The headlines read that housing has bounced back and that sales have bottomed. Recently the Commerce Department reported that sales of new, single-family homes rose 3.4 percent in November. The government even revised the previous three months, showing stronger activity. The national and local news reporters are all claiming housing is on the rebound. But has it?

The Devil is in the Details

The Census Bureau surveys home builders and specific housing starts. If a home is built and at some point goes under contract for sale, then it is considered sold. They do not go back and check if the sale actually went through.

So what happens when Lennar, the # 3 homebuilder, says the cancellation rates for home sales was running about 30% last quarter and KB Homes says their cancellation rate was 43%. Due to the way the Census Bureau performs their survey, they do not double count if the house sells at some point later to another person. In the long run, this method does not over or under count the estimate of the number of houses sold.

However, in the short run, as their web site explains "As a result of our methodology, if conditions worsen in the marketplace and cancellations are high, sales would be temporarily overestimated." So when the existing home sales data seems to be showing that the market is bottoming, we need to recognize that there are significant cancellations that are missing from the data. You can access the actual Census Bureau policy at the following link: http://www.census.gov/const/www/salescancellations.html

The headlines are missing the underlying trend. The cancellations in sales of new houses imply that sales are lower than being reported. It doesn't seem likely we have hit the bottom in new home sales.

What does History Tell Us?

We keep hearing that the bottom in housing is either here or very near. But is it? If the reported sales are skewed, then how can we tell? Maybe a look at some history will give us some insight. Guerite Advisors provides some excellent analysis that is useful. In the previous seven cycles since 1959, housing starts have dropped an average of 50.7% from peak to trough. Each time housing starts have fallen more than 25% from their most peak, a recession has followed. The only exception was the "credit crunch" of 1966-67 that ended in an economic contraction but not an official recession.

According to Guerite Advisors, shows housing starts have dropped 34% so far from their peak in January 2006. To get to the average we need to have another 20% drop in starts. This is another bad sign that contradicts the headlines.

Professor Robert Shiller of Yale, author of Irrational Exuberance: Second Edition, tracks housing prices, adjusted for inflation, since 1890. This index is for existing houses and not new construction. It presents housing values in constant terms over 156 years, factoring out the effects of inflation. The benchmark was set to 100 in 1890. This means that a house that sold for $100,000 in 1890, inflation adjusted to today's dollars, receives a value of 100. The equivalent house that sold in 2006 for $199,999 would receive an index score of 199, or 99% above the price in 1890. Here is the link to Shiller's site: http://www.econ.yale.edu/~shiller/

Since 1997 the price index has risen 83%. Professor Shiller believes that the index is mean regressing, meaning that over time it tends to return to the mean from the extremes. It is easy to understand that the spike in home prices is substantially above the mean, no matter how you would calculate it.

What concerns me the most is the wide disparity between the price of houses and building costs. I would think this has to narrow over time. Either the price of houses comes down, or the cost of building a home goes up or some combination of the two. In any case, it implies that housing will face more difficult times.

The Case for Soft Landing in Housing

So what will mitigate the problems in housing and help it find a soft landing and then move up. In the past declines in housing were accompanied by high unemployment, high interest rates and a general slow down in the economy. So far we have low unemployment and low interest rates long term interest rates. Even if unemployment were to rise in 2007, keep in mind that approximately 80% of the US economy is in the service sector, which tends to experience fewer layoffs than the manufacturing sector during an economic slow down or recession.

The 10 year treasury rate, the primary index for mortgage rates, is still in a relatively low 4.6% area. As long as mortgage rates stay relatively low, housing should be able to hold its own.

Conclusion

The headlines on housing sales are misleading in that they are based on data that is in accurate over the short term. Calling a bottom in housing on these misleading numbers gives a false sense that the industry is doing better than it is. Be careful making investment decisions on this data. There are factors that continue to push housing down further and there are factors that are helping to moderate the fall in housing. It will take time for this to work itself out.