You can only infer whether they have debt at all from the fact that they have deducted interest expenses. So it is simply another data source. We think the HILDA survey has been a great initiative and we have certainly learnt a lot from it. It is a sample survey. There is always going to be a difference between a sampling survey and the population, which is what the Tax data is. We are data mavens and we look at all the sources of data and weigh up the relative reliability of all of them. That is really all we can do.

Mr HUSIC: Sure. But sometimes you observe in the public space the reliance on the ATO data and you see some arguments being made about the way in which negative gearing is being accessed by different income groups. For instance, some have argued, using the ATO data, that people on incomes as low as zero or $20,000 are big beneficiaries of negative gearing. I am trying to work out why this may or may not be an accurate reflection of investor wealth. It would be helpful if you could explain that to the committee.

Dr Ellis : I would underline that the Tax data is about taxable income. As I mentioned earlier, there are a number of people with a very low taxable income who may have high actual incomes, although it is hard for us to say on the basis of the Tax data, because it is about taxable income—because that is what is in people's tax returns. I think it is fair to say that the Reserve Bank has invested quite a lot in learning how to analyse the HILDA data. It is a complex dataset. There are both individuals and households. We have to send our young analysts to do special training and they have to sign special waivers in order to access this data. So there is a lot of investment in understanding the HILDA data, whereas the Tax data is easier for someone who is new to the area or wanting to gain a quick understanding. But, as I said, we are not favouring one dataset over the other for any particular reason. It is really just that we use what data is available and put the weight on it as best we can given the reliability of different datasets. If there were another source of data available, we would certainly look at that as well—we love data.

Mr HUSIC: I certainly understand and appreciate that, Dr Ellis! It is just difficult to understand how people on zero incomes are possibly able to invest. It is probably a reflection of the fact that they have been able to use negative gearing arrangements to be able to get to the point where they are able to do so.

Dr Ellis : Certainly, that would be part of the story. Again, as I mentioned, it is entirely possible that some of these people either have temporarily low incomes—so they are not zero income all of the time—or it is possible that many people have non-taxable sources of income, such as superannuation streams and other income, particularly income that flows through on a non-taxable basis to older Australians.

Mr HUSIC: Okay. You have made reference to lenders mortgage insurance—the LMIs—appearing tomorrow.

Dr Ellis : Yes.

Mr HUSIC: I noted in the media today—in particular, in The Sydney Morning Herald—concerns around the future of LMI in Australia and whether there should be capital concessions extended on the basis of the existence of LMI. I can see both sides of the argument, but I note that FSI did make a recommendation that capital concessions should be extended. Does the bank have a view about the extension of capital concessions for LMI? I would be interested in understanding if your organisation does take a position in respect of this.

Dr Ellis : We have written a fair deal about lenders mortgage insurance—obviously, not in this submission but in our submission to the financial system inquiry and, as you noted, in the financial stability review. Australia is quite unusual in the fact that we have private sector lenders mortgage insurance that is not government guaranteed. It is quite unusual for that to be the case. The capital concessions that exist for the smaller lenders are no doubt part of that. In the past it has also been regarded as being an important part of credit risk management and credit protection in the residential mortgage backed securities.

You mentioned that there are arguments on both sides, and I think that is right. I would not like to offer a position on whether the existing capital concessions should be extended to the larger banks—the ones that run their own models to determine their capital. I think that is a matter for APRA. But I certainly would encourage you to raise that with Wayne in the next session.

Mr HUSIC: It is my intention to do so. But it is also something that I have been considering, whether there is a longer-term issue to stability within the system—for instance, as has been noted today, if you have some major banks deciding not to maintain existing LMI arrangements and to self-insure. Basically, you can see a shift in the risk profile to LMIs taking on what would be considered more difficult or bad risk and whether that ultimately would lead to them making a decision about whether to maintain a presence in the business and what that does in the longer term for financial stability. Hence my question, because I do think that the RBA has an ongoing interest in this area of financial stability, obviously—

Dr Ellis : That is right.

Mr HUSIC: and whether this poses risks.

Dr Ellis : Yes. I think it is a complex question. The real issue here is how much capital there is against particular kinds of risks in the mortgage book. Typically, what you see is that the lenders mortgage insurance is called upon for loans that start off with an original loan-to-valuation ratio above 80. That is certainly what is baked into the rules for smaller banks, which are using the standardised Basel model.

It is interesting that actually even though the larger banks do not gain any capital benefit, that usually they do still require LMI. But they have the option of doing that through a captive, essentially saying, 'We'll hold the required capital, but we'll put it in this pocket and not in that pocket.' But within the group it is all there.

Our interest in financial stability is to ensure that the risk profile of the mortgage book is sound. That does not rule out high-LVR lending. I am moreworried about whether investors and existing trade-up buyers can have very low deposits, but I think there is an argument that the amount that you actually have to accumulate to get a down payment on a house is quite large, and so there is an argument to say, 'Well, you shouldn't expect absolutely everybody to have 20 per cent deposit. You might want five or 10 per cent of genuine savings for a first home buyer.' To the extent that that then presents greater risks in the event that house prices should fall at the wrong moment for that first home buyer, you may want additional capital held against that risk. I think it is an open question, whether you say that that additional capital should be in a separate entity where all of that high LVR risk is thereby concentrated or whether it is just that you want the banks to hold more capital against those exposures. I think both of those outcomes would be consistent with promoting financial stability.

Mr HUSIC: But if you are requiring a greater capital contribution to offset risk, it has been noted in terms of Australia's financial system the size of loans in the home lending market is relatively high compared to other countries, so if you are dealing with that risk through a higher capital contribution, being able to have that capital concession because of the existence of LMI would no doubt help in the management of risk.

Dr Ellis : What I would focus on is the total amount of capital and the risk management practices around those exposures, and I think it is an open question about whether you have that capital all in the bank or split between the bank and an insurer. That is something that the inquiry did give some consideration to. I want to understand the statement you made about Australia having a larger lending market than other countries—

Mr HUSIC: As a proportion of what the banks lend out to the housing market.

Dr Ellis : That is a true statement, yes.

Mr HUSIC: There has been an argument in one of the submissions made to us that, quite contrary to the statements made by others regarding supply, there is undersupply. That is a view being advanced by a number of parties. One submission argued that there is an oversupply of land for home development. Does the bank have any view on this? You have watched these things for quite some time so I would be interested in your take as to whether or not you believe there is oversupply or undersupply.

Dr Ellis : It is fair to say that we follow both the data and the debate on this matter. There are certainly longstanding challenges to the Australian housing system meeting the strong demand that we currently face but it is also fair to say that building activity has been quite strong of late and therefore there is a risk, that we outlined in some recent issues of the Financial stability review, that you could have pockets of oversupply in some areas, and the areas we pointed to were, in particular, inner-city apartments in Melbourne and to a lesser extent Brisbane. It is always going to be difficult when you have very strong population growth and strong income growth to meet that demand without seeing some price increase. That is how you get the new supply in. The housing stock is a stock and that is what is supplied, and the amount of new supply through construction is a very small part.

Mr HUSIC: It is a small part?

Dr Ellis : Construction is just a very small fraction relative to the stock of housing, and if demand increases it is very hard for construction to gear up fast enough to meet that additional demand without seeing some increase in prices at least initially.

Mr HUSIC: Building approvals are roughly just over 200,000 annually. How does this emerging supply match up with demand?

