Residential property price indices

Welcoming remarks by Mr Stephen G Cecchetti, Economic Adviser and Head of Monetary and Economic Department of the BIS, at the Eurostat-IAOS-IFC conference on "Residential property price indices", Basel, 11 November 2009.


In his remarks, Stephen Cecchetti summarises why it is that he believes property prices are so important. He then goes on to outline some of the challenges in constructing accurate and useful indices for property prices.

Full text

Good morning, ladies and gentlemen. It is my great pleasure to welcome you to Basel and to the Bank for International Settlements for this conference on "Residential property price indices". We are very pleased to be hosting this event jointly with Eurostat, the International Association for Official Statistics and the Irving Fisher Committee on Central Bank Statistics.

Today is Armistice Day or Remembrance Day, the 91st anniversary of the end of the First World War. As many of you surely know, the aftermath of the Great War provided the impetus for the founding of the BIS. However - our mission and role as an international financial institution has fortunately evolved substantially since then and we are now engaged in a wide variety of cooperative activities. This conference illustrates one of the ways we have broadened our scope to include not only central banks, but also financial authorities, academia and the private sector.

But that particular history is not what brings us here today. Instead, the purpose of the conference is to address various aspects of residential property price indices - both theoretical and practical issues in measurement, as well as user requirements. This has always been an important topic, as property prices are an essential input into the proper construction of measures of GDP and inflation. But, as the crisis makes clear, if anything we have underestimated the importance of collecting property price data. In my remarks this morning, taking the view of a macroeconomist, I will summarise why it is that I believe property prices are so important. Then I will go on to outline some of the challenges we must face - or rather, you must face - in constructing accurate and useful indices for property prices.

Property prices - residential and commercial - have a variety of important uses. For national income accountants, they are used in calculating both investment and consumption. The former is about building costs, while the latter is about service flows. For consumer price index construction, residential property prices are an input into the computation of the price of housing services for owner occupiers. For bankers, as well as their regulators and supervisors, these property prices are important in helping them to estimate the value of collateral backing loans; similarly for people trying to estimate the value of securities backed by pools of loans collateralised by property. I'm sure many of you would be able to add a number of users to this list.

The crisis has clearly added one more: monetary and fiscal policymakers. The BIS is a central banking organisation, and my own background is as both an academic and a central banker. So, let me focus on what I believe to be the essential uses of property prices in the formulation of monetary policy. Over the last two years we have learned a very expensive lesson: policymakers ignore property price booms at their peril. The problem is a multifaceted one. First, property acts as collateral for loans. It is a leveraged investment. When prices rise quickly, there is a windfall to owners that allows them to borrow. So, as we have seen, property price increases tend to raise leverage - especially in the household sector.

It is worth pausing to think for a moment about this point. Households often have difficulty borrowing in order to smooth their consumption in the face of negative shocks to their income. They are, in the language of economic theory, liquidity constrained. An increase in the value of residential real estate helps to relieve that constraint, for at least some households. This should be a good thing, and under normal circumstances it is. In fact, the availability of a wide variety of financial products like home equity loans and reverse mortgages surely increases welfare by allowing for a more efficient intertemporal allocation of consumption. But when there are bubbles, when prices rise dramatically above levels justified by long-run fundamentals, then households are prone to miscalculation. Believing the transitory bubble to be a permanent increase in wealth, individuals will seek to shift consumption from the future into the present. That is, they will borrow, increasing leverage. When the bubble bursts, both the households and the investors who lent to them are in trouble.

Household overborrowing is only one of the problems created by property price booms. They also distort the allocation of capital by over-rewarding construction investment. The result is not just too many houses and other buildings, but too many construction workers and too much construction equipment as well.

And then there is the impact on fiscal authorities. Property booms are a source of increased tax revenue: not only are valuation-based tax receipts higher, but so are transaction taxes. A boon to politicians, these higher revenues lead to some combination of tax cuts and expenditure increases. But, when the bubble bursts, these receipts disappear leaving fiscal deficits that can be extremely difficult to manage.

