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 introduces a tractable heterogeneous agent New Keynesian (HANK) model with a housing sector and house price belief extrapolation. The model consists of a saver and a borrower, where the borrower is hand-to-mouth and uses housing as collateral. We identify four main channels through which house price extrapolation affects borrowers' demand: an indirect channel through aggregate demand, a direct collateral channel, a precautionary savings motive, and a fire sale motive. Turning to monetary policy we show that the central bank faces a trade-off between stabilizing borrower consumption and house prices (i.e., saver consumption). Since only one policy tool is available, both cannot be stabilized simultaneously, making monetary policy alone suboptimal tool in this context.