Transcript of Speech and Q&A Session Remarks to the Australian Housing and Urban Research Institute (AHURI) Panel Roundtable

Ms Ellis

Thanks very much Nicole and thanks to the organisers for inviting me, it's a really great pleasure to be here. I'm a little bit daunted because of course I'm trained as a macroeconomist and I'm here talking mainly I imagine to planners, so I hope that my perspectives are useful certainly the planning perspective is useful for me, it's really enlightening.

If there's one positive outcome, one silver lining out of the global financial crisis it's that we've now got more data. Policymakers and particularly international agencies are much more focussed now on the risks from housing, than they were in the past, so they want more data and there's been a huge number of international initiatives that have been put forward including the Data Gaps Initiative of the G20. The Bank for International Settlements has a new housing price database, the IMF and the OECD are collecting price-rent and price-income ratios.

And so there's a lot richer data than there used to be, and it means that for example many of the graphs in the paper are that supporting the presentations we've seen today can be updated. There were a few that I just didn't quite feel I could replicate. But I think the important point is that updating some of the numbers for the last five years changes a lot of perspectives and not so much the presentations we've seen today but certainly a lot of the international conversations around housing really haven't caught up with the last five or six years of data which is quite disappointing. It's not a comment about the presentations we've seen today.

But it does change the perspectives that we have and so for example just one thing I wanted to show people was this is housing prices deflated by consumer prices, so it's a real or relative price, it's not the price to income ratio or the price to rent. I could have shown the price to income ratio but the version we have is a complete spaghetti graph and I thought it wouldn't be as easy to see what was going on as in this graph, which I think is really instructive because you can see it quarterly it's much more up to date than the price income ratio stuff we have. And Australia actually is not one of the boomier places over the last 10 years. Prices have risen faster than consumer prices but they've not risen materially faster than household incomes in Australia and you can see three periods there, the period around 2004, the period just in the immediate aftermath of the crisis in 2009 and then a slightly more extended period around about 2011 where prices fell relative to consumer prices and in fact nationally prices fell on absolute terms in those latter two downturns and in Sydney they fell in all three occasions on a year-ended basis. So this is not a country where we haven't experienced house price falls, we have experienced house price falls, there is risk in the system that people have already experienced and so it's simply incorrect to kind of treat the Australian history as if there had never been any kind of downturn.

And it's interesting to compare the Australian line, which is the dark blue line on the left panel, with some of the lines in the right hand panel where you just haven't seen those kind of periodic moderations to the same extent. And this is a distinction that you don't see in the international narratives and I've talked about this publicly before, but there is a real tendency for some international observers to say, ‘well let's compare the house price income ratio over a very long period, over sort of two or three or four decades’. And well yes the house price to income ratio in Australia is higher than it was on average over four decades.

And there's a reason for that, I think that it's quite clear that in the US there is a strong tendency to mean revert. There isn't a level shift in the equilibrium house price to income ratio in the US, but in Australia and a number of other countries there is, and I can't count how many times I've explained this point, but it is really important to understand that for a long time the Reserve Bank has been saying there is a level shift in the equilibrium house price to income ratio. And part of that is financial liberalisation, and so maybe you might not feel comfortable with that, but you have to remember that the starting point of mortgage finance in Australia and in particular in many other countries, was in the 70s and 80s was one of excessively constrained provision of housing finance, so certainly an unmarried woman in their 30's or 40's would not have been able to get a mortgage even if she – well I don't think you could have female heads of financial stability back then, but it just wouldn't have worked.

But there were sort of credit constraints because of the regulatory system but there were also credit constraints because of high inflation, inflation is a credit constraint and when Australia went from being a high inflation country to a low inflation country that resulted in a level shift in the price to income ratio equilibrium and the price to rent ratio equilibrium and the debt to income ratio equilibrium. And this graph which I've shown before in other public talks is really the mechanism. If you think about it neither borrowers nor lenders know for sure who will default. So there are credit constraints even in a liberalised market, and the kind of credit constraints that sort of naturally come out of that concern about default relate to serviceability and they boil down to something like a debt servicing ratio for given circumstances.

