Professor Alan Ahearne:

Chairman and members of the joint committee, I thank you for inviting me here to testify today. In the invitation to appear before the committee, I was asked to discuss issues relating to early warnings, divergent and contrarian views in the context of the banking crisis in Ireland. In my opening statement, I will focus on the views that I expressed during the bubble years about house prices, the potential effect on the economy and the financial system of a reversal in property prices, and the evidence on which those views were based.

My own perspective on the Irish housing market during the bubble years was informed in large part by research on the international experience with housing booms and busts that I had done when I worked for the US Federal Reserve. In a study I wrote with several of my former colleagues, which was published on the Federal Reserve's website in September 2005, we identified no fewer than 44 episodes of house price booms and busts in 18 advanced countries since 1970. I have provided a copy of this paper to the members of the committee.

It struck me at the time that paper was published that given developments in the housing market in this country, the analysis contained in the paper could be relevant for Ireland. As I wrote in articles in The Sunday Business Postand the Sunday Independentin October 2005, shortly after I left the Federal Reserve to move back to live in Ireland, "Given the eye-popping gains in house prices in Ireland over the past decade, the foreign experience is particularly relevant". What the foreign experience analysed in the paper shows is that periods of prolonged rises followed by protracted falls are a surprisingly common feature of house prices in advanced countries. The study shows that certain financial conditions, such as low interest rates, ample liquidity, and financial deregulation, are usually present in past house price surges, though other factors such as demographics and buoyant income growth also help explain these booms.

At the time, interest rates in Ireland were at very low levels, with the European Central Bank’s main policy interest rate at 2%, having been cut from a peak of 4.75% in 2000. It was clear at the time that such low interest rates were not appropriate for Ireland’s rapidly growing economy.

A common feature of housing booms and busts is that around six to eight quarters before the peak in house prices, interest rates begin to move up. The ECB began to hike interest rates in December 2005, and within 18 months the ECB’s policy rate had doubled. Writing in 2007, I pointed out that, "The blow to affordability from rising interest rates and the knock-on effect on house prices should be obvious". The 2005 Federal Reserve study of the international experience of booms and busts shows that, after reaching a peak, real house prices subsequently fall for about five years, on average, and their previous run-up was largely reversed. Put simply, the bigger the boom, the bigger the bust.

The study found that swings in house prices can have important implications for both economic activity and financial stability. We found that in the past, major declines in house prices were often associated with economic downturns, slumps in consumer and investment spending, and at times contributed to financial distress, particularly when nominal collateral values also declined significantly.

Looking across countries, we noted that a historically high number of countries at that time were experiencing abnormally rapid rises in house prices. We warned that, "If these prices follow the same patterns as before, house prices in a large number of these countries are likely to decline in real terms at some point in the not-too-distant future". One question that arises is whether there exist indicators that can act as reliable telltale signs that housing is overvalued. The evidence suggests that comparing house prices and rents provides a useful, but by no means infallible, benchmark for valuing houses, in the same way that the ratio of stock prices to dividends is commonly used to measure valuation in the stock market. Rents are a key determinant of the value of housing and as such should not move too far out of line with prices.

House prices that are unusually high relative to rents may indicate that housing is overvalued.

Writing in October 2005, I noted: “Ireland’s price-rent ratio is currently higher than at any time for which we have reliable data, having soared since early 2002 as rents began to decline. In the first quarter of 2005, the average price paid for a house nationally was about €256,000 and the average annual rent was €8,800. The resulting price-rent ratio of 29 stood roughly 2½ times above its level in 1996.”

I concluded that this unusually high level of house prices relative to rents was mainly supported by large expected increases in house prices. The property market was pulling itself up by its own bootstraps. Property investors, for example, were not too bothered that rents were low, since they anticipated hefty capital gains on property. Once investors came to realise, however, that those rosy expectations were going to disappoint, it became clear that house prices were badly misaligned with rents and the market went into reverse.

Another question we address in the Fed paper is how house price reversals affect different sectors of the economy. We found that home buyers appear to be the most affected by fluctuations in house prices, especially if they lose their jobs in a downturn. We did note that low initial loan to value, LTV, ratios offered some protection to home owners. From that perspective, I expressed concern in 2006 and 2007 about rising loan-to-value ratios for mortgages. Data showed that one in three new home buyers in 2006 took out a 100% mortgage. Moreover, the number of first-time buyers taking out loans with little or no deposit doubled in 2006 from the year before. In an interview for Irish Property Buyerin July 2006 I said: "What we should be seeing are moves to restrict the loan to value ratios for mortgages, for instance. That would be a good start. Giving people 100% mortgages is a move in completely the wrong direction. It brought in people on the margins and in the event of a downturn they are going to be very exposed." Worryingly, the data also shows that nearly two thirds of all new home mortgages taken out in 2006 had amortisation periods of 31-35 years or longer. Such heavy borrowing rendered many households very vulnerable to a downturn.

Mortgage lenders are also affected by swings in house prices, though we found that their exposure to house prices does not, in and of itself, pose a significant risk to financial stability. We identified three factors that generally limit the prospects of credit losses on mortgage loans. First, loans are not typically made for the full value of the property, that is, loan to value ratios are usually low. Second, mortgage lenders can substantially reduce exposure by securitising a significant portion of the loans that they originate. A third factor is that nominal house prices are less volatile than commercial property prices. In the case of Ireland’s banks, these three potential mitigating factors were of limited help in containing credit losses. As I mentioned earlier, loan to value ratios were high, a significant portion of loans were not securitised but rather stayed on the banks’ books, and the banks were heavily exposed to commercial property, including speculative property development.

In the Fed paper, we also examined the recent historical experiences with banking system stress associated with declines in property prices. In particular, Japan, Sweden and Norway experienced significant financial system stress in the early 1990s, including, at least de facto, major bank insolvencies. Although declines in the value of commercial property collateral were a factor in these episodes, residential mortgage lending was not. As I put in a piece I wrote in July 2007: “The most important point, however, is that the banking crises in Scandinavia was more directly linked to drops in the value of commercial property rather than to the decline in house prices. A struggling homeowner that hands back the keys of the house causes a mild sting to a bank; a property developer that folds owing the bank a packet inflicts a terrible pain.”

Finally, we pointed out that typically the residential construction sector, home building, is very vulnerable to corrections in house prices. The evidence suggests that booms and busts in residential investment can be pronounced.

To conclude, I would like to note that, notwithstanding the patterns that we observe in the data, we did note in the paper that housing bubbles are intrinsically hard to identify, especially when they are occurring. This is because it is very difficult to differentiate between price changes coming from underlying economic fundamentals, some of which are unknown, unobservable, or unquantifiable, and those based on so-called “irrational exuberance”. That concludes my prepared remarks.