House Price Growth Interdependencies and Comovement

Abstract

This paper examines house price diffusion across metropolitan areas in the United States. We develop a generalization of the Hamilton and Owyang (2012) Markov-switching model, where we incorporate direct regional spillovers using a spatial weighting matrix. The Markov-switching framework allows consideration for house price movements that occur due to similar timing of downturns across MSAs. The inclusion of the spatial weighting matrix improves fit compared to a standard endogenous clustering model. We find seven clusters of MSAs that experience idiosyncratic recessions plus one distinct national house price cycle. Notably only the housing downturn associated with the Great Recession spread across all of the MSAs in our sample; other house price downturns remained contained to a single cluster. Previous research has found that housing cycles and business cycles are intertwined. To examine this potential relationship we apply our spatial Markov-switching model to employment growth data. We find that house price comovement and employment comovement are distinct across cities.

Keywords: housing cycles, Markov-switching, cluster analysis

JEL Classification: C31, C38, R30