The Office of Federal Housing Enterprise Oversight's HPI numbers for the third quarter are out today, and registered a quarterly decline for the first time in thirteen years, with the aggregate index dropping 0.4 percent versus the second quarter of 2007. The HPI purchase-only index dropped 0.3 percent on a seasonally-adjusted basis, OFHEO said. Year over year, the HPI showed an increase of 1.8 percent -- the lowest four-quarter increase since 1995. “While select markets still maintain robust rates of appreciation, our newest data show price weakening in a very significant portion of the country,� said Lockhart. “Indeed, in the third quarter, more than 20 states experienced price declines and, in some cases, those declines are substantial.� According to the HPI data, the states with the largest depreciation were Michigan (-3.7%), California (-3.6%), Nevada (-2.4%), Massachusetts (-2.3%), and Rhode Island (-2.2%). The MSAs with the largest depreciation for the same period were Merced, California (-13.0%), Punta Gorda, Florida (-11.8%) and Santa Barbara-Santa Maria-Goleta, CA (-11.6%). Seventeen of the 20 cities exhibiting the most depreciation in OFHEO's measures were in Florida and California; the other three were in Michigan. The HPI report also contained some interesting analysis on foreclosures and pricing trends, challenging the notion that a local decline in housing prices is linked to increased foreclosure activity:
A strong positive correlation between foreclosure filings and price declines is shown across the 50 states and the largest 100 metropolitan areas in the U.S. Using zip-code level foreclosure data, the analysis then looks within several high-foreclosure cities to determine whether prices in neighborhoods with particularly high foreclosure activity show greater price weakness. Although one might expect such neighborhood-level effects to be present, the limited empirical review suggests that price declines have been quite similar for high-foreclosure neighborhoods as compared to other areas.
In other words, price declines aren't correlating at the local level as much as you'd expect. The reason? OFHEO notes that prices are often sticky -- so while a pricing/foreclosure relationship will hold when aggregated sufficiently, it often breaks down when you drill down to a more local level. After all, areas seeing alot of foreclosures probably haven't seen many of these properties sell, meaning the price measures used haven't yet reflected any drop in prices. As Calculated Risk notes:
Translation: There are large price declines coming.