Home Prices, Rating Agencies & Common Sense

As I have noted many times over the last 18 months, maintaining a stable relationship between median incomes and median home prices is essential. Between 1965 and 2000, median home prices and median income held fairly steady, averaging about 3 to 1. This meant home prices rose based on an increase in income. For instance, if median income was $30,000 in 1965, the median home price as $90,000. As median income grew from $30,000 to $40,000, median home prices rose to $120,000. This meant home prices were supported by a stable relationship between home prices and income, which enhanced the home owner's ability to make his monthly mortgage payment on time.

Between 1965 and 2000, the U.S. experienced five recessions, including deep recessions in 1973-74 and 1981-82. Even when the unemployment rate climbed to 9.0% as a result of the 1973-74 recession, and 10.8% in 1982, median home prices did not fall. This underscores the importance and durability of maintaining this 3 to 1 ratio between median income and median home prices. The current crisis began with the unemployment rate well below 6%. Why?

Between 2000 and 2006, median home prices rose from 3 times median income, to 4.5 times median income. In some markets - Florida, Arizona, Nevada and California, the ratio was stretched to more than 6 to 1. It is no surprise that the markets with the highest ratio of home prices to income are experiencing the largest price declines. History suggests that mortgage lending policies should be geared to maintain a 3 to 1 relationship between median home prices and median income, nationwide and regionally.

Since the Great Depression, nationwide median home prices had never declined, even during the deep recessions of 1973-74 and 1981-82. Noting this record, the rating agencies (S&P, Moody's, Fitch Services) assumed nationwide home prices would never decline. Incredibly, the risk model they used to evaluate the quality of mortgages didn't even have the capability to assess loss levels, if home prices did actually fall. In the world of the rating agencies, market driven home prices could only go up!

In the future, the rating agencies will use a risk model that does allow for home price declines. Their risk model should also incorporate the relationship between median home prices and median income, nationwide and regionally. As home prices rose from 3 to 1 times median income, to 4.5 to 1 in 2006, and to more than 6 to 1 in a number of markets, the rating agencies failed to recognize the obvious.

As home prices moved 50% above their long term average of 3 to 1, the risk of a significant decline in home prices had risen substantially. And in those markets that had stretched to more than 6 to1, a decline in home prices was virtually guaranteed. Any first year Statistics student learns about regression to the mean. The financial engineers at the rating agencies obviously flunked their introduction to Statistics class.

The other significant flaw with the rating agencies is that they were compensated by the issuers of the debt they were evaluating. No risk of conflict of interest in that relationship! Between 2002 and 2007, the ratings agencies collected billions of dollars in fees. In the future, the ratings agencies should be compensated by those who buy the securities they rate. This will eliminate the conflict of
interest potential.

 |  |