With a potential housing bubble still a possibility in the minds of a few, analysts at GMAC RFC Securities recently issued a new quantitative measure to determine which metropolitan statistical areas (MSAs) may be subject to a drop in housing values. Their finding: The vast majority of U.S. housing markets have shown no evidence of a housing bubble, the test of GMAC RFC shows.
The persisting question of a housing bubble arises from the persistent asymmetry between the rate of change in housing prices and personal incomes in the country, wrote Michael Youngblood, director of research at GMAC in analysis last week.
For instance, the median existing house prices rose by 8.8% and nominal personal incomes increased 3.3% in the year ending September 2003. Furthermore, in the past year, housing prices increased an average of 5.0% in 318 MSAs. In contrast, personal incomes rose by an average of only 3.1%. The report added that the rate of change of house prices is more than the rate of change of personal incomes in 218, or 68.5%, of the nation's MSAs.
To identify a bubble in an MSA, the movement of house prices should considerably exceed the movement that is caused by fundamental factors. Merely rapid or prolonged increases in housing prices are not enough to constitute a bubble.
The new test uses the underlying distribution of the ratio of housing prices to per-capita personal income, including in its analysis the ratio in each MSA as of September 2003. Those performing the new measure tested the validity by applying it to the housing markets of Boston from 1985-1989 and Los Angeles from 1988-1993. The test confirmed that during these time periods, bubbles had actually occurred in these states. After making sure of the test's validity, analysts calculated the new test for bubbles in 318 MSAs as of September 2003.
The report confirms the existence of housing bubbles in 29 MSAs, including 13 California regions. The aggregate population of the 29 MSAs is 28.7 million, constituting 10.2% of the U.S. population in the 2000 Census. Aside from these MSAs, a considerable percentage of areas do not have housing-price bubbles; the vast majority of the population - 88.8% - lives in these MSAs.
MBS participants would benefit from knowing which markets are stressed in order to make sound investment and underwriting decisions. As one example, this would allow servicers to focus their time and resources in markets where appraisal prices have more variability. A servicer could focus on appraisal verification only in markets where there is likely to be overvaluation instead of merely conducting a random sampling of, say, 10% of all appraisals in a given month.