This paper shows that regional disparities in house price growth are more pronounced during house price busts than during booms. To explain this observation we construct a two-region currency union model incorporating a housing sector and extrapolative belief updating regarding house prices. To solve the model, we propose a new method that efficiently handles extrapolative belief updating in a wide class of structural models. We show that intensified extrapolation in busts and regional housing market heterogeneities jointly explain elevated regional house price growth dispersion in busts and muted dispersion in booms. Consistent with our theory, we provide empirical evidence that house price belief updating is indeed more pronounced in busts and document that regional heterogeneities on the housing supply side affect regional house prices. Quantitatively our model can match the empirically observed elevated regional house price growth dispersion in busts. Moreover, we demonstrate that a monetary authority targeting house prices may reduce the volatility of output and house prices, as well as regional house price growth disparities. This policy is welfare-improving relative to an inflation-targeting benchmark.

This paper examines the interaction between housing and business cycles in a tractable two-agent New Keynesian model featuring extrapolative house price beliefs. The model includes a saver and a hand-to-mouth borrower who uses housing as collateral. We identify four key transmission channels from housing markets to aggregate output: consumption, residential investment, collateral, and fire sales. Under rational expectations, output volatility is limited and primarily driven by consumption. In contrast, extrapolative beliefs significantly amplify output volatility, mainly through residential investment. Finally, we propose a novel solution approach for two-agent models with asset trade and asset price extrapolation, which is essential for solving models with fire sale motives.