Dr Ellis : I think the answer to that question rests very strongly on what you think is going to happen to population growth. At the moment the official forecast is for population growth to pick up because of the quite rapid growth in net migration particularly of students, but we have also seen that those forecasts have not quite been met in the last couple of quarters—partly because the student numbers have not increased as much and partly because New Zealanders are not coming in net terms as strongly as they used to. There were some arriving and some going home, and there are more going home than there used to be. The net migration of New Zealanders has come off a fair way in recent quarters. So your assessment of that hangs on what you think is going to happen to population growth

Mr CRAIG KELLY: I have some questions on your initial comments. You mentioned, correctly, that housing costs are a large part of household spending. In some of the analysis the Reserve Bank does, do you make international comparisons about how much the average Australian spends on housing compared with other countries?

Dr Ellis : It is hard to get total housing costs. You can use household survey data to look at the total housing cost kind of metrics. What I would observe is that the purchase price to income ratio is in the middle of the pack of a range of countries. It is about the same as Canada but it is lower than France—I can dig out the graph and read off some those countries if you would like me to. Both in terms of the debt to income ratios and the housing price to income ratios Australia is in the middle of the pack of a range of industrialised countries. In the sense that a lot of us live in big cities and we have quite big houses relative to many other countries, we are more likely to live in detached houses than in apartments than people in many other countries. Yet we are still in the middle of the pack in terms of what that looks like relative to our incomes.

The interest payments, relative to income, are also relatively low relative to historical averages in Australia. To the best of my knowledge the absolute level of interest payments, relative to income, of the Australian household sector is not that unusual compared to the historical norms in other countries, although many other countries have very low interest rates at the moment, including on their mortgages, so at the moment they have very low interest payments relative to their incomes. But if you take the broad sweep of history, I would not regard it as being supported by the data to say that Australians spend absolutely more on housing than other countries. Although we actually get more housing for that than many other countries.

Mr CRAIG KELLY: That leads into another question. When you are looking at the data on inflation of housing over a number of years, does the Reserve Bank look at changes in what the actual quality of the housing is. I know that when the ABS looks at other data they make adjustments for the changes in quality. You mentioned that houses are getting bigger in general, but then again anecdotally it would seem that our block sizes are getting smaller. Would you like to comment on that?

Dr Ellis : Sure. A number of the housing price measures that we follow in the Reserve Bank are what is called hedonically adjusted. So they do make a best-efforts attempt to adjust for those kinds of things. The average house is now further out than it used to be as the cities expand. Block sizes tend to be smaller, but the houses tend to be bigger and nicer. One thing we know from the ABS survey of income and housing is that across the age groups and across most household types, other than single-parent families, the average number of spare bedrooms actually increased between 1997 and 2011-12. For a given household type and size, people have more house than they used to have. It is also the case that although there has been some shift towards having more apartments in the Australian housing stock, it is still a relatively low share compared to many other countries. Australia and New Zealand have very high shares of detached housing in our housing stock. So we are kind of getting more housing and, at the margin, the quality of housing has increased over the period in which prices, even adjusted for that,—certainly, a median price—do not really adjust for that well. So if the median house has become nicer, better appointed and more has a more modern kitchen and four bedrooms rather than three, that will not be captured in a median price. To some extent you will see some of that adjusted for in some hedonic measures that some of the private sector suppliers of this data produce. So you can triangulate against all of those series.

Mr CRAIG KELLY: How about the other side of the coin, where we see the block sizes are reducing in size. A family might have less room to play cricket in the back yard or to kick a football around. That is a negative effect, while a better kitchen or better quality bathroom might be a positive. What about the negative effects of a smaller block size? Are they also considered in the data?

Dr Ellis: Block size is one of the pieces of information that are recorded in housing sales and land titles numbers. On the real estate websites or some of the apps that are available to estimate house prices will tell you the block size. That is something we know something about. But, of course, the block sizes get smaller for new housing. As I mentioned, new build is a small fraction of the total. So the quarter acre blocks that were built upon in the post-war period and the decades thereafter are mostly still quarter acre blocks. There are places, Melbourne in particular, where you see a lot of dual occupancy on some of those larger blocks. People are allowed to do dual occupancy in Victoria, so you do see people making that choice. But it is a choice. If you think about 1,100 square metres—I hope I am not misleading you with my recollection that that is roughly a quarter acre—that is actually quite a large block size. I think it is also important to remember that there are different kinds of families and households in Australia and not everyone needs 1,100 square metres. Once you get to a certain age and the kids are off your hands, you do not want to be looking after an 1,100 square metre garden any more. I would point out that there are still larger blocks, but there is now more variety in block sizes within the housing stock and people can make a choice about whether they would prefer to have a three-bedroom brick veneer home with a very large block, which was the typical housing form that you saw in the 1960s or 1970s, or whether they would prefer the five-bedroom house with a media room and parents retreat and all of those things on the smaller block. I think there is more variety now in the housing stock than there used to be as a consequence.

Mr CRAIG KELLY: I have another question on your interpretation of data. For example, at Moorebank in my electorate, a house recently sold for $1.5 million. That obviously kicks the median and average prices up and sets a new benchmark. But that house was being sold for redevelopment. Someone was not buying it to move into it. Therefore, can that figure artificially skew the housing affordability numbers up to make them look worse, when in fact it is actually a developer buying the proper to demolish and build units, rather than to occupy the house?

Dr Ellis: It tells you something about land values. To the extent that there was arbitrage between land values for houses you live in and houses you knock down, it still tells you something. If there were just one property being sold for that amount, that would not actually change the median, although it would change the average.

Mr CRAIG KELLY: There were several properties sold, because of a rezoning application. But it is not actually affecting affordability for people buying into the market, because that one block, which was $1.5 million may have half a dozen townhouses built on it, which might have a lower price. Therefore, the person who is trying to get into the market to buy a house to live in is not paying that $1.5 million. But could those higher prices possibly skew the figures, and does the RBA make any allowance for that or look at it in any way.

Dr Ellis: I think the hedonic numbers do produce some adjustment for that, just because of the way you actually do those. But it is the case that we are seeing very strong prices paid for development sites, so essentially no value or a negative value is being attributed to the structure, because you have to knock it down, and that costs money. That is telling you something about the land value and the willingness of developers to pay those kinds of prices. Every property is different and you do the best you can. That is why we look at a range of different data providers that are compiling statistics on a range of different bases.

Mr CRAIG KELLY: In your opening statement you talked about frequent moves that renters might have to make, and the cost of that. On the other aspect, what about the high cost of stamp duty. Take someone who lives in Sydney who wants to move from one side of Sydney to the other because of employment opportunities, or someone who is moving from a country area to a city area. Does the RBA have any concerns about the economic constraints from the high cost of actually moving, because of the stamp duty?

Dr Ellis : The cost of stamp duty is a low single-figure per cent of the price. You also have to add into the cost of buying and selling homes not just the stamp duty but the legal fees, the real estate agent fees and so forth. I think it is the case everywhere that buying and selling property involves considerable transaction costs. I wish we had more comprehensive data on this, but one of the things I have observed in the 15 or 20 years I have spent watching the housing market is that typically in Australia a real estate agent fee is about two per cent of the sale price—give or take, depending on which firm you go to and the nature of the property. As best as we can tell, in the US and Canada it is a number more like five or six per cent. So you are still paying the same amount of transaction cost, but you are paying it to a real estate agent instead of to the state government. I am not exactly sure—

Mr CRAIG KELLY: So are you saying the real estate agent fees are lower in Australia or much higher in Australia?

