Daily Speculations The Web Site of Victor Niederhoffer and Laurel Kenner



Robert Z. Aliber Cont'd

Write to us at: (address is not clickable)


One reason for the surge in the inventory of unsold homes is that would-be buyers are "on strike" because they anticipate that home prices will decline. The November 5th edition of the Valley News, the daily paper for Hanover/Lebanon/White River Junction, reported that a family that has just moved in the area is renting because they anticipate that the asking prices for homes will decline further.

The National Association of Realtors recently reported that the median price for existing homes has declined by 3.3 percent from October 2005 to October 2006. One of the local realtors said that the inventory of unsold homes in the Upper Connecticut Valley reached "huge" levels, the highest he has seen in the last three real estate cycles. "Many properties are getting price reductions of 5% to 10%, and the price of one property was reduced by 45%." The median price of vacation condos in Vermont declined by eight percent, and the decline was especially pronounced in the lower and middle-priced units.

The inventory of unsold homes is likely to increase for six or more months (a guess) as developers complete construction of single family houses and condo apartments that they started ten to twenty four months ago. Moreover the reported inventory of unsold homes almost certainly is an understatement because many would-be sellers have taken their properties "off the market" until prices "come back"-since prices aren't going to come back for a long time, these individuals will stay put if they stubbornly refuse to recognize the economic realities. If they recognize the realities, they will reduce their asking prices.

The stalemate between the buyers who are waiting for prices to decline and the sellers who are waiting for prices to increase has led to a decline in real estate transactions of ten to fifteen percent.

The Sunday Business section of the New York Times of November 12, 2006 carried a full page ad sponsored by the National Association of Realtors with the heading "It's a great time to buy or sell a home". The ad noted that "Housing prices are forecast to post gains in 2007, surpassing median home sales values in 2006". (Those who wrote the ad can point to a statement by Curley, Larry, or Moe or some other stooge that prices would increase; these forecasters are first cousins to those who wrote the books "The Dow at 36,000" and "The Dow at 100,000" in the late 1990s.) The ad also summarizes a line from a recent speech of Chairman Greenspan that "housing prospects are looking up. 'Most of the negatives in housing are probably behind us. The fourth quarter should be reasonably good, certainly better than the third quarter.'"

If Chairman Greenspan and the realtors are proven right, then the analytic model of asset price bubbles will be wrong. Moreover if they are right, then many first-time home buyers will spend forty five to fifty percent of their incomes on housing---which means that their spending on other goods and services must decline.

On Saturday November 18th both the Wall Street Journal and the New York Times summarized a press release from the Department of Commerce that new housing starts declined by 14.6 percent in October and were at the lowest rate since September 2000. Moreover data on housing starts in the previous several months were revised downward.

The National Association of Realtors has just reported that in October the inventory of unsold homes reached a new cyclical high of 7.4 months supply. Traditionally-before this froth or bubble in housing had began in 2001--about five to six million homes were sold each year and the inventory of unsold homes averaged about five months sales. If the inventory had been five months sales, and sales had averaged 500,000 units a month, the inventory would have been two and one half million units. Consider this a ballpark number.

Now the inventory of unsold homes is 3,750,000 units, or twenty five percent larger than the average value. These 750,000 excess units in the inventory are seven to eight months production of new housing units at the rate that prevailed before the current boom. That's a big number, a very big number. Moreover these inventories will inevitably increase for the next few months because the construction pipeline probably ten or twelve months long (okay, I'm guessing) and perhaps half of the units in the pipeline were started when developers were still bullish.)

To reduce the inventory the developers are reducing prices sharply; during the last six months the prices of new homes have declined twice as rapidly as the prices of existing homes. (The "portfolios" of the developers and the builders are less extensively diversified then the families that own one or two homes; moreover their carrying costs probably are higher.) Moreover this data estimate understates the "true decline" in the prices of new homes, since the developers are offering many more amenities at the "same prices"-more upscale appliances, a finished basement, cash to help with closing costs.