So the right ratio for one person may be different from the ratio for someone in a different set of circumstances. It won't be the same number for everyone but for a given set of circumstances it's roughly some sort of debt servicing ratio is a good way of thinking about the serviceability test.

And what this graph shows you what a particular kind of repayment to income ratio test would mean, given a 3 per cent real interest rate on mortgages, which is about right, on a 25 year loan, which is not outrageously excessive, and then what inflation is. And so you can see that as inflation falls as you go to the left-hand side of this graph, the kind of loan to income ratio you get is higher. This is not news – I see quite a few nodding heads, this is not news to most of the people in the audience. It is news to a number of international observers and the reason I have this graph on hand is I have to sort of drag it all the time.

And so those who pine for the days of loan sizes at three times income are also assuming that 30 year loans are bad (Really? not if you're 25), debt servicing ratios maybe shouldn't exceed 30 to 40 per cent no matter who you are and no matter how much disposable income you have, and presumably they also assume that inflation is somewhere between five and seven, which in Australia it is not and has not been for nearly 20 years; actually more than 20 years now.

So what does all this mean? What is this change in the loan size that pops out of this credit constraint, what does it mean? Well it has important macro implications. The housing wealth or overall net worth, which is the blue line in the top left panel, well, that goes up relative to incomes, I've shown a real net worth relative to the CPI because that's kind of easier but house price to income ratio goes up for a period and then it stabilises. You can certainly see it in the orange line in the top right panel: debt to income ratio went up quite significantly and then it stabilised and in Australia it stabilised, the turning point is actually about 2006, it was before the crisis – again, not well understood in many quarters.

Interest payment to income ratio, that did go up but when you have this effect you of course once inflation is lower the people who have already borrowed are not inflating the debt away as quickly. So the kind of the repayment to income doesn't decline as quickly as it did when inflation was high because nominal wages aren't growing as quickly.

So you do expect that to come up a bit, but the important point of course at the moment in Australia, mortgage repayments are sort of lower than the average of the past 10 or 15 years. So again are you thinking about what it means to have affordability or unaffordability, the affordability of a current mortgage, as we can see in these macro numbers from the national accounts, is actually not unusually high at present.

And the final point is of course if you're going through a period where you're increasing your debt to income ratio, well that probably means that growth in your spending – in your consumption – is outpacing the growth in your income. That's just kind of a natural counterpart to the increase in your debt. And then at some point you stop doing that and your consumption growth starts tracking something more like your income growth. And the difference between consumption and income in the national accounts is of course the household saving ratio. And it's calculated as residual so it bounces around a little bit, but the green line in the bottom right hand panel shows you through the period where we were increasing the debt to income ratio in Australia, that household saving ratio was coming down, it hit zero. It was never going to stay there, that was not a sustainable situation. But look how rapidly it came up, now there's a few bumps because of some fiscal transfers during the crisis, but again if you look through those you can see again the turning point, the point at which household savings started rising rapidly relative to income, was before the crisis, about 2005.

So the narrative we've had for a very long time has been: we expected house prices to rise relative to income and then hopefully stop. And we nearly kind of got into trouble in 2002/03 when we did have quite a bit of speculative and investor housing and fortunately that particular phase of our housing market ended more or less with a fizzle rather than a bust. But that experience has really coloured our thinking about what the risks around housing were, and I think I showed a graph a couple of speeches ago of the share of pages in countries' financial stability reviews that are devoted to households and housing, and everyone else has kind of caught up with us. We've always devoted a lot of time in our financial stability reviews to talking about households and housing and a lot of other countries have started putting similar weight in their financial stability publications as well.

So it's important to remember that equilibriums can shift, but the risks are really potentially quite large.

One of the really helpful things about tonight's presentations and the paper that was behind them is just how important it is to have those cross country comparisons. In Australia we should not assume that we're special, we're special, we're different. But sometimes the reason particular other countries have had different experience is that they're special, and not in a good way.