Dr Ellis : Lower, surprising as that may seem. We do not have comprehensive data on that, but I have certainly asked many US housing experts about this over the years, and they will always tell you, 'Yes, realtor fees are six per cent.' Often there is a buyer's agent fee as well, so the cost of transacting in that market is, if anything, higher relative to the house price. Transaction costs do make it more expensive to move, and that is, presumably, a disincentive to buy and sell properties if you want to move. Of course, one of the things that is true about the Australian system is that it is also very easy to rent out the house you already own. One quite typical pattern is that if someone is moving for a job for a while but they plan to come back to that area, then they will rent out the house they own and previously lived in, and they will either buy another house or rent where they are currently working with the expectation that they will come back to their old house. In fact, both my parents and my sister and brother-in-law have done exactly that in pursuit of job opportunities. It is quite a typical thing for people to do. The fungibility between owner occupation and rental in the Australian housing market I think adds to that flexibility. You do not actually have to sell your home to take advantage of a job opportunity.

It is also worth noting that the disincentive to buy and sell posed by all of these transaction costs is a disincentive to flip. From a financial stability point of view, flipping—the buy and sell quickly; give it a lick of paint and then sell it on very quickly—is a really speculative activity that we saw a lot of in the US during their boom. To the extent that there is a disincentive for that kind of activity, I would regard that as being amenable to promoting financial stability.

Mr CRAIG KELLY: On investment in housing, you say:

Investor interest in property has been especially strong in recent years, no doubt partly encouraged by low interest rates and the prospect of … capital gains.

Do you think that prospect, where people are anticipating future capital gains, is perhaps overly inflated?

Dr Ellis : I do not want to provide a forecast of house prices. I certainly would not like to do so. I would point out that, as long as you have got trend nominal income growth and prices of other goods and services rising, it would make sense to have at least some nominal price growth in housing services and, therefore, the housing that provides those housing services. Some capital gain of some amount over a very long horizon can be anticipated, but there will be periods when house prices fall. As I have previously said publicly in speeches, because that transition to the low-inflation, financially-liberalised world with a higher house-price-to-income ratio is now finished, we actually anticipate that nominal price growth will be slower on average than it was over the 15 years to 2005 and certainly than it was over the high-inflation period where there was an inflation component to that.

It then stands to reason that if housing prices are cycling around a lower average, you are more likely to have brief periods where housing prices fall in an absolute sense. In fact, if you look at Sydney housing prices there have actually been three such periods since 2003. People do have to bear in mind that risk, but in a very long holding period trend sense you could anticipate that there will be some capital gains and those capital gains are concessionally taxed, as I mentioned in my opening statement.

Mr CRAIG KELLY: On the issue of negative gearing, you mentioned that it might be something that needs to be looked at in housing. Are there any other asset classes in which negative gearing should be looked at? Or is housing a special, stand-alone issue?

Dr Ellis : We are not suggesting that negative gearing be looked at in isolation. We think a holistic review of all the tax incentives to engage in leveraged asset accumulation are worthy of review.

Mr CRAIG KELLY: You are saying not housing in particular but all asset classes?

Dr Ellis : I think the important point to make there is that property, both commercial and residential, is the asset class you can leverage the most. The tax system does not discriminate against asset classes in terms of your ability to negatively gear them. It is just more possible to negatively gear property, because you can gear it more, because of the amount of leverage you can take is higher.

Mr CRAIG KELLY: But therefore wouldn't prohibiting negative gearing on one type of asset class against another potentially cause some distortions?

Dr Ellis : That is the point. We are not making any particular suggestions about the overall principle of treating assets equally—about how the tax system might be changed or not changed at all. We are simply saying that the interaction of concessional capital gains tax and the ability to deduct all your expenses, including the expenses involved in taking leverage, is worthy of a look. We are not suggesting that any particular asset class should be treated differently from any other asset class. There are a number of fundamental principles about the ability to deduct expenses that you do not necessarily want to distort for particular asset classes.

Mr CRAIG KELLY: But you would agree that if different asset classes had different treatments as far as capital gains were concerned then that could lead to some distortions in the economy?

Dr Ellis : I think the distinction is not that different assets are treated differently; it is just that not all assets offer the prospect of capital gains. And to the extent that you can gear an asset it is treated the same, but it is not feasible to gear all assets to the same extent as can be done for property. It is not that property is treated differently; it is just that the effect is quite particular.

Mr GILES: Thank you, Dr Ellis, for a submission that I found very helpful. I have a couple of questions, firstly to clarify in my mind the evidence you gave in response to the chair's opening questions on negative gearing. I do not want to put words into his mouth or your mouth, but he was effectively suggesting to you that at a time when interest rates are lower than rental yields negative gearing is not a significant issue. Your evidence, as I think I understand it, is that both investors and lenders are looking at a longer perspective, so that assertion on the part of the chair is not one that you agree with. Is that right?

Dr Ellis : I think we can make a synthesis of both of those statements. In the current environment, the mortgage rate—actually, let me just double-check the numbers—

Mr Aylmer : The average outstanding interest rate on mortgages at the moment is about 4.7 per cent. That is across all banks. The rental yield is I think about 3½ per cent, but that is a real yield, as opposed to the interest rate on mortgages, which is a nominal rate.

Mr GILES: And if you go to page 20 of your submission—this is data that is obviously a little bit out of date—the suggestion is not that negative gearing is decreasing in the period from 1999 to the most recent data that is available.

Dr Ellis : I think I would refer to the tax data that says that there are people who are declaring a net loss, so, definitionally, they are negative gearing.

Mr GILES: And rather more of them in the most recent year for which we have data than in the late 1990s.

Dr Ellis : That is also correct. I do not think the rental yield was in the submission.

Mr Aylmer : It is about 3½ per cent, but it is a real yield.

Dr Ellis : Here is the data I was looking for: yes, it is 3½ to four. At the moment the rental yield is still below the average mortgage rate. Whether someone is negatively geared or not depends on the extent of the other expenses. There is insurance and maintenance and all the council rates—water rates and so forth—that you have to pay, and they are also deductible expenses. How much you are deducting on interest depends on how much gearing you actually have on that particular property. If you have very low gearing, very high equity in that property, you are not deducting much interest. Individual investor circumstances will differ. It is certainly the case that as interest rates fall fewer people will be negatively gearing, on average, but the tax data does suggest that there are still people negatively gearing.

Mr GILES: I want to take you to another aspect of this debate and to a speech made by Philip Lowe in November 2013. He touched on, amongst other things, the importance of infrastructure investment, without going directly to this but focusing on outer suburban areas—those areas where new home ownership purchasing is significant. Do you have any comments to make on that issue in light of broader public policy challenges around productivity, maximising agglomeration benefits and the like?

Dr Ellis : I would also refer to a speech I gave on that topic in May last year. I made the point that Australian cities are unusually low-density, so buying at the fringe is buying a lot further out than is buying at the fringe in many other countries, and certainly a lot further out than buying at the fringe used to be. For that reason, it is imperative to make sure that people right through the city can access all the affordances of that city, including the job opportunities. So yes, the deputy governor and I would agree on the value of having good transport infrastructure right though the city. And we would note that that is actually quite difficult to do, given how low-density our cities are. That comes back to the issue of block size that we discussed earlier.

Mr GILES: Indeed. In a country as urbanised and as suburbanised as Australia is, the consequences of this issue are more profound than in other OECD nations.

Dr Ellis : I think that is a reasonable conclusion to draw.

Mr GILES: There is one other matter I want to touch on very briefly. You are probably aware of a paper put out by the Australia Institute in April this year on negative gearing and capital gains tax.

Dr Ellis : Yes.

Mr GILES: It in part touched upon the commentary of the RBA, the then-most-recent minutes of the Reserve Bank board, which presumably would be March or thereabouts. The commentary is:

The speculative nature of Sydney’s residential investment property, affected by negative gearing and the CGT discount, is making interest rate changes less effective. The consequences of this might be that the Australian economy is deprived of further monetary policy stimulation and remains weaker for longer because negative gearing and the CGT discount encourage speculative behaviour in the housing market.

