Endogenous Market Segmentation and the Volatility of House Prices

We study an economy where households face transactions costs of participating in the housing market. In response to these costs, households choose to buy and sell houses infrequently. This, in turn, implies that the value of the typical transaction is large relative to income. In this way our model captures two essential features of household-level adjustments to residential capital. Moreover, it implies that, in any period, only a fraction of households are active in the housing market, and that this fraction evolves with the aggregate state of the economy. We find that this endogenous market segmentation amplifies and propagates the response in the relative price of housing following aggregate and sectoral shocks. Because those households currently active in the real estate market must absorb changes to the aggregate housing stock in equilibrium, market segmentation exacerbates changes in house prices. Moreover, it implies large and persistent changes in the distribution of households following an aggregate shock, which themselves cause additional movements in prices. Thus households' optimal (S,s) policies, driven by nonconvex transactions costs, not only induce lumpy adjustment at the individual level, but also explain a nontrivial fraction of observed excess price volatility in real estate markets.