And so if you think about the price crashes and the housing foreclosure rates, I mean what went wrong in the US mortgage market is sort of my Mastermind special topic, so don't ever bring it up at a dinner party, you'll never hear the end of it. And it's clear that it was not just that there was overproduction and a huge overhang, which I'm pleased to say there were people who knew that internationally, I mean the Bank for International Settlements Annual Report in 2008 made it quite clear: look, the US, maybe Ireland, less so Spain, not at all the UK or Australia. It wasn't enough just to look at prices you had to look at that over-build, you had to look at the increment to quantities. And so some international agencies at least were getting it right even as early as then.

And I agree, all of the speakers would agree, that you can't just take the experience of those three countries and translate it direct to Australia and you know Christine's section of the paper said don't translate UK experience to Australia just word for word because it is different and understanding why it's different is important. You can't just translate it, but you can learn from it none the less.

I couldn't help wondering: where's Canada in all this, though, because I feel like Canada's a much better comparator for Australia. I said geographically I mean you've got almost all of the Canadian population is within 100 miles of the border, almost all of the Australian population is within 100 miles of the coast, so topographically they're a cold straight line and we're a warm circle with some bumps. But there is a lot of similarities, the very similar banking structure, the supervisor is quite conservative relative to the global median, the urban hierarchy is not dissimilar, the cities are a bit denser but not drastically so, the cities here are not dense at all compared to many other countries. So there's a lot more touch points of similarity and it would be really interesting to sort of see this work extended and reach into what Canada's doing. The two main differences I can see are: one, they have fixed rate mortgages, not variable rate mortgages, and I agree with Christine that that is enormously important for the resilience of the household sector in a downturn.

And the other thing is that the government takes the tail risk of the mortgage book because it's a government funded mortgage insurer, and so that's extremely important as well.

One of the key messages I drew from the presentations and the supporting paper was that housing crashes do not improve affordability. Housing prices are lower, yes, but capacity to pay is also lower, and again that's something I wish a lot of people would take away from this. I mean, again, this audience is nodding and saying yes of course, but again it's something that's not well understood outside these sort of professions.

And we need to focus more on where real affordability issues are which is – and certainly this is the US experience – low income households in private rental, probably people who because of the reasons of the variability of their income would actually be worse off if they were in owner occupation and that was the lesson of trying to put people into subprime mortgages, they were worse off than if they stayed in private rental. But the rent is also really expensive and also in the UK it's even more expensive to rent than own, and I think that's the affordability issue we should be giving more focus to than we currently are doing.

I'm really glad to see a focus on the risk of overbuild. I mean in Australia all of the usual narrative is about housing shortages and all the problems of high prices would be solved if we could just build more. And of course the usual assumption is that means if we could just build more detached houses on the fringe because suburban living is the ideal type of housing for everybody, apparently, that's the assumption and, again, you get it, right?

But what the US, and Ireland and Spain have shown is that the really painful outcomes happen when there's an overhang of excess supply, and again this is something I've been banging on about for a number of years. A number of international agencies have been saying as well, and I've certainly spoken about it at length on numerous occasions. But I should make it quite clear: Australia is a long way from oversupply nationally, that's absolutely right, supply is increasing, and that's an expected result of the current stance of monetary policy. It is needed to house a growing population and it's intended as part of the handover from the end of the mining investment boom, in that other sectors need to and now have the scope to expand more quickly than they did when that boom was in full swing.

So building approvals at the moment are running at an annual rate of about 200,000 so we're building as many houses as you would like to and we have a much smaller population, I mean that's extraordinary. It's quite a bit higher than in the past decade. And detached housing is still a bit over half of that but there has been a real switch to medium density and high density becoming more important in recent times.

And a lot of the reason for that is of course we've got very strong population growth, and population growth matters for the amount of housing supply you need it's not just the level it's the growth rate, but the composition of that population and its location matters enormously. There is an assumption here that we need more housing to accommodate the people that are arriving, but it's little appreciated just how concentrated that population growth has become in students and former students. So this is net immigration by visa type and this is a graph that maybe not many of you would be as familiar with, but it's quite extraordinary firstly what a big swing up and down there was in student numbers and this allows for the fact – it's net – so some of them go home but an awful lot of student people on student visas do get permanent residency after they complete their studies. I think anyone in an Australian university could not fail to have noticed that there are a lot of foreign students. Education is one of our big – it's coal, iron ore and university education is more or less the ordering of exports by volume in this country. And you can see that in the last two years its picked up again really rapidly and again part of that's the flow of new students and part of it's more of them are inclined to stay.