Do you have any comment on that, or is that perhaps a matter you might take on notice?

Dr Ellis : I am not familiar with the specific wording of that report. There are a lot of reports to keep on top of, so I think I would prefer to take that one on notice.

Mr COLEMAN: Thank you both for being here today. I just want to take a step back in the context of this discussion about housing. Australia's mean household wealth is, according to the ABS, over $700,000. And according to various global indexes from Credit Suisse and others we have the highest mean household wealth in the world, most of which is held in property. Doesn't that suggest that the existing policies, from the perspective of building wealth in the community, are working quite effectively? And shouldn't that make policy makers reluctant to make changes?

Dr Ellis : It is an interesting way to frame the question. Coming back to our mandated policy responsibilities for monetary policy and financial stability and the fact that housing is an important part of the wealth of the Australian household sector, to the extent that it is financed with leverage, most of those mortgages have a variable interest rate. That makes some channels of monetary policy very powerful, so that is probably a good thing. I think from a financial stability point of view the important question is the sustainability of the current level of that wealth—essentially the sustainability of the level of prices and avoiding mass destruction. I guess one needs to be alert to the prospect that a particular element of that household wealth portfolio is at unsustainable valuations. I am not suggesting that is the case, but you do not necessarily want to push too hard to push prices up in a way that may not be sustainable.

Mr COLEMAN: I think it has been the case for some years, though, that our average household wealth has been either the highest in the world or very close to the highest in the world, so this is not some new thing that has spiked up in the past 12 months. You do obviously have that stability goal, which is very important, but you also have a goal of economic prosperity and welfare for the people of Australia. Presumably the actual accumulation of wealth for average households is a very positive thing for the economic prosperity of Australia.

Dr Ellis : I would make two points in relation to that. One is that I think the best contribution we can make to the prosperity of the Australian household sector is to avoid having the kind of financial crisis or meltdown that we have seen in many other countries. That is how you end up destroying the wealth and damaging the prosperity of the Australian people. To see that, you only have to look at the unemployment numbers and the household wealth numbers for countries that were more affected by the crisis.

The other point I would make is that distribution matters. The average might be one number, but, in thinking about the prosperity of the Australian people and in particular their financial resilience and financial stability, it matters how that is distributed. That would be something we would be looking at very carefully as well.

Mr COLEMAN: Indeed it does. That is why it is important that it is our median wealth in addition to our average wealth which is the highest in the world. This is not a situation where we have a very spiky, lumpy distribution of wealth, as seen in other nations. Certainly that is an important point. I want to ask also about negative gearing. Negative gearing is something that is discussed often but perhaps not often plainly described. Negative gearing is effectively the capacity to deduct from your income tax a loss from interest relative to the income you get from that income-producing asset. It seems to me that that is a fairly fundamental point in the economy generally—that if you have a loss or a business expense then that is something you can deduct against your income, because it is a cost that you had. I guess I am just interested in the idea of why you would be open to the notion of treating that form of expense—interest—differently to other forms of expense and why interest potentially should be taxed, that there should effectively be a tax increase in relation to costs associated with interest. I am assuming that you would not advocate that other costs of holding a property or holding shares or whatever should also be subject to less-favourable tax treatment.

Dr Ellis : We have not advocated any particular tax measure, and we certainly have not specifically advocated treating interest expenses differently from other expenses. You are quite correct that all negative gearing really means is that if you have expenses involved in earning certain types of income then you can deduct them against all your income, not just that income. There are plenty of countries where that is not a provision. In the UK and the US, for example, you can deduct your property-related expenses against property income but not against wage income, so different countries have come up with different solutions to this. We are not advocating any specific change, and that is why we said in our submission that we are not suggesting that it is all about negative gearing; we are saying that it is worthy of a holistic review, particularly given the combination of concessional capital gains tax and the ability to deduct all expenses against all income—to look at what that does for the incentive to leverage into capital gains producing assets.

Mr COLEMAN: Couldn't that argument logically extend to all business interest expenses? Because if your argument is that interest expenses are deductible, and the ultimate sale of the asset is subject to so-called favourable capital gains tax treatment, that, as I understand it, is the case for the operation of you average Pty Ltd company in Australia—they can deduct interest expenses in the cost of running a shop or whatever. Why would we seek to potentially treat those things differently?

Dr Ellis : We are not advocating that they would be. I would, however, note that in a case of a corporation, the corporation can only deduct its expenses against its own income. Whereas, if you think about an individual, they may have a business renting property to people and a wage. What the Australian tax system allows is that you can have business expenses in the form of the expenses that you have against your rental income business that can then also be deducted against another source of income. I think that would be one way of making that distinction but, as I emphasised earlier, we have not made any specific recommendations about whether anything should be changed or what should be changed. We have simply said that it is worth a holistic review.

Mr COLEMAN: Sure, and on that distinction between a business and an individual, I think that from an individual's perspective, they are one economic entity, whether they are happy to have an investment property or a wage, or whatever—they are not split personalities, so to speak. One last question, and it is a hard one to answer. It is quite notable that yields have declined, and therefore the price earnings ratio on housing has increased quite significantly—I guess you are looking at about a 30 times PE ratio on investment property, when historically there were much lower averages, as you have noted before. In your view, is that largely a function of low interest rates meaning that the serviceability enables people to cope with a lower rental yield, or is there something more fundamental going on in the economy where investors have somehow re-rated investment property and determined that lower yields are more acceptable than they were historically?

Dr Ellis : I think that there is some element of truth to both of those explanations. Real interest rates have come down over the last few decades, and you can see that in market interest rates—indexed bonds and so forth, as well as rental yields, which are a real yield, as Chris mentioned, so there is some element of that. The rental yield, on the best measurement we have, has been cycling around a relatively flat number for about a decade. Again, this is one of those things that there is a point of inflection around, because between 2003 and 2006 that transition ended, and one of those things was that the rental yields stopped coming down. So at the moment the rental yield is at the bottom of that cycle, but it is cycling around a flat trend, and there had previously been a shift.

Mr Aylmer : I think that is right. It is marked—that change around 2003. You even see that in leverage as well, because if we look at the rate of growth of investor housing, it is currently about 10½ per cent. The average over the past 10 years has been about eight per cent. In the previous decade, the average growth in investor housing was 23 per cent. So you see this incredible structural shift from about 2003 onwards.

Dr Ellis : And it is worth pointing out that that was a structural shift that the Reserve Bank predicted ahead of time. There are a number of papers—going back through 2001, 2002, 2003—and, in particular, there was a Bulletin article in March or April 2003, and a speech by Ian Macfarlane around the same time period on the topic of household debt. We talked about that transition and explained that all these things were in transition and that transition was going to end. I think that we got the orders or magnitude and the behaviour right.

Mr CONROY: I am really interested in your observations around housing affordability for young people and what has driven the change. That is something you have covered that others have not. Congratulations. On page 8 of your submission you said you believe one of the key indicators of declining housing affordability for young people is the rise in inequality in Australia over recent decades. Could you amplify that point?

Dr Ellis : We would not regard ourselves as being the expert on this particular issue. In fact, you might notice that on that page there is a citation to Yates (2011)—and I understand that you will be speaking to Judy Yates tomorrow. Judy Yates has done a lot of work on this over the decades. It is one thing that may be a contributor. I think the big-picture thing that has driven the decline in home ownership rates for the younger age groups is that people are getting married at a later age. Before you get married and settle down you do not necessarily want to buy a house that then is not appropriate to your needs once you are in a family with children. So, typically, people want to buy a house at around about the time they settle down. The average age of marriage—and, of course, not everybody does that—has come up by about seven years over the last three decades. So it has basically gone back to the more historic norms of marriage age rather than the anomalous period of the postwar years when the median age at first marriage for a woman was under 21. That is certainly not the case now. So, essentially, people are making choices to settle down later. So it makes sense for them to buy their first home later.

