Do Real Estate Investment Trusts (REIT) act effectively as a proxy for the real estate asset class? In their June 2011 paper entitled “REITs and Underlying Real Estate Markets: Is There a Link?”, Andrey Pavlov and Susan Wachter estimate the strength of the relationship between REIT and underlying real estate returns. Specifically, they relate REIT returns (unlevered based on debt and total capitalization data) to returns for shadow portfolios of Commercial Property Price Indices (CPPI) matched REIT-by-REIT based on geography and property type. They consider control variables such as interest rate changes, credit and term structure spreads, currency exchange rates, industrial production growth and commodity price changes. Using quarterly data for 2001 through 2007 for 71 REITs with a predominant property type, CPPI shadow portfolios, control variables and stock indexes, they find that:
- There is a strong positive relationship between REIT returns and real estate returns for the office sector.
- The relationship between REIT returns and returns on underlying real estate is weak for the retail, industrial and multifamily sectors
- Relationships between changes in REIT net asset values (in place of returns) and returns on underlying real estate is very weak.
- There is a significant positive relationship between REIT returns and returns for broad stock market indexes.
- Except for office REITs, a hedge portfolio that is long REITs and short the S&P 500 Index is largely ineffective in providing real estate exposure (as indicated by shadow portfolio returns).
- Findings are robust to model specifications and control variables, which relate more strongly to REIT returns than underlying real estate portfolio returns.
- A potential explanation is that the performance of the best REIT managers during economic downturns weakens the relationship between REIT returns and returns on underlying real estate, with exceptionally strong regional effects preserving the relationship for the office sector.
In summary, evidence indicates that REITs are generally not a good proxy for the real estate asset class.
Cautions regarding findings include:
- The sample period is short, perhaps lacking in variety of real estate market conditions and ending before full realization of the real estate market collapse and associated 2008 financial crisis.
- Statistical significance and investment materiality may differ.
- Statistical tests in the study assume well-behaved return distributions.