So where do students want to live? And where are former students like recent graduates likely to want to live? Well, it's not big family homes on the fringe, it's apartments in the inner areas near the universities. It's so important to understand that, and so of course a lot of our supply and again if you're wandering around Sydney you cannot fail to notice that we are actually getting a fair bit of construction, but it's in that sort of apartments, close to universities, locations. There's a consequence to that though. This is where all the populations coming in, is to inner areas and so the premium to being close in has increased: so this is the ratio of the sort of medium price in an inner ring suburb to the medium price of outer ring suburb, and this is detached houses – apartments are not even in here. So even in detached houses, you can see that in the last sort of half a dozen years or so there's always a premium to being close in but it's gotten larger recently. So I think that's really important.

I think I talked about this a bit in a speech in May and in fact I just used the same graph. I think there's a lot for financial stability policymakers to learn and to consider about urban geography. I think that we tend to look at national aggregates and one of the things I've learned over the last decade is you've got to look at the local area stuff as well. I think the planning profession can certainly contribute a lot to that thinking, and the whole thing of Dublin wasn't building enough and there was all this, all this building going on in sort of random villages that you [Michelle Norris] can pronounce. So Dublin's very much the primate city, it's the economic centre and yet it wasn't building enough housing and yet there was way too much housing elsewhere. It's all very well to build more housing to accommodate the population but it has to be broadly comparable with where the people want to live. Building houses people don't want to live is counterproductive.

Actually according to Kate Barker, the former Deputy Governor of the Bank of England [sic – former member of the Monetary Policy Committee], who did a review of housing supply in the UK, about a decade ago, has recently put out a new book which I've been reading and I couldn't help noticing that part of the issue in UK planning policy is building new towns. Now since when have we built new towns in Australia? I mean not since Canberra, we haven't actually gone ‘well yeah here is a place for a city’. We don't do that anymore and we certainly don't say look here's a town that's currently got 30,000, 50,000 people but really we could put half a million here and maybe we should have a long term plan to have more. We have this sort of decentralised policies that have been tried but there's no sense in which we need more cities that can create spaces for entrepreneurship, good jobs for people who are university educated. I mean, there's plenty of good universities outside the major centres but they don't keep their population in the way that university towns in the US do or university towns in the UK do. It's really quite extraordinary.

One of the things about supply is, again, you can't always compare with other countries. In Australia you don't tend to have a lot of spec building, people have to wait to be asked to build a house. But I can't help thinking that there's a bit of a trade-off: yes spec builders create an overhang of excess supply because they're willing to add new supply into the system whenever prices are above cost. It's a levels thing.

But if instead builders are waiting for people to ask them to build something and if much of the demand is coming from buy-to-let small investors, then what we have is a situation where supply might only increase if we have expected price growth. And if expected price growth tends to come from recent price growth well then that isn't a great outcome either. And so I think we need to be cognisant of all of the institutional differences. We need to learn from those differences and from the points of commonality, so I'm really pleased to see this kind of work being done and I can assure you that it does get attention in the policymaking world.

Question & Answer Session


Yeah Frank Stillwell from the Political Economy Department here at Sydney University, my Question relates to capital accumulation and taxation. We've heard a lot about the failure of housing markets; it seems to me that the root problem is that people are participating in markets, not purely to access housing but for the purposes of capital accumulation. Is therefore the only solution in the long term to drive out that kind of housing demand through taxation, whether it's inheritance taxation, land taxation, capital gains taxation that makes housing markets purely a means of transacting housing services rather than getting rich.