The Wall Street Journal story also noted that "A major factor driving up the supply of unsold homes is buyers getting cold feet and canceling contracts for new homes often at the last minute." (My first "data point" that the housing bubble had been pricked was in August 2005 when a real estate agent in Hanover reported that her firm was reducing prices and that many buyers were walking away from their contracts. The change in the last fourteen months is that the proportion of buyers that are reneging on their contracts and in many cases forfeiting their deposits has increased sharply.)

The Journal story notes that "The depth of previous downturns suggests that construction may have yet to hit bottom. In October, starts of single family homes were down 32% from a year earlier, compared with 45% (decline) in January 1991 and 52% (decline) in March 1980. (The "declines" in the brackets in the previous quote have been added to facilitate understanding.)

Because of the surge in real estate prices between 2001 and 2006, the increase in the ratio of the market value of U.S. residential real estate to U.S. GDP was two and one half times the increase in the same ratio during the cycle that peaked in the early 1980s and three times the increase in the ratio in the cycle that peaked in the early 1990s.

The ad sponsored by the National Association of Realtors noted that "The national median price of homes bought ten years ago has increased 88 percent." (Note the embedded sales pitch-if prices increase at the rate of seven or eight percent a year and the interest rate on your mortgage is six percent, you can live "rent-free" if you buy since the increase in the market value of your home exceeds the cumulative interest payments on your mortgage.)

This dramatic increase in household net worth has led millions of Americans to increase their spending relative to their incomes; the flip side of this statement is the decline in the household saving rate to a slightly negative value. In the last several years much of the growth in the U.S. economy has resulted from the increase in household consumption spending that was paid for with the money obtained from re-financings and home equity loans-the generic term is "equity withdrawals."

The U.S. economy now is beginning to adjust to decline in home production; new home sales peaked at the annual rate of 1,361,000 in July 2005 and the sales of new homes were 1,000,000 in November 2006. Developers appeared sluggish in reducing housing starts to reflect the economic realities. . Residential real estate construction has declined by thirty percent since the peak when spending reached 6.5. percent of U.S. GDP; now spending 4.5 percent of U.S. GDP, slightly above the long run average of four percent of GDP. The vacancy rate-the number of unoccupied homes-has increased; there are many more "for rent" signs.

The inventory of unsold homes will begin to decline only when the production declines relative to final sales. The implication is that construction spending will decline below its long run average value. This decline will occur continuously over the next four to six months as new starts decline further in those regions that have not yet experienced a surge in the inventory of unsold homes. The speculative purchases of homes in the last several years boosted the rate of growth of GDP by three tenths or four tenths of one percent of GDP; spending stimulated by this construction may have added an additional tenth of one percent. For the next several years production of new homes will decline relative to the final demand, and the rate of growth of GDP will be depressed relative to the long run trend.

The Tuesday November 7th edition of the New York Times had a front page story with the heading "After Arizona's Housing Boom, 'For Sale' is a Sign of the Times." The story has familiar features-sales cancellations are running up to forty percent, double the rate a year ago, the number of unsold homes has soared, "builders are pulling back as fast as they can". In the boom years, the median home price in Phoenix had increased by more than sixty per cent while household incomes increased by five percent; one third of the newly constructed homes were sold to speculators. Twenty to thirty percent of the employment growth in Phoenix has been in construction. Janet Yellen, President of the Federal Reserve Bank of San Francisco, described some of these new developments in the Phoenix area as "ghost towns".

One of the key questions is the magnitude of the forthcoming decline in home prices. Prices in virtually every market are set "on the margin"; relatively small changes in demand relative to supply lead to large changes in prices. The ads in the Wall St. Journal and the Chicago Tribune and the New York Times suggest that there are more real estate auctions. One recent ad in the Tribune for a new luxury home just off Lake Shore Drive has a "teaser" that characterizes some of the other ads-"Originally Priced at $2 Million-Suggested Opening Bid: $925,000." Moreover relatively more of these auctions are "absolute", that is, without a "reserve price." (The seller in a absolute auction agrees to accept the highest bid however low it may be; in contrast if the seller stipulates a reserve, the seller agrees to accept the highest bid only if it is above this reserve price.)

