Standard theory predicts a positive relationship between risk and return, yet recent data show that housing returns vary positively with risk in some markets but negatively in others. This paper rationalizes these cross-market differences in the risk-return relationship for housing, and in so doing, explains the puzzling negative relationship. The paper shows that when the current house provides a hedge against the risk associated with the future housing consumption, households are willing to accept a lower return to compensate for risk, thus weakening the positive risk-return relationship. Further, in markets with less elastic housing supply and a growing population, hedging incentives can be sufficiently strong to make the relationship negative. The empirical analysis confirms these predictions, suggesting that hedging incentives, housing supply, and urban growth are indeed central to understanding the risk-return relationship for housing.

We perform a comprehensive neighborhood-level analysis of housing supply. Predictions of floor space and housing unit supply elasticities using our estimates average 0.5 and 0.3 across all urban neighborhoods in the United States, exhibiting greater variation within than between metro regions. New construction accounts for about 50% of unit supply responses, with important additional roles for teardowns and renovations. Supply responses grow with central business district distance mostly from the increasing availability of undeveloped land, flatter land, and less regulation. Identification comes from variation in labor demand shocks to commuting destinations, as aggregated using insights from a quantitative spatial equilibrium model.