A0585
Title: Housing bubble contagion across US cities
Authors: Michael Owyang - Federal Reserve Bank of St Louis (United States)
Jeremy Piger - University of Oregon (United States)
Daniel Soques - University of North Carolina Wilmington (United States) [presenting]
Abstract: A multi-region qualitative vector autoregression (QualVAR) model is developed to study the transmission of housing bubbles across US cities. Each city transitions between fundamental and bubble regimes based on a continuous latent variable that depends on lagged regime indicators from other cities. The model allows for the possibility that the likelihood of entering a bubble regime in one city is influenced by the regime history of others, enabling the detection of contagion in housing market dynamics. Fundamental house prices are determined by regional economic drivers, while bubbles are defined as persistent deviations from these fundamentals. Bayesian estimation is conducted using panel data on MSA-level housing prices and regional fundamentals. The model provides a tractable structure for analyzing the spatial diffusion of housing bubbles and supports the real-time identification of speculative episodes. This approach contributes to the understanding of interdependence across regional housing markets by allowing for regime interactions in a dynamic, multi-region setting.
