Ripple effect and contagious effect in the US regional housing markets

Abstract

The existing researches show that the relationships across regional housing market in the USA are asymmetric, and the ripple effect may exist among regional housing markets. To study the ripple effect among regional housing market, this research uses vector autoregression model, Granger causality test and the forecast error variance decomposition to study the relationships within regional and cross-regional housing markets. Besides, this research also uses copula model to inspect the contagious effects between regional housing markets. The main empirical results show that, first, among the metropolitans under studied, Los Angeles in western region, New York in eastern region and Miami in southern region are the source of shock in their respective region. Second, apart from those in southern region, conditional contagious effects between regional housing markets are dynamic during the study period and there exist obvious contagious effect during the period of global financial crises started from 2008 to 2009. Third, contagious effects are more significant within the same region but less obvious in cross-regional housing markets. We also found that housing price comovement in down markets are rare to observe than housing price comovements in up markets.