Using State and Metropolitan Area House Price Cycles to Interpret the U.S. Housing Market
This brief examines the numerous house price cycles in states and metropolitan areas since the 1970s, drawing lessons that may be informative for analyzing and projecting national patterns. It finds that house sales volumes, new home construction, and mortgage delinquencies have provided leading indicators when a statewide house price boom was nearing an end, but that house prices have rarely decreased in the absence of a state recession. The median relationship suggests that the national OFHEO house price index could keep increasing well into 2007, given the sales and construction declines and the increases in delinquencies observed to date, and absent other factors that weaken the economy. However, the lead-lag relationships have varied considerably across states and time periods, indicating that turning points in house prices are difficult to predict precisely.
Next, the brief examines the empirical relationship between metropolitan-area house prices and measures of their deviation from justifiable values, as derived from economic models. It finds that the probability of a house price decline in metropolitan areas has depended on both the extent to which housing was overvalued two to three years earlier and on changes in market fundamentals that affect housing. An extrapolation of these results suggests that national house price increases are likely to be approximately flat for the remainder of 2006. Estimates for 2007 vary, depending on assumptions regarding mortgage interest rates, personal incomes, and apartment rents, and on one's views on how such factors influence housing markets. In addition, the brief also cautions that the economic models developed to date provide only partial explanations of past house price movements.
On the whole, the results can be interpreted as adding support for the view that average U.S. house prices are likely to be fairly flat in 2006 and 2007. Because house prices are subject to inexplicable movements, this conclusion should be viewed as a plausible extrapolation using historical evidence rather than a forecast. An additional caveat is that mortgage markets and other institutional factors may have changed sufficiently so as to alter the relationship between house prices and the economy compared with what existed in past cycles.