RBA Annual Conference – 2012 Opening Remarks Glenn Stevens

Introduction

Welcome to Sydney. It is a great pleasure to see you here and to open this Conference. For the Reserve Bank of Australia this is the 24th annual conference; however, it is the first one that we have co-hosted with the Bank for International Settlements (BIS). It is a great pleasure to do that.

This year the theme is ‘Property Markets and Financial Stability’. Recent events have highlighted the importance of that linkage and how great the dangers to economic stability can be if we do not appreciate the linkage between property markets and stability. In earlier conferences we have touched on these themes but we have not devoted a full conference. So the topic is welcome – it is also timely.

It has, of course, increasingly been a theme of research by the BIS. The BIS Office in Hong Kong started to focus on the theme in 2010 as a result of a request by the BIS Asian Consultative Council. A year ago there was a workshop hosted by the Monetary Authority of Singapore to review the progress being made on the contributions to this research program. A number of projects that were commissioned as a part of this program have come to fruition as research papers that will be presented over the next two days.

There is little doubt that property price swings and financial stability are intimately connected, our own experience in Australia certainly confirms that. To start things off I would like to talk very briefly about that and about some of the simple, and occasionally painful, lessons that we have drawn along the way.

Australian Experience

The history in Australia dates back to the 1890s when there was a very pronounced property price cycle that was very important for the city of Melbourne (Fisher and Kent 1999). In a way, that episode is of some interest for this Conference because some of the drivers of that cycle, initially at least, were population growth and urbanisation which are important factors in parts of Asia today. But the episode in Melbourne was also fuelled by foreign capital inflow, speculation and a decline in lending standards. The term ‘Marvellous Melbourne’ was coined to describe the apparently easy prosperity of that time. I have read that Melbourne claimed to be the biggest city in the British Empire after London in those days.

In Australia the 1890s depression, particularly acute in Melbourne, was much worse than the 1930s on several metrics, particularly financial ones. Something like half of the trading banks suspended payment and a large number of non-bank financial institutions failed. So it was a very severe experience.

Our more recent experience with cycles in property markets occurred in the 1980s, the 1990s and the 2000s against the backdrop of broad-based deregulation of the financial sector. This deregulation removed the strictures of very tightly constrained regulation on banks and the credit rationing to the community that resulted from those regulations (Battellino and McMillan 1989). Deregulation saw banks freed from those shackles and looking to compete. Liberalisation of the capital account increased access to offshore funding which allowed credit growth to be faster than deposit growth. This persisted for many years, though it is changing now. A more liberal policy towards granting banking licences also saw foreign banks enter the market, heightening competition, especially in the area of corporate lending.

As you might expect, with this deregulation occurring after a long period of financial repression, risk management practices in financial institutions were not necessarily all they could have been. The initial result was a boom in commercial property – finance and prices – in the late 1980s that unwound quite painfully in the early 1990s when a recession hit. This episode exposed the earlier poor lending, with parts of the banking system coming under severe stress, and once again, interestingly enough, the city of Melbourne was quite prominent in that dynamic, though the problems were not confined to Melbourne.

So the initial experience of liberalised financial markets in my lifetime in this country was initially painful for banks and for the economy.

Subsequently, in the long recovery, banks were looking for other forms of lending which were safer. They looked around and saw an opportunity in households which, to that point, had been relatively constrained in their lending choices and were probably, by international standards, rather under-leveraged. For the ensuing 15 years or so, that is where the growth was. The profitability and stability of that business saw new entrants to the market and heightened competition, and banks' margins were eroded. The net result was that households, which had been credit constrained for a couple of generations, found they could borrow more, and that is exactly what they did. The ratios of debt to income and dwelling prices to income spent about 20 years rising to new, higher levels.