The increase in the ratio of market value of U.S. residential real estate to U.S. GDP in this housing boom and the two previous housing expansions is shown below; the increase in this ratio in this cycle is twice the increase in the 1980s cycle and nearly four times the increase in the 1990s cycle. (This ratio is one measure of "affordability"; as the ratio increases, homes are less affordable.) The decline in the ratio in the 1980s cycle was twenty percent of the increase in the previous expansion, while the decline in the ratio in the 1990s cycle was seventy five percent of the previous increase.

Price Changes in Three Real Estate Cycles
1980s 1990s 2000s
Base Year 1974 = 100 1983 = 100 1994 = 100
Peak Year 1982 = 126 1989 = 115 2006 = 156
Trough Year 1983 = 121 1994 = 104 2008/9 = 120?

The analytic model suggests that the market response to the exceptionally large inventory of homes-for-sale will be that prices will decline until the ratio of the market value of U.S. homes to U.S. GDP again is in the same ballpark as in the 1980s and the 1990s---which suggests a peak-to-trough decline of forty or so percent. The declines will be much sharper in those sections of the country that experienced the surge in speculative demand. Many of those who bought homes in 2004 and 2005 in search of speculative gains now have monthly carrying costs that are much higher than their rental incomes. Many in this group will become distress sellers and some will go bankrupt.

The prices of the futures contracts on housing that began trading on the Chicago Mercantile Exchange six months ago are at a annualized discount of six to seven percent relative to the current spot price; the inference is that the sellers of the futures contracts would rather take a seven percent haircut because they believe there is a non-trivial probability that the market prices will decline by an even larger amount during the next twelve months. The observation from a large number of markets is that futures prices and forward prices almost always "under-predict" the changes in the market price when the changes in the market price are large.

The New York Times article on Phoenix reports that some new homes can be purchased for their construction costs of $75 to $100 a square foot; buy the house and "get the land for free."

The second question--the banner headline on the cover of the November 13, 2006 issue of Fortune-is "Will the Housing Bust Kill the Economy?". The surge in home prices and home equity wealth in the last five years has been much greater than in any previous period in U.S. history--just as the increase in the market value of U.S. stocks in the late 1990s had been much greater than the run-up in stock market values in any previous period. Moreover the share of newly produced homes that was purchased for speculative gain has been much greater than in the 1980s and the 1990s real estate cycles. The surge in home prices had two major impacts on the growth of GDP; one was the increase in the number of homes constructed and the other was the equity withdrawals. Home construction is a proxy for sales of furniture and appliances, brokerage services, mortgage banking services, title insurance, and even moving.

Spending on residential real estate construction was $800 billion in 2006. If instead spending on had been at the four percent rate, the dollar value of spending would have been $520 billion.

The logic of the analytic model is that there will be two primary adjustments to the excess supply of housing---one is that the household savings rate will increase as the decline in homes prices leads to a reduction in household wealth and the second is that is that new home construction will decline until the excess inventory more or less disappears.

First consider the impact of the decline in household wealth on household saving. In the "good old days" the household saving rate had been in the range of five to six percent. Households can spend more than their incomes for one or two years, but they can't spend more than their incomes for an extended period. (Consider that if one group of households spends more than its income, another group of households must spend less than its income; this second group acquires the assets and the loans from the first group.) In recent years the second group of households has been in Japan, China, Russia, and many of the oil producing countries.

Each increase of one percentage point in household saving amounts to nearly a $100 billion reduction in the rate of growth of U.S. GDP; the larger the increase in the saving rate, the less rapid the increase in household spending, and the less rapid the rate of growth of the economy. The household saving rate might increase by two percentage points a year-which means that economic growth would slow dramatically unless there is a more or less compensating increase in the spending of some other group-the business sector, the government sector, or the foreign sector (an increase in the spending of the foreign sector would involve a reduction in the U.S. trade deficit.)

