New home sales for July, announced Thursday, were down 21.6% from the previous July, confirming that the real estate bubble of 2001-05 has ended. Nevertheless, conventional wisdom remains that the housing market is due for only a modest correction, before resuming its strength – after all, the United States hasn’t seen a nationwide house price correction since the 1930s. As usual, conventional wisdom appears to be wrong; a housing slump probably lies ahead, although inflation may prevent house price declines from reaching 1930s levels.

The National Association of Realtors doesn’t agree. Its Chief Economist David Lereah forecast in June a housing market in 2007 of 6.7 million sales, down from 2005 but marginally up from 2006, with the market strengthening through the year after a cool 2006. “Most metropolitan areas have not experienced price declines in the periods of modern recordkeeping” the NAR website claims proudly – recordkeeping, of course, was notoriously un-modern in the 1930s. As with stock market analysts, the NAR’s optimism is only to be expected, but there is no equivalently well financed lobby calling for low house sales and price declines. As usual, there are thus no unbiased “experts” and the enquiring mind must delve deeper.

One of the principal supports to NAR optimism is its “housing affordability index,” which was used in 2003-4 to prove that housing affordability was at record highs, so we shouldn’t worry about a downturn, and in the last year has “proved” that housing affordability, while down from its highs, is still above its base of 100 (set in 1970) and hence no problem.

The Housing Affordability Index calculates whether the family on a median family income can afford the median house, given a 20% down payment and a limitation of 25% on the ratio between mortgage payments (principal plus interest) and (pre-tax) family income. As an index of whether real people can afford real houses this has a number of problems. It focuses on the pre-tax income, thus a rise in tax or social security payments or a change in mortgage interest deductibility has no effect on it. It ignores the difficulty of getting together the initial 20% down-payment, which is greater if house prices are high. It allows rises in house prices to be counterbalanced by declines in mortgage interest rates but on the other hand it looks at only the initial affordability of the house, and does not calculate the effect of inflation in reducing the real value of a fixed rate mortgage payment.

When you look at the HAI over a long period, say since 1980, the effect of these problems becomes clear. The HAI bottomed in 1981-82, at about 65, then rose sharply throughout the 1980s and early 1990s to a level around 130-140, at which it stayed until dropping back recently. However house prices in terms of real income have behaved quite differently to inverted-affordability; they rose in the 1980s, dropped back in the early 1990s (as incomes rose while house prices stagnated) then have risen rapidly since 1995 to record levels. Thus while the HAI is currently reassuring, suggesting that affordability is somewhat above the middle of its historic range, house prices themselves are far above average in terms of median income.

The difference between the two trends is explained by the behavior of nominal interest rates, which peaked in 1981-2, declined steadily until 1993 and more gradually thereafter, bottoming out in 2003 but not rising significantly until this year. Even though house prices were low in the early 1980s, nominal interest rates were so high that “affordability” was very low. Conversely, in 2003-05, even though house prices were at record levels and continuing to rise, nominal interest rates were so low that “affordability” remained very high.

The bottom line is as follows: (i) the real affordability of housing is much lower now and was much higher in 1981-82 than the HAI claims, (ii) affordability declines sharply, as does the HAI, if real interest rates increase and (iii) the HAI declines sharply as nominal interest rates increase, but affordability doesn’t if real interest rates remain low – homeowners just have a short term cash flow problem until inflation erodes the real cost of their mortgage payments.

The composite HAI for June was 103.7, down sharply from an average of 130 in 2003 and from 110.3 in December 2005. Its future trend depends on inflation, interest rates and house prices and itself affects house prices directly. Real interest rates, in the form of the yield on the 30 year TIPS (Treasury Inflation Proof Securities) were 2.11% Friday, below their long term average level of about 3%, particularly when you take into account that the inflation figure on which TIPS indexation is based is about 0.5%-1% too low due to the index’s spurious “hedonic pricing” adjustment. Reported consumer price inflation for the 12 months to July 2006 was 4.2%, and seems likely to increase since interest rate increases have been paused while the Federal Funds rate remains at 5.25%, a real rate of just 1%.