Ms Ellis

I'm not sure that as central banker I really want to opine on tax policy. Although the Bank has talked about this at length in its 2003 submission to the Productivity Commission's inquiry into first home owner ownership affordability and my boss Malcolm Edey talked about this when we were up before the Senate Committee a couple of weeks ago and if my answer departs from his in any way it's simply because I didn't memorise it. But I think one important take away I would mention is that, of course, there are people for whom owner occupation is not currently in their best interests because their income won't support it or because their income is too variable that it exposes them to too much risk, particularly given that there is a certain tendency for that to be debt financed. So some people need to rent for at least some part of their lives, and someone else needs to own those homes, but I think one of the interesting points about, certainly, the Australian housing market, I mean we don't actually call it buy to let here – I always find I have to translate when I go to the UK – here we call it investor housing. Maybe this is not the right audience but is anyone willing to admit to owning an investment property here?


Out in the lobby we'll do a census there'll be a tax, drinking tax for all your investors out there.

Ms Ellis

But this is the thing - - -


We'll charge in beers Nicole.

Ms Ellis

- - - yeah but I think the important thing is people you know when you meet someone and say oh I own an investment property, they don't say I'm a landlord, or I have a small side business providing housing services to customers. They certainly would never say what my landlord in Basel said when I was working at the Bank for International Settlements, which was, ‘I see my tenants as my clients’. Can you imagine anyone here in Australia who owns an investment property saying that? I think you've touched on an important point and I think I wouldn't go beyond what the Bank said about taxation policy because it isn't our remit to go any further than that. But I think there's an important grain of truth there.


Okay, Michael Janda from the ABC and unsurprisingly my Question is for Dr Ellis. Now you mentioned that lower inflation and financial deregulation resulted in a step-up in Australian house prices and allows for increased housing debt to pay for housing. But in Australia that increased debt relies on people overseas being willing to lend our banks the money to lend people for that housing, at least at the margins. So my Question would be what happens if people overseas decide to stop lending Australian banks money for Australians to invest more in housing? And also I noticed that the RBA has a strong tendency to use house price figures from 2003 onwards, wouldn't you say that the step up in housing prices should have been factored in in the 1990s when the deregulation took place? Why did we have that small spike in the early 2000s?

Ms Ellis

Well I think the first point is its certainly true that a counterpart to the transition, the one off transition to a higher debt–income ratio was that that meant that investment was stronger than savings and so that was a current account balance that was being funded by something. But of course again this is where updating the graphs is very informative, there is the sort of narrative that Australian banks rely on offshore funding. Australian banks might wish to diversify their funding offshore and I think that's entirely appropriate, but if you look at the net funding, net wholesale funding by Australian banks overseas over the last several years it's basically been zero. They're issuing debt but they're also paying off debt and so again the increased use of wholesale funding we saw prior to that turning point was all the counterpart, the low household savings ratio, the current account deficit, the strong growth in debt – they were all part of a parcel, and that is how you got that transition. But again update the graph and you'll see something different. Look I admit that that particular graph we used it's one that I pilfered from a previous Financial Stability Review, but I think that's an important point because we don't actually regard the late 90s up to about 2002, we know that was the period of the transition and so the important point is what's happened since then. Because we don't think that the 70s and the 80s is the appropriate comparative for things like house price to income ratios. So, look, I admit the particular graph was one that we've used before and so I didn't change the start point it starts getting messy on a two panel graph. I could have brought the spaghetti graph of house price to income ratios, that goes back further. But the important narrative and the number of times we've said this publicly, I mean it's in the Financial System Inquiry initial submission, it's in about five speeches of mine, it's in several speeches by other people more senior than me, numerous research papers that we've been writing and in fact that whole idea that lower inflation meant that you could then service a bigger mortgage. Glenn Stevens gave a speech on that back in 1998 [sic – 1997], we always knew that this was going to be a one-off transition and we've been talking about it for over 15 years, so the fact that housing prices have basically tracked income since 2003 in Australia should finally be starting to sink in. I mean I again if I'm getting this quote wrong it's only because I haven't memorised it, but the Governor did talk about this in 2012 and it's kind of like, ‘well, after a decade or so, perhaps people might – including some of the most you know dire financial circumstances – maybe there's some truth in that analysis’.