Finally, there is the impact on the financial sector. Property price booms, because the assets are leveraged, put lenders at risk. Impairment of these assets, then, means impairment of financial intermediaries' balance sheets. Moreover, in recent years, both residential and commercial mortgages have been the raw inputs into various securitisation schemes that chop their payments streams and risks into a wide variety of pieces.

This ability to separate finance into its most fundamental pieces - the financial analogue to the particle physicist's subatomic particles - has had profound implications for the way in which risk is bought and sold. The idea was, and many of us believed, that atomistic finance would ensure risk went to those most willing and able to bear it. This would mean added efficiency for the economy as a whole, allowing individuals to take on risks that they would not have taken on before, but that improve social welfare overall.

On the plus side, I am convinced that the development of the financial system is an important source of the Great Moderation - the reduced volatility of real growth in the developed world over the past 20 years. Faced with income volatility, individuals today can use the financial system to ensure that their consumption remains smooth. Amazingly, this adds up to macroeconomic stability.

But with the ability to sell risk cheaply and easily came the ability to accumulate risk in almost arbitrarily large amounts. Combined with compensation schemes in which money managers share the gains but not the losses of their investment strategies, this creates incentives to take on huge amounts of risk. Without risk, there is no reward; and without big risks there are never big rewards.

The result is that small numbers of individuals have the potential to jeopardise the stability of the entire financial system. They do this not because they can fail themselves - the right to succeed in the capitalist system is the right to fail - but because of the knock-on effects they will have when they fail. It is the interconnectedness of the system that is the biggest challenge. And the more complex the system becomes, the bigger this risk becomes.

Returning to property, as we have seen, even residential and commercial mortgages that are securitised remain in large measure on the balance sheets of financial intermediaries. And, because of a lack of history (or lack of guile), their riskiness was grossly mispriced. The immediate implication is that in the 21st-century financial system, the connection between property price dynamics and financial stability is stronger than ever.

This brings me to what I believe to be one of the most important lessons of the crisis: bursting bubbles, especially in the prices of leveraged assets, are extremely damaging.

The debate over what to do about this is lengthy and complex. There will be changes to financial regulation and supervision. There will be changes to the nature of financial instruments and markets. And there will be changes to the way in which stabilisation policy is conducted. And everywhere, property prices - by which I really mean property price statistics - will play a central role.

This brings me to the last item I will discuss today: the challenges associated with the construction of property price indices. In preparing these remarks, I looked around a bit and discovered some of the complexities of your work. You have to face the fact that every house and commercial building is unique. While many differences can be observed, many cannot. One response to this is to develop repeat sales indices. If you think about this for a while, you realise that this only solves the problem up to a point. For example, between sales, there can be an unobservable improvement or deterioration in quality. And in many countries and regions houses and buildings change owners very infrequently, so the problem can be quite large. The technical problem is that this creates heteroskedasticity in the error in measurement of the price change that depends on the time between sales.

During bubbles, as we have seen, there is the added problem that the composition of both the stock of property and the properties that tend to sell can change. For residential real estate, larger, more expensive homes tend be involved in more transactions. This can skew the information available from transactions data.

But you all know all of this. And you know much more. You know about the difficulty of using hedonics to value characteristics of different buildings and locations. You know about the problem of sampling. You know about the seasonality created by the fact that sales tend to occur at specific times of the year. You know about the difficulty of handling sales of new homes. And on and on.

The message that I want to leave you with is that even though the task of accurately measuring property price appreciation is difficult - at times it may seem impossible - it is extremely important. In fact, the crisis means it is more important than ever.

I therefore wish all of you a productive and rewarding two days of discussions. I am sure that the contribution of this conference to the development of an International Handbook on Residential Property Price Indices will be a significant one.