Time-Varying and Spatial Herding Behavior in the U.S. Housing Market: Evidence from Direct Housing Prices
Posted: 31 Mar 2017
Date Written: March 30, 2017
This paper investigates herding behavior in the US residential housing market. The sample period is 1975M01 to 2015M06. The study utilizes the housing price index of each of the 50 states and Washington DC to form nine census region-based markets, or portfolios and then employs switching and quantile regressions to examine the spatial and time-varying disparities of housing return dispersions and investors' herding behavior. The study find that the degree of herding varies across regimes, regions and conditional distributions. The regime-specific herd formation may be partially originated by extreme housing market conditions, bull and bear housing market conditions, uncertainty in national financial markets, economic recessions and uncertainty of economic policies. The bull housing market exhibits stronger effects on return dispersion than down markets, which is consistent with the "flight-to-safety" consenus behavior of investors. The study also finds that positive and negative linear and nonlinear returns magnify dispersions in an asymmetric manner. The increase in co-movement and interdependence of state and regional-level housing markets returns among geographically diverse states and regions offer little hope of successful geographical portfolio diversification strategies for U.S. housing market investors. Moreover, time-invariant modeling may yield incorrect inferences regarding herd formation in regional housing markets.
Keywords: Herding; CSAD; Housing market; Regimes; Switching regression
JEL Classification: G14, G15
Suggested Citation: Suggested Citation