Let’s do the calculation. If real long term interest rates increase to their average of 3% (plus ½% for the effect of hedonic pricing) reported inflation increases modestly to 5% per annum and the risk premium of 30 year mortgage rates over 30 year Treasuries remains 1.03% as it is today, then the prime 30 year mortgage rate will move from its current 5.96% to 9.53%. This would raise the monthly payment on a $300,000 mortgage from $1,790.94 to $2,529.13 and drop the HAI from 103.7 to 73.4 if house prices and incomes remained the same.

Let’s be generous. Assume that the market takes 3 years to reach the above level and that inflation and median incomes rise at 5% per annum in those 3 years. Then in 3 years time, if house prices haven’t changed, the HAI will be 85. For housing affordability to reach its long term average of 100, still lower than at present, house prices would have to drop by 15% from their current levels in nominal terms. In other words, a house that sells for $200,000 today would sell at $170,000 in 2009.

Of course if Fed Chairman Ben Bernanke was doing his job and controlling inflation properly, the decline in house prices would be greater because money incomes (and ability to service debt) wouldn’t be rising so fast – so maybe Bernanke has a hidden agenda he’s not telling us!

A 15% decline is a nationwide average; since house prices in the Midwest and South have risen less than average, and affordability remains greater, the decline around the big cities of the Northeast and California would be greater, perhaps of the order of 30%. In those areas, a house that sells for $500,000 today would sell at $350,000 in 2009.

That would bring the house price/median income ratio back to a little above the 1990 level (but well above that of 1981-2) reversing most of the 15 year surge that took house prices up 125% in 1990-2005 while median incomes rose only 50%

A house price drop of 15% nationwide and 30% in the major bicoastal metropolitan areas would have a correspondingly severe effect on the U.S. economy. Homeowners’ equity, currently 55% of household real estate, the lowest it has ever been and compared to 70% in 1985, would decline to 47% of household real estate even if take-out re-mortgaging ceased altogether. Housing wealth would decline from its current $22 trillion to $18.7 trillion, which would be equivalent to a 25% drop in the stock market, wiping $3.3 trillion off the nation’s net worth. Unemployment in the construction sector would soar. Home mortgage defaults would soar even more, as the levels of homeowners’ equity plunged to historically unprecedented levels while payments on adjustable rate mortgages soared.

This is not just bearish fantasy; last week Toll Brothers Chairman Robert Toll described housing oversupply conditions as the worst he’d seen in the 40 years he’d been in business. However, it’s certainly not discounted by investors, as stock prices remain close to their May 2006 peaks, and the Dow Jones index remains within 5% of its 2000 high. A rise in interest rates which devastated the housing sector to this extent would also affect the stock market; the combined negative wealth effect on the U.S. economy might be double that from housing alone.

Outside the United States, it’s likely that a broadly similar effect will occur, as most countries have seen a run-up in house prices fueled by the cheap money of the last few years. In Britain, for example, central London house prices are now at a level that is completely unaffordable for anyone not in receipt of a munificent City bonus, and such bonuses themselves are likely to be hard to come by in the years to come – in central London, a drop of 50% in house prices is probably needed for equilibrium to be restored. (51% of London houses selling for more than 2 million pounds are bought by foreigners, according to the Financial Times Friday — I suspect that represents outsiders’ tendency to buy at the top.) Only in Japan, where the era of 100 year mortgage finance is nearly two decades in the past and house prices have declined by about 40% since their 1991 peak, is a further house price decline unlikely.

It’s not the 1930s; the decline in house prices during the 1930s was more than 15%, largely because that era saw deflation rather than inflation. However if you combine the effects of a stock price downturn that is rather larger than in any recession since 1972-74 with a house price downturn that is larger than any since the Great Depression the overall result will not be pretty.

-0-

(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)