Return distributions of private real estate investments.

摘要

This dissertation investigates risk in private real estate markets, offering a theoretical and empirical analysis of the distribution of returns accruing to individual owners of residential real estate investments. Previously, most research in investment real estate concentrated upon large institutional real estate owners. Dominating the prior body of research are the finance paradigms, tools and methodology of the stock market.;Individual real estate investors, faced with an inefficient market of heterogeneous assets that are illiquid, difficult to divide in kind, impossible to sell short and trade after private negotiations at a local level, find implementing MPT implausible if not impossible. The ability to diversify away non-systematic risk is severely restricted. The use of a two-parameter model such as the mean-variance rule applied to financial assets is insufficient to describe the motivation of these investors. As this market is characterized by unavoidable non-systematic risk, it is reasonable to assume that return distributions contain more variance, in the form of heavier tails. One type of heavy tailed distribution is the family of Stable-Paretian (SP) distributions. The assumption that returns are SP distributed provides a four-parameter model sharing with the normal the important properties of stability under addition and linear transformation.;The theoretical position of the dissertation is that investors accept the greater risk implied by Stable, non-normal distributions in return for a positive skew. Thus the rule becomes: More variation is acceptable provided it is right tailed variation.;Using a unique data set covering only apartment investments of non-institutional owners over a 15 year period, the findings are that returns are likely to be Stable-Paretian distributed with a characteristic exponent, alpha, of 1.4218 +/- .0681 at the 95% confidence level. Stable distribution theory suggests that alpha at this level cannot support MPT, its efficient frontier or the gains purportedly to be made from diversification. A second important finding is that the SP distribution is skewed right. The measure of skewness for SP is beta ⊂ (-1, 1). Positive beta means that the distribution is skewed right. Data in the research here was estimated to have a beta ≈ .2579.