Obviously also people are staying in education longer—school retention and university participation have gone up. But if you have a larger fraction of people with more variable employment arrangements—perhaps less secure employment arrangements—that makes it harder for you to accumulate a deposit and qualify for a loan. By and large, in the big picture, the decline in the home ownership rate for younger groups happened before the big increase in housing prices relative to income. That means it cannot be because of that—so it must be because of something else. We think the most plausible explanation for the bulk of that shift was the demographic choice after the baby boom ended.

Mr CONROY: On page 9 of your submission you say increased labour market flexibility and other factors leading to employment and income insecurity are also driving that. As you said, that predates the price explosion in the late nineties and early 2000s; it began in the early nineties to some extent. So you accept that that is probably a factor as well?

Dr Ellis : Yes. My own view is that the dominant factor is the demographics and the change in education, participation and age at first marriage. Certainly based on the work that Judy Yates has done on this in the past, we find it quite compelling that some of these things would also have contributed.

Mr CONROY: I do not want to confuse causation and correlation. You are saying the key factor is that people are getting married later—for lifestyle choices or their ability to have a double income and save a deposit. Is there any validity in switching the argument around and saying that people are getting married later because their jobs are less secure and they need more education to get decent jobs because they cannot afford homes? Are all those factors driving people to get married later? They are living with their parents longer.

Dr Ellis : It is an interesting hypothesis. One of the interesting facts that we have uncovered since making the submission is that the share of young adults aged between 25 and 34 living with their parents is no different in Australia than countries such as the US, the UK and Canada—it is around 15 per cent. We have not done reams of research papers on this. Our best analysis at this point is that the decision to settle down drives your decision to buy a house. You might meet the right person and live with them while renting. It seems to me more likely that coupling decisions drive decisions to buy rather than the other way around—because you could still couple and rent.

Mr CONROY: Switching from young people to inequality more generally, you make the point that the decline in home ownership rates has been most marked and concentrated in the third and fourth income quintiles. So, clearly, inequality is driving this to some extent across all age classes. The top quintile is stable. It is really the bottom, third and fourth quintiles where the home ownership issue is becoming more marked.

Dr Ellis : Yes, but I do not know for sure whether that is not just an artefact of where the younger households are in the income distribution. I would anticipate that people at the very top of the income distribution are probably not 25 years old—on average.

Mr CONROY: I want to go to the behaviour of investors. I asked Treasury about this when they appeared. We had a big explosion in house prices in the early 2000s. One factor clearly was a decline in inflation expectations; people accepted that the period of low inflation was here to stay so they could speculate a bit more. But, surely, a big factor was the capital gains tax concession being reduced in 1999. To the extent that investors were driving up house prices, do you accept that that was one of the key factors?

Dr Ellis : There were a number of things going on at the same time. The one that we have pointed to in terms of a secular shift has been financial liberalisation and disinflation. We believe that was a one-off shift that is sustainable. Of course, it took time for house prices to get to their new equilibrium relative to income. Every boom starts with something real and fundamental, but you can then get people wanting to extrapolate that. It would be difficult to say how much of that investor participation was simply in the late stage of a real boom becoming more speculative—in the way that Charles Kindleberger wrote about so many years ago—or whether it was also the change in the tax. I would not like to attribute a share of the causation to one or the other. I cannot rule out that the tax change made a difference, but I do not have a view about how much of it was just the extension of the boom and then people extrapolating that versus tax incentives.

Mr CONROY: But you do accept that it was one of the factors?

Dr Ellis : Yes, it is entirely possible that that was one of the factors.

Mr CONROY: We are also seeing changes in investor behaviour. In the eighties the investor share of dwelling finance was between 10 and 20 per cent. It is now between 40 and 50 per cent, depending on the market. But your submission makes the point that their share of new dwellings is roughly the same. So, clearly, they are showing a predilection for buying existing dwellings. Have I read that correctly in your submission?

Dr Ellis : Based on what we know about the nature of the property being secured from the loan approval numbers—and these are loan approvals, so it is the flow of new lending. Chris, do you want to talk about that?

Mr Aylmer : Yes. The numbers are for the stock of housing loans. About one-third of those relate to investors. But the 40 to 50 per cent number that you referred to is the flow of new loans. What that is saying is that, over time, that share, which is currently one-third, of outstanding housing loans will get bigger because there are more investors in the market. The other interesting thing about that is that we have traditionally thought of first home buyers going into the owner-occupier market, but our sense from liaison is that, increasingly, they are appearing as investors as well.

Mr CONROY: Let's leave aside what this increased investor focus is doing to rental yields—suppressing rental prices. Speculation is driving investor activity but it is not increasing new housing stock—they are not buying new houses at a greater ratio. So investor behaviour is not adding to housing supply more than it has done historically?

Mr Aylmer : I am not sure about that assertion about investors not adding to supply more so than they did in the past. Investors end up being disproportionately represented in the new construction sector—in particular, units.

Mr CONROY: Your own statistics say that the share is roughly the same as it was 20 years ago.

Dr Ellis : That is correct.

Mr CONROY: Leaving aside their impact on rental yields, if the tremendous investment flow is into existing dwellings then it is not boosting new housing supply more than it has done on a long-term basis?

Mr Aylmer : Established housing is normally where you get the initial response; it flows through to the prices of established houses and then you get a supply response. The next leg of that would in fact be more supply coming on because of the incentive. Even if you look at the Sydney market, you are already seeing signs of that in the prices. According to RP Data, the increase in detached dwellings over the past year or so has been about 23 per cent but the increase in the price of units has only been eight per cent—which is suggestive of a supply response coming through. That marked differential is also true in Melbourne, where those numbers are eight per cent and two per cent.

Mr CONROY: Who are these new investors? My colleagues talked about the difference between ATO data, which is net income, versus the HILDA surveys and everything else. You made a good point in your submission that has not been picked up by other people, which is that one of the big drivers of the discrepancy could be retirees post the Howard-Costello decision to exempt them from the tax system for their super income. This means they might appear as very low income earners but in fact they are just very low taxpayers. Is your instinct that that would account for a lot of the people whom some in the housing industry say are very low income earners who are negative gearers?

Dr Ellis : That is the way to square the two different datasets and we think that is what is going on. I think we even had a footnote somewhere in our submission that made that exact point.

Mr CONROY: Yes, I saw that footnote.

Dr Ellis : So that is certainly how you would reconcile those two datasets. In the HILDA survey we get the unit data, so we can say make a distinction between a retired person with a low income and a younger household with a low income. My recollection of the HILDA data is that most of the lower income people who say they own investment property in the HILDA survey are older people.

Mr CONROY: On investment incentives—and obviously my colleagues were talking about negative gearing and the impact there—you say in your submission that investment in dwellings is advantaged in comparison to other classes of investment due to the ability to leverage greater. So anything that increases investment concessions—such as the CGT concession—will have an amplified effect on dwelling investment rather than share investment? Is that a reasonable proposition, since it is amplified by the leverage issue?

Dr Ellis : We are not tax experts but I think that is a reasonable supposition to make. I think we would also make the point that the treatment of investment property is different from the treatment of owner occupation. Australia does have a tax system that provides some advantage to being a relatively un-leveraged owner-occupier—because obviously you cannot deduct mortgage interest, unlike in the US. So people have a really powerful incentive to build up offset accounts, redraw accounts. We think that is really positive for the financial resilience of the Australian household sector. Of course, they do not pay tax on the imputed rent either. It is regarded as being somewhat advantageous because you are not paying tax on imputed rents or capital gains on owner occupied property. But, at the same time, you cannot deduct expenses. So you have this symmetry in the treatment even though they are treated differently.

