Academic journal article Journal of Housing Research

Understanding Systematic Risk in Real Estate Markets

Academic journal article Journal of Housing Research

Understanding Systematic Risk in Real Estate Markets

Article excerpt


A one-factor pricing model is employed to investigate the internal consistency of single- family home and professionally-managed property prices. The risk factor used here is the U.S. real estate index, which has much stronger explanatory power than the S&P 500 Index for real estate assets. Empirical tests with this model lead to several surprising results. First, portfolios of East Coast or West Coast cities have negative risk-adjusted returns (alpha), while a portfolio of all inland cities has positive alpha. Second, a momentum strategy does not outperform the U.S. real estate index on a transaction and risk-adjusted basis, despite its ability to pick the largest-growth cities. Third, high-beta cities have negative alpha, while low-beta cities have positive alpha, even after considering transaction costs. Fourth, high rental yield cities have positive alpha and vice versa, even after transaction costs. Fifth, large cities have negative alpha, while small cities have positive alpha. Finally, expensive cities have negative alpha and vice-versa. A possible explanation for these abnormal returns is that some cities are systematically neglected by investors.

(ProQuest: ... denotes formulae omitted.)

Despite the massive size of the residential real estate sector in the United States and its connections with the macroeconomy, little is known about differences in systematic risk across city home prices. Due to the infrequent trading of real estate properties, there are aggregation and smoothing issues that make it difficult to estimate a meaningful correlation with common risk factors in the stock and bond markets. The classical CAPM, which uses the S&P 500 Index as a risk factor, cannot explain differences in risk across cities. Several authors (Case and Shiller, 2003; Glaeser and Gyourko, 2006; Smith and Smith, 2006) have noticed a difference between the remarkable growth and volatility of coastal metropolises and the more staid advance of inland cities, but no comprehensive risk-return tradeoff has been provided.

We introduce a one-factor model, using the U.S. real estate index as a risk factor instead of the more common S&P 500, for assessing the risk-return tradeoff of real estate assets across U.S. cities. There are several reasons for including the U.S. real estate index as a factor. First, asset pricing models such as the CAPM require that the market portfolio be formed of all existing assets, not just the S&P 500. Stock market assets, as leveraged claims on business enterprises, capture a large part of the risks in the economy. But so too are housing price dynamics as evidenced during the current crisis. When assessing the systematic risk of real estate across U.S. cities, we should at a minimum include a U.S. real estate index as a risk factor.

Second, the U.S. real estate index has much more explanatory power in one-factor model regressions than the S&P 500 Index. This may be because price series for individual cities and for the U.S. real estate index are constructed by a repeat sales regression method, which might introduce aggregation and smoothing. The U.S. real estate index explains about 0.60-0.70 of city-level return variation, whereas the S&P 500 has only marginal explanatory power. Multifactor regressions have difficulty explaining the performance of the overall real estate market. Several risk factors such as the S&P 500, size, book-to- market, stock momentum, bond term spread, and bond default spread (Fama and French, 1992, 1993) do little to explain real estate excess returns. Even lags of these risk factors (Scholes and Williams, 1977; Dimson, 1979) cannot illuminate on the source of excess returns. The literal implementation of these multifactor models leaves an unexplained alpha of about 7% per annum for single-family homes and about 6% per annum for professionally managed properties. Empirically, we find that the U.S. real estate index works in explaining systematic risk across U. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.