The Specter of Deflation

U.S. house prices rose 13% in the year to Q3, including an astonishing 42% leap in Nevada, 27% in California and 23% in Washington DC. Prices have risen a long way on the coasts over the last 7 years with gains of 134% in California, 103% in Massachusetts and 92% in New Jersey and 89% in New York. Inland regions have generally been more stable so the nationwide average gains since 1997 is a more moderate 65%. Nevertheless, with house price inflation accelerating, it looks as though the United States is in the early-to-middle stages of a bubble. In the U.K. and Australia more advanced bubbles are key factors in economic performance and monetary policy. The United States is likely to go the same way.

One of the causes of the bubble is that people seem to have forgotten that house prices can fall as well as rise. And the risks of a significant fall are more acute now than for over 50 years because of the low rate of inflation in consumer prices and the threat of deflation. Between the 1950s and the mid-1990s falling consumer prices, deflation, was virtually unknown anywhere. The world's attention was focused entirely on battling rising prices, inflation, which had become the number one economic problem. But by the late 1990s the battle against inflation was won and deflation had emerged in several countries in Asia including Japan.

Deflation is a new and troubling threat for all of us, brought up in an era of continuous inflation. Almost nobody alive today, even the venerable Mr. Greenspan, was an active market participant or policy-maker in the 1930s, the last time the United States suffered deflation. Yet, during the 19th century and right up to the 1930s, deflation was common, indeed even normal, while inflation was usually only seen at the height of economic booms and in wartime.

In the U.S., deflation is still only a hypothetical possibility, but in Japan it is a painful reality. Japan's stock and property bubbles deflated rapidly in the early 1990s and a series of short-lived upswings were each soon ended by a new downturn. In this weak environment, inflation gradually dropped to zero and then deflation set in, starting in 1995. As of the end of 2004 Japan's price level has fallen a cumulative 10%.

A world of very low inflation, and potentially deflation, makes the current house price bubbles more dangerous than in the past and, from an investor and homeowner point of view, means that houses are a more risky investment. After past price bubbles, house price adjustments were limited in nominal terms by the cushion of high underlying inflation. Indeed in the United States, the nationwide price index has never fallen in nominal terms. In fact, there was a 10% adjustment in real prices in the 1990s, but it was hidden by the high consumer price inflation of the time. In some regions, the real price adjustment was greater and so nominal prices fell too. For example, Californian home prices fell 10% in nominal terms in the early 1990s, with a 24% decline in real terms.

How much effect would a fall in house prices have on the economy? The bursting of the 1990s stock market bubble wiped about $5 trillion off U.S. household wealth. It would take a 33% fall in home prices to have the same impact. A decline of this magnitude cannot be ruled out if valuation ratios for housing, such as the house price-earnings ratio or the house price-rents ratio returned to past cyclical lows, but it would only be likely in the context of a serious recession and a new rise in unemployment. However, wealth effects from declining house prices are usually found to be more virulent than those from falling stock markets, so a fall of "only" 10–20% in house prices could present Mr. Greenspan, or his successor, with a similar headache to the aftermath of the stock crash.

But a housing crash would have other effects too. In past housing downturns residential investment fell sharply, by 40% in 1980–82 and by 24% in 1988–91. This is reflected in the monthly housing starts data, which typically halve during recessions. But starts only ticked down briefly during the 2001 recession and have since risen close to past peaks. Residential investment accounts for about 5% of GDP, so a severe house-building recession would be enough to cut GDP by 1–2% on its own.

How likely is a U.S. housing bust? The economy enters 2005 with considerable momentum and with interest rates still low so it seems likely that house prices will continue to rise for a while, inflating the bubble further. Good news on the economic front will support house prices while rising mortgage rates (likely as bond yields move up) will threaten them. The outcome of these opposing forces will depend partly on how much mortgage rates do in fact rise. Continued good news on consumer price inflation would keep bond yields low and make higher home prices more likely. But house prices will also depend on whether the growing signs of a bubble mentality, now evident in some regions, extend further. When a bubble reaches the euphoric phase, rising interest rates may have little effect because people are entirely focused on the prospect of quick gains.

The ideal outcome from here would be a period where house prices were broadly stable, allowing earnings and rents to catch up and valuations to moderate. A small fall in the market of 5–10% would help that process along, without causing too much hardship, though a nationwide 5–10% fall would almost certainly imply falls of 10–20% in parts of California and New England and other particularly high-priced areas. The most dangerous scenario is if house valuations are still extended when the next major shock hits the U.S. economy. Stock prices would likely be falling too, so that the economy would face a double dose of asset prices effects adding up to a much more lethal mixture than in the aftermath of the stock market bust.

A large correction of house prices at some point, 20% for example, would be a painful process for homeowners as well as investors in housing. Moreover prices would likely only recover gradually since inflation and incomes growth would likely be very low at that point. Hence it is probable that prices would not return to their peak levels for 15 years or more. This might not worry some owner-occupiers. Many will have bought before the peak of the bubble so that, while they will see some erosion of their equity and perhaps suffer some disappointment, they may not be losing much, except on paper. Moreover, since mortgage rates would likely fall, people would be able to refinance at lower rates.

However, people relying on future appreciation to help fund their retirement could be very disappointed. Moreover some people would find the value of their house falling below the outstanding on their mortgage, i.e. negative equity, because of the greater decline in nominal house prices.

For an investor in housing the scenario above would, to say the least, be a huge disappointment, because there is no capital gain for more than 15 years. Of course, provided he could find tenants and provided rents did not fall, his net rental yield should be positive so there would be some income after costs, though not much given the low level of rental yields, especially in the more bubbly areas. It is difficult to define exactly where the investor would end up, because a great deal depends on how big a loan he has and what rent he could obtain. But there is no doubt that this is what disappointed investors call "a very long-term investment," or in other words a mistake! The choice is either sell and accept the loss or wait it out, but then miss the opportunity to make money elsewhere.

A big adjustment like this is most likely when we see a sharp slowdown or recession and especially if house prices continue to rise rapidly in 2005, as seems likely unless mortgage rates rise very rapidly. The United States avoided a major recession in 2001, with the help of massive fiscal stimulus and rapid cuts in interest rates. But another downturn will come one day and, if house building and consumer spending crash too, the recession will be more severe than in 2001. In a low inflation world, housing bubbles are a much more dangerous phenomenon.

December 23, 2004