Spillover of mortgage default risks in the United States : evidence from metropolitan statistical areas and states

Abstract

This paper offers a new perspective to the analysis of spillover transmission in the housing market, specifically dealing with mortgage default risks. To do this, the recently developed generalized forecast error variance decomposition (FEVD) methodology proposed by Diebold and Yilmaz (2014) is utilized to investigate the degree of interconnectedness of mortgage default risks in metropolitan statistical areas (MSAs) and states of the U.S. The empirical findings, based on a real-time mortgage default risks index, reveal complex interconnectedness across twenty MSAs and forty-three states. Our study finds that Chicago, New York, and Los Angeles are net transmitters of spillover effects to other regions in the housing market investigated. This study also corroborates with the central place theory (CPT), as Washington DC serves as a key player in the housing market among the MSA's. Amongst the states, Minnesota, followed by Arizona, Pennsylvania, New York and New Hampshire, are found to be the main source of mortgage default risks spillovers. Implications of our results are discussed.