Thursday, September 11, 2008

The Link Between Foreclosures and House Prices

Rising foreclosures will not cause U.S. home values to plunge, despite widespread concerns to the contrary. That's the conclusion of a new and first-of-its-kind study, The Foreclosure-House Price Nexus: Lessons from the 2007-2008 Housing Turmoil (NBER Working Paper No. 14294) by Charles Calomiris, Stanley Longhofer, and William Miles. Although the authors recognize that other factors not captured by their analysis could weigh on home prices, the effects of foreclosure shocks - which promise to grow over the next several months, and which have been a source of worry to homeowners and economists - seem to be smaller than many have feared. Even under their most extreme scenario, in which foreclosure rates would substantially exceed current forecasts, the resulting average drop in home prices between the national peak in the second quarter of 2007 and the fourth quarter of 2009 would be less than 6 percent.

The authors emphasize that house-price declines vary across states and argue that headlines pointing to extreme circumstances in a few states can be misleading about the United States as a whole. Despite increased foreclosure rates throughout the country, only 12 states are projected to see foreclosure-induced price declines of 6 percent or more through 2009, led by Nevada, Florida, California, and Arizona. "This suggests that home prices are quite sticky, and that fears of a major fall in house prices, with all of its attendant negative macroeconomic consequences, typically are not warranted even in extreme foreclosure circumstances," they write.

Part of the reason that foreclosure shocks have small effects on house prices is that these shocks tend to occur late in the housing cycle, after housing starts have declined and the supply of existing homes on the market has fallen sharply. These effects largely offset the price consequences of a supply surge caused by foreclosures.

Another contributing factor to the observed stability of house prices is the measure of price change chosen by the authors. The authors argue that it is appropriate to focus on a house price measure related to the prime conforming segment of the mortgage market (which accounts for more than three quarters of American homes). The authors seek to measure foreclosure effects on the values of homes sold by typical sellers, not the declines in prices of homes undergoing foreclosure-induced distress sales. They argue, therefore, that the house price index from the Office of Federal Housing Enterprise Oversight (OFHEO) -- which does not include subprime home sales -- is the most reliable and useful dataset for their purposes.

Using quarterly data for each state going back to 1981, the authors model the dynamic linkages among five variables: foreclosures, home prices, employment, permits issued for single-family homes, and existing home sales. Using state-level data makes it possible to measure linkages using the frequent and significant ups and downs that occur in state and regional housing markets. In contrast to the aggregate national market, individual states have seen larger and more volatile swings in foreclosures and house prices since the 1980s. By concentrating on the states, the authors also can take into account the effects of widely varying employment growth during that period -- an effect that continues to define important regional differences, in particular between housing trends in the Rust Belt and the West.

One limitation of the authors' model is that it assumes that rising foreclosure rates have the same incremental effect on house prices regardless of whether the foreclosure rate is high or low. In fact, the incremental effect of increases in foreclosures on prices is much larger when foreclosure rates are high than when they are low. The authors adjust their model to account for this by increasing their assumed foreclosure forecasts for 2008-9 by 53 percent. To test the sensitivity of their results to even greater foreclosure risks, they also build an "extreme-shock" scenario and boost the foreclosure projections by 75 percent. These two scenarios create modest downdrafts in home prices that average 4.7 and 5.5 percent, respectively, through 2009.

"We do not have a crystal ball," the authors conclude. "Our estimates are based on relationships among house prices, foreclosures, and other variables observed in the past. It is conceivable that unusually tight consumer credit conditions, or other factors, could weigh on the housing market and produce more price decline than we estimate." But "based on the past experience of the housing cycle, even when one proverbially bends over backwards to inflate estimated foreclosures and take account of…" their effects on house prices, there is no reasonable basis "…for believing (as many commentators do)that the housing wealth of consumers has fallen or will fall by much more than 5 percent," they write.

No comments: