Stocks Are from Mars, Real Estate Is from Venus

Abstract

How does the expected total return, or opportunity cost of capital (the discounted cash flow discount rate), differ across different types and locations of properties? This question is fundamental to institutional real estate investment portfolio and asset management. The cross-section of the long-run investment performance of real estate assets should be related to plausible measures of risk. The authors study this question using original, customized investment performance indexes constructed from the property-level NCREIF Index database. They find that a multi-factor model similar in structure to that of Fama and French predicts quite well the cross-section of long-run performance within NCREIF. But the real estate model is curiously opposite to that of the stock market. Is this an arbitrage opportunity?