In this paper we use a standard neoclassical model supplemented by some frictions to understand large price swings in the housing market. We construct a two good general equilibrium model in which housing is a composite good produced using structures and land. We revisit the connection between changes in interest rates, credit conditions as measured by maximum loan-to-value ratios and expectations in influencing housing prices in a setting in which the stock of housing can be used as collateral for borrowing and credit markets are segmented. We find that changes in interest rates and credit conditions can generate significant price swings. Under rational expectations (perfect foresight) our model is able to explain 50% of the recent movements in U.S. house prices. When we allow shocks to expectations, the model’s ability to match the evidence increases significantly. Contrary to conventional wisdom, we show that standard asset pricing formulas seem to correctly describe the behavior of house prices if the appropriate pricing kernel is used.