Asset Prices, Local Prospects and the Geography of Housing Dynamics

Abstract

I uncover a novel link between economic fundamentals and real estate returns. First, using a measure of (i.e. MSA) growth prospects from both industry level employment and financial data, I am able to explain a substantial portion of variability in excess real estate returns and price-to-rent ratios. A one standard deviation decline in local growth prospects is associated with a 120bp drop in the housing risk premia and a 500bp increase in the price to rent ratio. Second, I establish the underlying dynamics of relative, housing versus non-housing, consumption in the MSA-level cross section of economic agents. I document that a one standard deviation decline in local growth prospects is associated with a 1000bp rise in level and 700bp decrease in variance of relative consumption. In contrast, a one standard deviation increase in local growth prospects is also associated with a rise in relative consumption. This suggests that housing acts as a hedge against long-run economic growth. I then investigate these empirical facts through the lens of a consumption based, asset pricing model. I combine a persistent component in consumption, Epstein and Zin (1989) preferences, and nonseparable consumption of non-housing goods and housing services. Consumption of housing services provides a hedge to long-run growth prospects. The model is able to replicate the excess returns and price to rent ratio regression results from the data.