Real Estate and its Role in Asset Pricing
61 Pages Posted: 30 Nov 2001
Date Written: November 17, 2001
This study examines whether residential and commercial real estate risks carry positive risk premiums. Real estate assets have been excluded from most of the empirical asset pricing literature because of perceived data and measurement problems. This paper shows, on the contrary, that the available data are sufficient to capture risks inherent in holding real estate assets. Using both Fama-MacBeth cross-sectional regression techniques and a stochastic discount factor GMM framework, I test whether the cross-sectional explanatory power of well-known asset pricing models can be improved by adding a real estate factor. I find strong evidence for the hypothesis that both residential and commercial real estate risks are priced by the market and therefore have a definite role in empirical asset pricing specifications. The main finding that returns to real estate improve the performance of empirical asset pricing specifications is not sensitive to the choice of assets being priced and holds for size and pre-beta sorted portfolios, size and book-to-market sorted portfolios, and portfolios in which assets are sorted initially by market value and then by sensitivity to a real estate index. The research set forth in this study is not only relevant for a better understanding of the empirical performance of linear asset pricing models but also has implications for the development of optimal investment strategies. Since most equity market investors are homeowners, the existence of a common real estate factor in asset prices has to be considered in portfolio choice decisions.
Keywords: asset pricing, capital asset pricing model, real estate, linear factor model
JEL Classification: C51, C52, G12, L85
Suggested Citation: Suggested Citation