Mr CONROY: But the point about the CGT concession and other investment incentives, such as amplifying investment in housing versus shares because of the leverage issue, anecdotally that would be supported by the fact that, as is pointed out in the Treasury submission, since about 2001 rental yields are well below share dividend yields. That was the first time that had occurred since statistics had been collected, and now it is permanent. So if you are an investor choosing between these investment classes, the fact that your gains are amplified because you can leverage more on property investment combined with a CGT concession argues that there is an in-built bias towards dwelling investment.

Dr Ellis : It certainly changes the incentives, but in comparing different asset classes one also needs to consider the relative riskiness and the volatility of the prices. As we have seen over the last several decades, the absolute volatility in equity prices is higher and so your risk adjusted return may differ. That is something that an investor would bear in mind.

Mr CONROY: The RBA is one of the two most pre-eminent economic advisers to the government of the day. Do you have any view on why capital gains should be privileged in their treatment versus other income streams? Is there any economic justification for that at a broad level?

Dr Ellis : As we have said in the past we are not tax experts, so I would not want to offer a view on the rationale for that. It is worth bearing in mind that, prior to 1999, there was not full taxation on the whole gain. The taxation was only on the real gain—the CPI adjusted gain. You need to allow for the fact that some capital gain is compensation for inflation. There are different ways that people may seek to achieve that.

Mr CONROY: Leaving aside the real-versus-nominal aspect, is there an economic justification for privileging capital gains over income from work, for example?

Dr Ellis : I do not think you can leave the real-versus-nominal aspect aside on this one. It is not uncommon for countries to have concessional taxation of capital income. I think it is for the political system to decide whether that is something it wants to encourage or not. I do not think the Reserve Bank has a view.

Mr CONROY: My final question is on the historical experience in the 1980s when negative gearing was briefly removed. Some people claim that drove a spike in rent. But I have seen further analysis that made the point that the increase in rent was localised to Sydney and Perth, which had other factors driving that. If you look at other capital cities, rent fell or was flat. Can you form any conclusion on the impact of removing negative gearing on rental prices in the 1980s? Do you have a view about whether it did cause a rent spike?

Dr Ellis : The first point I would make is that we are not advocating for the removal of negative gearing.

Mr CONROY: I understand that.

Dr Ellis : The second point I would make is that the word 'briefly' is quite important here. My understanding of the rental statistics is that you have characterised the outcomes during the 1980s correctly. That is what happened. But I think it is very hard to extrapolate from that experience to the hypothetical in which there was such a change simply because it was only for a short period of time and so people would not have had time to adjust their portfolios in response to the preferential tax treatment.

Mr CONROY: So, from an economic policy point of view, you cannot draw any conclusions from that experience?

Dr Ellis : You certainly could not draw strong conclusions from it.

Mr CONROY: I appreciate that.

Mr BUCHHOLZ: Today we have heard a lot about the Sydney market. The inquiry is into homeownership in Australia. I just want to go for a run around the country and get your overview. We will start with Western Australia; how do you see the regional areas in the real estate sector? At a glance it looks like they may have softened as a result of the softening of the resources sector. Have you any comment there?

Dr Ellis : Sure. It is correct that, outside of Sydney and Melbourne, the property markets have been rather more moderate over recent years. It is also true that, in areas where there had been a substantial run-up when there were big population pressures and income pressures due to the mining boom, that has all receded now. So the fundamentals themselves were very volatile and then you have seen that in the price.

Mr BUCHHOLZ: In Queensland, let us take a run up to the top end of our coast, to Townsville and Cairns. Mostly their incomes would be hinged on the tourism sector. We are just starting to see some lights of recovery now, on the back of the softening Aussie dollar. As to housing prices up there, we have not seen any spikes or a return to strengthening numbers just yet.

Dr Ellis : Yes, the data we have available would suggest that. As I mentioned earlier, outside of Sydney and Melbourne there really has not been a huge amount of strong growth. Individual locations are related; people can make choices to move to cheaper areas, should they want to do so, so there is some connection between all the different geographic markets. But each of those markets is subject to its own pressures and different supply-and-demand fundamentals. So you will get different outcomes in different areas.

Mr BUCHHOLZ: People do have the option of moving, and I would encourage anyone who could not afford to buy a house in Sydney to shift to Queensland—particularly to the Gold Coast, where we would love to have them. Given that there is such a diversity in the markets, where you have a couple that are hot but the bulk of the country is really in a softening position, how does the Reserve Bank balance the levers to reduce the tension in one without adversely hampering the bulk of the country, which is looking for some type of stimulus?

Mr Aylmer : Basically what the board aims to ensure is that the broad structure of the interest rates that are faced by households and businesses is consistent with the desired stance of monetary policy. So there really is only one instrument, and, when the board is targeting an interest rate, they are thinking about the overall interest rate structure faced by households and businesses. The cash rate, effectively, is an anchor for that broad structure of interest rates.

Mr BUCHHOLZ: Do you have any comment on some of the movements by APRA, as to some of the barriers to entry, to try and offset the heated markets? The liquidity around banks is a bit separate. But what about the loan-value ratio shifting out a bit? What effect would that have in both markets?

Dr Ellis : By and large, the prudential standards and the supervisory settings that APRA is responsible for are either at the national level or at the consolidated group level, in terms of capital. Some of our banks have international operations and they are also covered by APRA's prudential requirements. To date, neither the Reserve Bank nor APRA have been in the business of saying, 'Do this in this town and do that in that town.' But those of us who spend a lot of time understanding banks' lending policies and practices will see that it is part of good credit risk management to take into account the differential experiences of different areas. So one of the things we started to see was that, while there was still the boom going on in prices in some mining districts, the banks were saying, 'As part of our management of our business and of managing the risks in our business, we do not want to offer loan to valuation ratios in mining areas that are as high as we will offer in other regions because we are concerned about this forthcoming price cycle and getting caught on the downside.' Banks manage their geographic exposures and banks that have very concentrated geographic exposures obviously have to hold more capital for that because you then have a more correlated risk in your mortgage book, but that is part of good credit risk management in any lending institution.

Mr BUCHHOLZ: At the moment there are a lot of empty new homes in mining towns that were built at the top end of the market curve. I just ask for that to always be considered: Australia is a large country; it is bigger than the Sydney market, it is bigger than the Melbourne market. I understand the emotional debate that exists in those two localities, but the rest of Australia does not have the same strength of market. On a different line of questioning, I would like your opinions. There may be no data to support this, so it might be just an opinion. For those entering as first home owners, is there a higher expectation today of what a first home looks like to what it was, say, when I first purchased. I can describe my first time. It was somewhat of a housing commission home: lowset, weatherboard, three-bedroom. I went to a display village on the weekend purely to try to prepare for this inquiry. They were beautiful brand-new homes. I asked, 'Who is your target market? Who is this home designed for?' They said, 'The first home owner.' This was a beautiful four-bedroom home with double lock-up garage, ensuite, modern kitchen with appliances and fully decked out. The yard was done and some have pools. Has the Reserve Bank noticed this? Is there any data to support this—that the expectation of a first home owner is potentially vastly different to that of their parents?

Dr Ellis : That is a really interesting question. There is no data on people's expectations of what they should have, but the data on the survey of income and housing that I mentioned earlier does suggest that, on average, households have more bedrooms and more living space than they used to—certainly more than the house I grew up in. It is not just about the actual structure; it is also about the services that we expect in new developments. I was always told as a child that, when we first bought the house, there was an unmade road and a septic tank. Nowadays, for a new subdivision pitched to first home buyers, they would not be saying, 'By the way, the road has not been made yet and there's a septic tank. That will come later.' Incomes have increased a lot since the 1960s and part of that is that people expect to be able to consume more housing, given their higher incomes.