We are still debating whether that is a new, stable and sustainable higher equilibrium. I think it is plausible that moving from a constrained credit equilibrium to an unconstrained one is, in fact, a one-time move to a higher-equilibrium level. But it is difficult to know how high that equilibrium could be. One of the difficulties in this debate is that of getting accurate measures of dwelling prices and incomes and all the other fundamental factors. It is particularly hard to do this when we want to do international comparisons and that is a thing I will return to later.

However, to begin it is worth reviewing what conclusions we have drawn about property values, policy and financial stability from these experiences, beginning a century ago, but also in more recent memory.

Lessons Learned

The underlying monetary policy framework

It is customary to say that it is crucial to get the underlying monetary framework right. Indeed maintaining overall macroeconomic stability is the most important thing that monetary policy can do. Macroeconomic instability, and certainly lack of growth, drives many of the proximate causes of borrower distress, which in turn are very damaging for macroeconomic stability. There is no doubt that the RBA's capacity to achieve macroeconomic stability has been enhanced greatly by the implementation of the flexible inflation-targeting regime that we have had for nearly 20 years now. This provides a framework for clearly communicating the goals of monetary policy, building credibility by demonstrating that our actions are consistent with these goals, and ultimately providing a nominal anchor for the economy.

But I would also say that establishing a credible and enduring monetary policy framework around controlling consumer price inflation, though critical, is not enough. Financial stability and monetary policy are related – the relationship is not simple, it is complex, but you certainly cannot divorce the two. We came to this conclusion based on our own experience; however, one only needs to look around the world to see that lesson has been writ large in many places.

Asset prices

The second lesson is that asset prices, and property prices in particular, do matter. And they matter more than just for the information they contain about future economic activity and prices. Everybody agrees that asset prices are important and should evoke some response to the extent that they convey information about future output and consumer price inflation. But that is not saying anything more than we should build the best forecasting model. Asset prices actually matter more than that. They matter in ways, and over time horizons, that are not well captured by the standard Phillips curve framework that I think still underlies much of the standard inflation-targeting framework in operation around the world. They matter because property holdings tend to be leveraged. Big swings in asset values where the holdings are not leveraged will not, I conjecture, matter all that much. It is property that is being used as collateral for significant lending by financial institutions that makes property prices so important. It is actually the leverage that matters.

Now there has been a long debate about whether monetary policy should respond to asset prices – the so-called lean-versus-clean debate. A lot of people taking part in this Conference have contributed to this debate, which has been going on at least since the Japanese bubble economy ended, and that is about 20 years ago. Should monetary policy lean into the asset price upswing? The case to do that is that the growth cost might be small and the benefits of avoiding future pronounced instability might be great. The argument against is that we just do not know enough about asset price and credit dynamics to do this with any precision or with the hope of much success, at least without risking serious damage to other parts of the economy.

I would have thought that by this point we have to conclude that simply expecting to clean up after the credit boom is not sufficient anymore; the mess might be so large that monetary policy ends up not being able to do the job when the time comes. Moreover, if the monetary policy clean-up after the asset price bust involves interest rates low enough to prompt some other sector of the economy to leverage up in order to spur the growth, then the clean-up itself might leave its own toxic consequences.

The debate has moved, it seems to me, some way towards doing a bit more leaning. Monetary policy cannot surely ignore any incentive it creates for risk-taking behaviour and leverage. The risk-taking channel, as some now call it, is very important. But, at the same time, monetary policy cannot succeed in managing cycles, asset values and leverage by itself.

Prudential supervision

This is where we would draw a third lesson from our own observation and that of other countries in recent years. Policymakers have discovered – or maybe I should say rediscovered – the importance of prudential supervision.

Currently, there is a big debate about the appropriate structure of supervisory and regulatory arrangements. In this country, following the early 1990s' experience and following a financial system inquiry in the mid 1990s, we decided to put all the prudential supervision of banking, insurance and pension funds into one institution. That was thought to be the structure most consistent with achieving the appropriate policy objectives.

