US disposable personal income and housing price index: A fractional integration analysis

This paper examines the relationship between US disposable personal income (DPI) and house price index (HPI) during the last twenty years applying fractional integration and long-range dependence techniques to monthly data from January 1991 to July 2010. The empirical findings indicate that the stochastic properties of the two series are such that cointegration cannot hold between them, as mean reversion occurs in the case of DPI but not of HPI. Also, recursive analysis shows that the estimated fractional parameter is relatively stable over time for DPI whilst it increases throughout the sample for HPI. Interestingly, the estimates tend to converge toward the unit root case after 2008 once the bubble had burst. The implications for explaining the recent financial crisis and choosing appropriate policy actions are discussed.