Transatlantic differences in real estate bubbles?

With the outbreak of the so called ‘sub prime’ crisis the attention of policy makers has been focused on the US housing market. It is by now common knowledge that the problem started with a bubble in US house prices. However, it is not widely appreciated that in Europe housing prices have also increased over the last decade, and, in many cases, even more than in the US. The same has happened in a number of other OECD countries and emerging markets, where rapidly increasing incomes have put pressure on house prices.

Over the last decade a curious phenomenon has thus emerged: with a few exceptions (essentially Germany and Japan) housing prices have risen almost everywhere to levels never seen before. How could such a global cycle emerge when real estate is the most local of all assets? Recent research suggests that the global housing cycle was tightly linked to the unprecedented increase in the supply of liquidity by the major central banks. Financial innovation, like giving mortgages to sub prime debtors might also have played a role, but sub prime mortgages would probably not have been supplied on the same scale if central banks had not created an environment of ample liquidity and persistently low interest rates.

The ‘sub prime’ crisis has so far affected mainly financial markets in the US and Europe. This is not surprising if one considers that on both sides of the Atlantic, prices have reached historical peaks and on both sides the upward movement had, until recently, accelerated. This is not mere coincidence. US and European house prices have always been correlated. Figure 1 shows that over the last three decades, prices in the US and Europe have tended to follow three slow-moving related boom-bust cycles. All previous peaks were followed by several years of declining housing prices (in real terms) and there is no reason to assume that this cycle will be any different.

What does the long run correlation between US and average euro area housing prices imply for the future? Since last year, prices are declining in the US. The euro area data become available only much later. Hence it is not possible to determine whether housing prices have already turned on this side of the Atlantic as well. But if the past pattern holds house prices should be starting to fall very soon in (continental) Europe as well.

How far do prices have to fall? As figure 1 shows, prices rose 20-30 % above their longer term average values both the US and Europe. The variable shown in figure 1 is the ratio of house prices to rent, a sort of ‘p/e’ ratio for the real estate sector. This ratio should definitely be stationary. The large overshooting of 2006/7 is thus likely to be followed by some undershooting. This would imply that prices will probably have to fall at least by 20 – 30 %, but possibly considerably more before they find a bottom.

Should European (and in particular euro zone) policy-makers be concerned about the behaviour of house prices? There is little evidence in the euro area of large-scale ‘sub-prime’ lending.

But, another, seldom mentioned, difference is even more important: in the US most mortgages are ‘no recourse’, which means that the lender (the bank) has no recourse to the owner of the house. If the value of the house is lower than the mortgage on it the borrower can just walk away, and simply send the keys to the bank. This is called ‘jingle mail’, and it is spreading rapidly in the US as house prices are declining almost everywhere. This ‘no recourse’ nature of US mortgages means that a fall in house prices leads to severe problems for the banking system since mortgages still make up for almost one half of all lending by US banks. By some estimates the US banking system might lose all of its capital if house prices fall by the magnitude estimated above 20-30 %.

By contrast in Europe, borrowers cannot just walk away from a mortgage since they remain liable for any difference between the value of the property and the amount of the loan. In Europe a fall in house prices might make consumers poorer and less willing to spend, but does not threaten the stability of the banking system.

Another often overlooked transatlantic difference lies in a more subtle difference in the mandate of the two central banks. The key difference is not so much the emphasis on inflation, but rather that of the financial system. The website of the Federal Reserve proclaims proudly that it “provides the nation with a safe, flexible, and stable monetary and financial system.” This totally different in the euro zone where the stability of the financial system is not even mentioned among the secondary objectives of the ECB.

Given these two differences it is clear why the Fed had little choice but to slash rates, hoping that this would help the banking system, whereas the ECB has not moved an inch. However, since rate cuts cannot stabilise house prices in the short run it is also clear to the Fed that even cutting 300 bps cannot stop the crisis from spreading. This is where the second difference comes in: the Fed had to overhaul rapidly the instruments by which if provides liquidity to the banking sector. Until mid March it only accepted government paper as collateral. Since then it is accepting a wide variety of private sector assets, even the mortgage backed securities the market shuns. By contrast, the ECB has accepted private sector collateral already for years. There was little need to change instruments in response to financial market difficulties. Moreover, in the US the Fed has extended access to its financing window to investment banks and even primary brookers, whereas the ECB did not face the same pressure to widen access to its discount window because of the universal nature of banks in Europe.

But even if there is no immediate threat to the stability of the financial system in Europe there is still reason to be concerned in Europe as well because house prices have an important impact on domestic demand, albeit with wide variations among individual countries.

Almost everywhere higher house prices have been associated with strong consumption (and vice-versa in Germany where consumption and house prices have been weak). Lower house prices throughout Europe are thus likely to be accompanied by weaker consumption demand, much like what can be expected from the US.

The impact of house price on construction activity has been more varied. In Spain and Ireland, for example, construction investment has increased to a level (18-20% of GDP) way above that of the US and not seen in any other OECD country except Japan before the bubble of 1989 burst. In these two countries, lower housing prices are likely to be associated with a sharp and prolonged drop in domestic demand which should be stronger even than what can be expected in the US. By contrast, in France and Italy, where house prices have increased almost as much as in the US, there is no evidence of a housing overhang. The negative impact of a downturn in housing prices in these two countries should thus be limited to a drop in consumption. Germany should be affected least because both house prices and construction activity have for some time been below trend.

All in all one can thus conclude that the coming downturn in housing prices will not be limited to the US, but its impact likely to be more concentrated in, and thus quicker to show up in the US banking system. European consumers and home builders will also be affected, but this will take more time to become apparent.

Figure 1. Transatlantic comovements in house prices: a long term view


5 Responses to "Transatlantic differences in real estate bubbles?"

  1. eparisi   March 27, 2008 at 10:17 am

    Very interesting post. It must be noted however that European banks are not doing much better than U.S. banks at all in terms of writedowns in dollar terms ($95bn in EU vs $110bn in U.S.) And house prices are just starting to turn in the EU unlike the U.S. where they fell 10% last year. With regard to the recourse feature and universal banking structure: Is it fair to say that higher banking resilience comes with larger spillovers into the real economy? Thanks.

  2. koteli   March 27, 2008 at 12:55 pm

    Very interesting post.Yesterday news in Spain were housing 27% down in January 2008 from a year earlier. Mortgages 25% down.Today Planner-Asprima 2008 report (realtors) says that in 2008 prices will go down 8% and new flat building down to 300000 from 600000 last year, and unemployment up 0,9%, GDP down 1,3%.And it’s just the beginning!

  3. David Jensen   March 27, 2008 at 1:39 pm

    -An interesting chart. Given that inflation has been systematically understated in the G7 and an acknowledged and prolonged gold leasing program by the same has muted gold’s ability to rise when a loose monetary policy has been in effect, it should not be surprising that asset bubbles have been created in all of the G7 by our central banks.Please see: David Jensen.

  4. Anonymous   April 5, 2008 at 2:41 am

    I do not agree that most US mortgages are non recourse. In some provences of western Canada this is the case and I have heard it is also true for purchase money mortgages in California. However, my experience is that in most other states mortgages are a personal obligation of the borrower.

  5. Flanders   June 21, 2008 at 1:03 pm

    Thank you for this article. I appreciate this a lot.