That decision paralleled one taken in the United Kingdom at about the same time. Ironically, the United Kingdom now seems to be moving back the other way. In the United States and Europe as well, there is extensive discussion of redesigning the architecture of supervisory and stability arrangements. All of this is important: getting all of the relationships right between the various entities; getting their charters properly clarified and aligned; and providing the communications between them. There is also the Basel III work and the burgeoning agenda of the Financial Stability Board, and beyond that, the G-20. This is a response to the inevitable post-crisis demand that the events of the last five years never happen again. All of this work is good, but we need to make sure that we do not neglect a few very simple eternal verities.

One of those is that serious supervision is critical. The most elegantly crafted rules will not make much difference if the capacity and will to enforce them is not in place. There is probably too much tendency to fall for the easy line that if only we can craft better regulations and bring the bankers under control then all will be well. Actually, it is the application of the rules and the framework that matter. After all everybody was using some version of Basel II or Basel I. This was more or less the same set of regulatory arrangements pre-crisis. But not everybody had a collapse of their banking system. In fact, I would say probably more countries did not have a banking crisis than did have one. The regulatory framework certainly is important, but the way it is applied matters. The quality of supervision may actually matter more than the exact detail of the black-letter regulations.

Those are a few observations that I would draw from experience. I have one final remark about property prices in particular and that is that measurement matters and measurement is difficult.

Measurement

Over the years looking at this set of issues in Australia we have found that some of the apparently easily available aggregate data on dwelling prices were rather less useful than you might hope. For one thing they are affected by compositional shifts in the nature of transactions. Having some transactions in very high-priced locations this month compared with very low-priced transactions last month will have a significant impact on aggregate measures of prices. Measures of prices can also lag what is actually happening in the market considerably due to the way they are compiled. These and other issues prompted the RBA to invest quite a lot of effort in trying to improve measures of dwelling prices. We worked with private sector data providers to come up with measures of dwelling price movements that abstract, as far as you can, from these problems by using matched samples or hedonic methods. As a result of these efforts, Australia now has better measures of dwelling prices than before. Hopefully this leads – although this is less clear – to a more informed debate. Commercial property prices, of course, are even more difficult to measure, being highly diverse and turning over more rarely so there tends to be much more reliance on subjective valuations in that sector.

One particular area where measurement matters is when you do international comparisons. The RBA has tried to look at ratios of dwelling prices to income across countries and compare the levels rather than just the movements from some base point (Stevens 2012). It turns out this is very hard to do; getting comparable data on dwelling price levels and for that matter disposable income levels is very difficult. It also highlights that such comparisons need to be done with due caution.

Although measurement matters, unfortunately we cannot always wait for the perfect dataset before trying to factor these important dynamics into our practical policymaking decisions. The assumption that if only we could get the perfect data then everything would then follow easily, I think is another delusion. We do have to try to get better data; however, in the interim we have no choice but to use whatever imperfect information we do have.

In conclusion let me say I am very pleased to see a conference devoted to these issues underway. It is fitting and timely that this occurs here in Sydney where we have been giving consideration to these issues for many years. We certainly have learned that getting our underlying policy frameworks correct is critical but we have also learned that it is not enough. The ability to manage financial stability risks is inextricably linked to the strength of the prudential supervisory regime and it requires all the agencies involved in that to work together effectively. We also have to work on getting better data, but after all that we remain acutely conscious that there is much we do not know.

I am pleased then to see that the papers that will be presented here today and tomorrow demonstrate the sort of careful analysis that this question of the interaction between property prices and financial system stability demands. As a result I am sure the discussions you have over the next couple of days will be productive and I wish you well in your deliberations.

References

Battellino R and N McMillan (1989), ‘Changes in the Behaviour of Banks and Their Implications for Financial Aggregates’, RBA Research Discussion Paper No 8904.

Fisher C and C Kent (1999), ‘Two Depressions, One Banking Collapse’, RBA Research Discussion Paper No 1999-06.

Stevens G (2012), ‘The Lucky Country’, RBA Bulletin, September, pp 75–83.