I just want to come back to the comment you made earlier. I think the Reserve Bank is very cognisant of the variation within Australia and, as Chris said earlier, we are making monetary policy for the whole country. If monetary policy were about one very small segment of the housing market, it would not look as it does, but it is not. I just wanted to emphasise that before going forward.

I think it is the case that, as incomes have risen, people want to consume more stuff and that includes more housing. People also focus on the ability of people to buy the median priced house. Of course, first home buyers have never bought the median priced house. One of the things we have focused on has been: what does the distribution under that look like? One interesting thing is that the bottom 25 per cent of dwellings in Sydney have prices that are noticeably, in percentage terms, lower than the bottom 25 per cent of houses relative to the medians in other cities. There is more spread of those. There is something for people to buy. There is the question: have we set the bottom of the standards too low?

I am neither a tax expert nor an expert on building regulation. Certainly the standards that are being put in for building regulation are all things like fire safety. Do we really want to say it is okay to reduce that? We cannot build fibro houses anymore, for very good safety reasons. We cannot build houses that are not safe for fires, for very good reasons. So is it really the case that the way for first home buyers to be able to purchase their first home is to find ways for them to have less housing? I guess in some sense the shift towards apartments has done that.

Mr BUCHHOLZ: Yes. I just want to add that I do not think there is anything wrong with the new homeowner having a high expectation—higher than their parents did. I do understand the direct link with household incomes today in comparison, but I was interested—and I think you covered it—to get your commentary on the idea that because a first home owner potentially in a hot market cannot access the median price home, which in Sydney could be $850,000 or whatever—let's call it $850,000—the whole system is broken and the only way we can fix it is to tamper with negative gearing. I do not suggest that that assists the whole Australian market to any benefit.

I have one other one, and you did touch on it with Mr Conroy. It was about what would happen to prices in the event that capital gains provisions were changed. I know that the Reserve Bank does not advocate a change to that, but Mr Conroy's question was about a period in the 1980s. I will just take that question, pick it up and bring it to today. If there were to be substantial changes to capital gains provisions today, does the bank have a position on what impact that would have, either positive or negative? If you were to take a number of investors out of the market, what would be the direct result for people in the lower quartiles of income who are renting properties? What impact would it have on their rents?

Dr Ellis : I would not want to make a quantitative assessment of that. I think it is fair to say that, if there are fewer investors, the house prices will be a bit lower. That will take some people out of rental and into owner occupation, so it will kind of all balance out. What the net effect is on rents is hard to say. As we were discussing—not in terms of capital gains but in terms of negative gearing with the earlier questions—it is just very hard to extrapolate from a temporary shift where investors had not had a chance to shift their portfolios to then make an assessment of what the effect of a permanent shift would be. I think there are tax modellers who would be better qualified to do that than we would be.

Again I think it is also important—coming back to something you said earlier—that it has never been the case that first home buyers have typically bought the median priced house. They have bought the cheaper properties in maybe the less salubrious areas—maybe the areas on the fringe. So part of the challenge is to make sure that there are enough lower priced properties in any one market. As I mentioned earlier, part of the way that that can be done is that we now have smaller properties called apartments, rather than everybody having to have the triple-fronted cream brick veneer on a quarter-acre block. So the presence of more variety in the housing stock does give people more options to choose more housing or less, and less expensive, housing.

Mr BUCHHOLZ: On that investor market, do you have any data that can split out how much of that investment sector consists of superannuation funds or mums and dads that might be utilising their super? I was of the opinion that that is what superannuation was designed to do: to build wealth to prepare for future retirement.

Dr Ellis : There are data from the Australian Taxation Office on investment in property by self-managed super funds and the use of gearing in that. That is as yet not that high. Commercial real estate is more common that residential real estate in those funds, although the residential component is growing very quickly, and we have commented about that in the past. I think the important point here is one of investor protection, and we have said this publicly in our financial system inquiry submission and in other forums. Superannuation was not really designed to be leveraged and superannuation also was not really designed to be concentrated in a single asset, often a single property. While property is an asset class that you may want to have as part of your retirement savings, having all of your retirement savings in a single unit or a single apartment or a single house is actually a pretty risky and noisy strategy, particularly if that is leveraged. From the perspective of the risk that that household is taking, that is something about which we have expressed concern in the past.

Mr BUCHHOLZ: Thank you, Dr Ellis.

Mr HUSIC: Every time someone says 'with due respect' it sounds like they are not respectful. But, with due respect, I really want to pick up on an element of your submission.

Dr Ellis : Sure.

Mr HUSIC: It goes back to this whole issue of negative gearing. You say that it has been a longstanding view of the bank that we need to have a look at the way negative gearing operates and at the way it operates in conjunction with other elements of the taxation system. But, with all due respect, you do not actually say: what next? The reason why I focus on this in particular is that every time there is speculation in the broader public that the RBA is going to make an interest rate cut, people then wonder what impact it will have on housing markets in Sydney and Melbourne in particular and on the inter-relationship between investors and owner occupiers, and even your submission today says that investors do have an impact at the margin on pricing. If you have this broader issue about whether or not you set an interest rate for the benefit of the economy, because you have to do it at a particular point in time as is being discussed right now, you look at the interplay of negative gearing, capital gains and other things in terms of the way that influences investor decisions. But there is no view from the bank to say: what next? How does this inform the decision makers? You say that it is up to us to consider this and make laws accordingly, so my question is: what do you think we should be doing? That is not really clear in your submission, and a number of times today you have said, 'I'm not a tax expert.'

Dr Ellis : And you are right; I am not a tax expert, and I think there are tax experts in Treasury. We have articulated what we think the incentives of the current tax system are doing, but we imagine there are a number of different ways to ameliorate that and, because there are many other policy objectives that the government would be trying to achieve, we would not purport to make a judgement about one of those policy objectives over another. We have our policy mandates. We have a mandate to promote financial stability. The incentive to leverage into property has an influence on the financial stability of the households and to an extent also of the lenders. So there are a couple of different ways of doing it, and I am very reluctant to talk outside my area of expertise and suggest precisely which way you could tweak things. There are a number of different ways that that could be done. We have mentioned that it is the interaction between negative gearing and capital gains, so you could do something to one or the other or both. When we talk about negative gearing, does that mean you can or you cannot, or there are particular types of expenses that can be—there are so many different ways of attacking the problem, and we simply do not feel it is our place or within our area of expertise to suggest a specific solution without having the expertise to do that analysis. Treasury has a whole tax division, and I am sure they will be much better placed to advise government on whether and how—

Mr HUSIC: I am sorry to cut you off, but it is enough of an issue to have an element of your submission focus on it. By the way, has the RBA's submission to this inquiry differed much from that submitted to the Senate inquiry that released its report earlier this year?

Dr Ellis : It did not differ that much, because I can tell you the first author of both submissions was the same, and I have not changed my mind in the interim. Just because of time issues, we did not talk as much about tax in that submission. We referred back to what we said in the 2003 Productivity Commission submission where we talked about it at much greater length. It is a longstanding view of the bank. We refer back to the 2003 submission. That is pretty much what we still think.

Mr HUSIC: So what elements of the 2003 submissions that are not in this submission would you highlight with respect to negative gearing or its interaction with other elements of the taxation system?

Dr Ellis : It is essentially the same position. The main thing that is different in the 2003 submission, working from memory, is that we presented a lot more figures about the rates of return—you would go from being negatively geared to positively geared—and we talked a lot about the fact that you could deduct non-cash expenses as well as cash expenses like interest. We did a lot more comparing and contrasting with the taxation systems of other countries.