Consider two "corner outcome" scenarios for the adjustment to the excess supply of houses; the scenarios differ primarily in terms of the speed of absorption of the excess inventory of 750,000 housing units. One is that house prices decline rapidly in 2007 so that the excess supply is absorbed in the next twelve months; residential real estate construction would be in a free-fall to one to two percent of GDP; if the mid-point of this range is picked, then construction spending would be $200 billion. The decline in spending on residential real estate would be nearly $600 billion. A recession would be inevitable, and probably a severe recession.

This outcome is very unlikely because construction spending will continue on the units that are already in the pipeline; the developers understand the "sunk cost" fallacy and will complete the construction as long as the prospective selling price exceeds the incremental construction costs. Nevertheless toward the end of the 2007 the construction could be at the annual rate of $300 billion, less than half the current rate.

The alternative corner outcome is that both home prices and the excess supply component of the inventory decline slowly and continuously perhaps at the rate of $200 billion a year; the absorption of the excess inventory would take nearly four years. During this four year period annual spending on residential real estate would be $320 billion, less than fifty percent of the current level. . (If spending on residential real estate averages four percent of U.S. GDP, then spending would be $520 billion on the assumption that U.S. GDP is $13,000 billion.) The decline in spending would be dramatic, and a recession would be likely.

The number of different combinations of outcomes that vary with the pace of the decline in home prices and the impact on the construction is large.

A large number of those employed in home construction will lose their jobs. Similarly many of those employed in mortgage banking and real estate brokerage and in the production of furniture and durables will be adversely affected.

The slowdown in the U.S. economy became apparent in third quarter of 2006, the rate of economic growth declined to 2.2 percent. The leading economic indicators began to trend down-albeit at a very sluggish rate-in January 2006 even before the bearish housing news made it to the front page. By themselves the leading indicators are not sufficiently robust so they can forecast which of the slowdowns is likely to morph into a recession.

Why might this application of the standard analytic model be off-base? The estimates of the excess supply of homes-more or less the number that was purchased for speculative purposes-- may be too high. As Americans become richer, the share of their income that they are willing to spend on housing may increase modestly; one aspect of this increase is the purchase of second homes.

Moreover, spending in some other sector might increase as spending on new homes decline. Still it seems unlikely that business investment will increase as spending on residential real estate declines; the more likely outcome is that firms will retrench their spending as their profits decline and their projections of the anticipated growth rate are revised downward.

The negative slope on the interest rates on U.S. Treasury securities (the yield on the ten year bond is 4.43 percent and the yield on three month bills is 5.05 percent) suggests a significant economic slowdown in the next six to nine months; since 1960 a negatively sloped yield curve has forecast five of the six recessions. The National Association of Home Builders Housing Market Index has been negative and seems to forecast a decline in consumer spending. The trend in new car sales has been negative. The ISM (Institute of Supply Management) index of manufacturing has moved from 51.0 to 49.5; a value below 50 is interpreted as an indicator that manufacturing activity is slowing. Corporate inventories have increased sharply, an indicator that sales growth has lagged the growth in output.

The U.S. stock market is an outlier to this forecast; it is as if the aficionados of the two markets have totally different views about the developments in the U.S. economy in the next twelve to eighteen months. Profit growth has been phenomenal in the last few years (much much higher than my expectations) and the profit share of GDP now about thirteen percent, higher than ever before, and much higher than the historic average in the eight to nine percent range. The price-earnings ratio for the S&P500 is nineteen hence the earnings yield is 5.1 percent and the dividend yield is 1.50 percent. The anticipated rate of return on stocks is the sum of the dividend yield and the anticipated rate of increase in stock prices; stock prices must increase at the rate of three percent a year to provide the same rate of return as that available on bonds. The payment for the additional risk associated with stocks seems small.

One of the fascinating aspects of the last few months is that the spreads in interest rates between the least risky securities and riskier securities have declined; the interpretation is that those investors that need income-that are desperate for income-have "reached for yield." The risk attached to these securities has not declined. As the economy becomes softer, some of the firms will tank; the value of their securities will decline, and risk will be re-discovered.

Keep posted, there will be a lot of economic news in the next few months and most of it will not be comforting. In the meantime, save your money, many goods and assets will be less expensive six months from now.