This study analyzes the time-varying nature and determinants of comovements in US housing prices using state and metropolitan statistical area (MSA) data. We employ dynamic factor models with time-varying loadings and stochastic volatility (DFM-TV-SV) to estimate the national, regional, and state factors. The timevarying factor loadings and stochastic volatility features enrich the dynamic factor model structures and are an effective tool to examine the comovements in housing prices. We find that the national factor is the dominant factor in explaining the movement of housing prices.The national factor accounts for 79% of the variation in state-level housing prices on average, with the greatest magnitude occurring during the housing boom and bust periods in many regions and states. We also find that the factors and synchronization effects are time-varying and heterogeneous across regions. The state-level housing prices contain higher national housing factor components in states with more diverse economies, higher wages and house prices, and lower unemployment rates.These findings shed light on the effectiveness of residential real estate diversification across the United States and the potential for elevated national housing risk amid economic downturns due to increased national housing price integration.

No abstract