Authors: Michael S. Young and Susan Annis

Start Page: 3
End Page: 28
Volume: 23
Issue Number: 01/02
Year: 2002
Publication: Journal of Real Estate Research

Abstract: This article is the winner of The Best Research Paper Presented by a Practicing Real Estate Professional manuscript prize [sponsored by the American Real Estate Society Foundation (ARESF)] presented at the 2001 American Real Estate Society Annual Meeting. 

The popularity of performance attribution in the publicly-traded equities arena may soon spill over to real estate markets. With that in mind, this study analyzes the practical and statistical problem that may arise when real estate managers apply this technique to their portfolios. The study involves three data sets: a portfolio of publicly-traded REITs, a single-client separate account and a multi-client private REIT. The findings indicate that there is no clear distinction between stock selection and sector allocation in any of the data sets (i.e., the portfolio impact of the manager?s sector allocation and asset selection decisions are, on average, indistinguishable). Also, for the publicly-traded REIT portfolio (the only data set with sufficient sample size), the monthly returns attributed to stock selection versus sector allocation do not display significant serial persistence (i.e., the manager cannot consistently attribute the portfolio returns to either the stock selection or sector allocation decision).