There is a sentence in the current submission where we said Australia is at the more generous end of treatment of individual property investors but not overwhelmingly so. That is essentially a statement that New Zealand does not have capital gains tax and does allow negative gearing, so it is more generous again, whereas a country like Canada allows negative gearing but not of non-cash expenses. So different countries have different positions on the spectrum. We did more of that detailed analysis back in 2003 and nothing had changed, so we did not repeat that work for the last two submissions.

Mr HUSIC: At the risk of verballing the RBA—and you can certainly correct me if I am wrong—the RBA thinks that negative gearing needs to be reviewed. Whenever you are making interest rate cuts, everyone wonders whether or not it will have an impact on the housing market, but you say your overall mandate is quite different, that you cannot be too focused on this issue. I have read many of your papers, Dr Ellis, on the interplay of urbanisation, infrastructure and the way in which housing lots can be improved or cities can be changed, all potentially having an impact on affordability. In spite of all that, the RBA does not have any indicators of where it thinks negative gearing might be reformed to improve the operation of the system, to potentially remove distortions?

Dr Ellis : As I said earlier, it is not negative gearing in isolation. It is the interplay between negative gearing and capital gains tax, capital gains being concessionally taxed. So there are a number of different ways that you could address the particular concern from our policy mandate, but we do not make specific recommendations, not only because we feel we have a lack of expertise but because we recognise that there are other policy objectives of a tax system that may direct you to one particular solution versus another. That is why we have said quite explicitly, repeatedly, in this meeting and also in the submission, that it is not about reviewing negative gearing in isolation. It is about a holistic review of the tax system and that therefore there may be more than one way of achieving the same objective. We do not purport to advocate. If we can get the same outcome in a variety of different ways, we would not have a strong view about which of those ways is preferable.

Mr HUSIC: On a separate issue, I want to ask about investors. The RBA has publicly noted that it believes economic growth within Australia is quite different, has fundamentally shifted. We may not necessarily have the high rates of growth we once enjoyed. We have already observed wages, for example, not growing at the rate that they once did and that the drivers of wealth within the economy are changing. The banks also observed and told us as a committee that it is possible, based on history, that house prices can go south rather than continuing to go north. So the question I would ask is: where economic conditions do not support or are not as favourable to investors moving into residential property and they decide to exit, what impact would a large-scale exit of residential property investors have in Australia?

Dr Ellis : In general I would not like to speculate on a hypothetical, particularly a hypothetical that would involve people crystallising a loss that they do not have to crystallise. I think there has been a large amount of analysis and research around the world that says, 'Oh, but what if baby boomers all sell, or what if investors all sell?' The answer is that you will only get a fire sale of property if you absolutely have to because you have defaulted. As long as they are still making a positive return, some people may seek to rebalance their portfolios over time, but I think it would be very unlikely that a whole bunch of investors would suddenly all engage in a fire sale and push prices down dramatically, because in doing so they are crystallising their own loss when they do not have to. So, on the idea that there would be a self-fulfilling rush for the exits: as long as the risk profile of investors is such that they can reasonably hold that asset, they are not going to crystallise a loss unnecessarily and generate a loss unnecessarily in doing so. So that is not a hypothetical that I think necessarily transpires.

But there is a broader point that you are making here, which is: yes, trend housing price growth is likely to be lower in the future than it was in the 15 years to 2005. It is reasonable to expect that there will be, as there have been over the last decade, more periods in which prices are slightly falling. That is a different environment in terms of the risk profile that you might want to take on, in terms of the amount of leverage that you want to take on and the non-price lending conditions. That is why it is so important and why the Council of Financial Regulators has placed so much emphasis on the risk profile of the entire mortgage book. Certainly there were poor practices that we saw in, say, the pre-GFC period in 2006-07. There was an awful lot more low-doc lending back then than there has been since. There has been a welcome sort of rebalancing of the prudence of the lending standards in the Australian mortgage market. Some of the analysis that APRA has recently done and that ASIC has recently done suggests that there may be some more work to do on that. We are a long way from a US-style situation, and that is a good thing. There may still be some work to do, but I think the broader point I would take from your question is that it is really important to have prudent lending standards in the Australian mortgage market, and now so more than ever.

CHAIR: Just to finish up: the last three items on our terms of reference are the proportion of investment housing relative to owner occupied housing, the impact of current tax policy, and opportunities to reform. In your previous answer, you said that there is a mandate to provide financial stability, and prior to that there was a comment about wanting to avoid a crisis and a comment regarding distribution of wealth. There was talk about the instrument of adjusting interest rates. We know that that impacts in this market, and we are aware now, some year or more into this boom, that APRA have provided some braek in the form of a reduction in the amount that they recommend banks should lend. There is also a qualification for homebuyers to be able to repay more.

In that in nearly every event prevention is better than cure, would the bank consider—to turn to the opportunities for reform, and I am not asking for an answer now, but would you consider—an instrument during times of volatility? For instance, as interest rates go down, prices are going up very sharply—in my region, maybe 50 per cent in the last 12 months or so. There could be an instrument that could have a similar impact to what APRA's advice has been to the banks, in that, if you reduce a borrowing of 80 per cent, you are going to cool that investor down. If that instrument could very finely adjust the deductibility for an investment loan—in fact, in this case, if the RBA had the opportunity to have an instrument to reduce, in line with what APRA has done, the deductibility of an investment in real estate to 80 per cent—and if this had been done during the last 12 months, and as they meet monthly to adjust the interest rate they could simultaneously have an offset instrument to provide stability in the housing market and to avoid this incredible increase, which may lead to an incredible decrease, is that worth considering?

Dr Ellis : Just to be sure I understand: are you advocating that the Reserve Bank actually get control of some parts of the tax system?

CHAIR: An instrument that would allow a control over the deductibility of the borrowings on an investment property, so yes, I guess—

Dr Ellis : It is part of the tax system, so it is a specific element within all of the configuration. I think, as I said earlier, the tax system is a whole system that is designed in a way to meet a number of policy objectives, many of which are outside the Reserve Bank's remit. We would not be comfortable with saying that one—it is up to government to say which of those should take priority.

As a practical matter—so aside from the surprising idea that the Reserve Bank would have control over some elements of the tax system—I think the tax system is something that is very hard to change quickly. As we all learn in undergraduate economics about monetary versus fiscal policy, monetary policy has long and variable lags in its effect on the economy but has an advantage that you can change it very quickly, whereas in the tax system, where you have to legislate, where you have a political process, where people only pay tax once a year in many cases anyway, so it is over the course of a year, it is much harder to finetune.

I guess the hypothetical instrument that you are referring to does rest on the idea that you can finetune this and that you know how to finetune it. It is not entirely clear to me. We have had 25 years of experience in seeing what happens when we change the interest rate. We have had zero years of experience in seeing what happens when we change an instrument like this. That references what is happening in the prudential sphere. There is a bit of an international fashion nowadays for declaring certain things to be macro prudential tools, but there actually is not a whole lot of experience in seeing what the effect of those instruments is on the objectives that you are trying to meet, and such experience that exists is often quite mixed.

It is a policy idea. It is one that the government may wish to consider, given that it is the government's taxation powers that are at issue.

CHAIR: Thank you. As there are no further questions, I thank you both for being so generous with your time. Thank you for your attendance here today. If you have been asked—which I think you were—to provide some additional material, would you please forward it to the committee secretariat. You will be sent a copy of the transcript of your evidence, to which you can make corrections of grammar and fact. I hope I have the same chance! Thank you very much for attending.

Dr Ellis : Thank you very much for the time.

Proceedings suspended from 11:28 to 11:38