A Test of the Construct Validity of Equity Real Estate Investment Trusts' Returns as an Operationalization of the Construct 'Real Estate Returns'

This research tests the construct validity Equity Real Estate Investment Trusts' (EREITs) returns as an operationalization of the construct "real estate returns". Numerous studies have attempted to correlate EREITs' returns with other measures of real estate returns such as the Russell-NCREIF (RN) Index returns with conflicting results. Prior studies have established correlation between the RN Index total quarterly returns and EREITs' total quarterly returns but have been unable to establish correlation between EREITs' income returns and the RN Index income returns. Prior studies of EREITs' returns have applied cross sectional statistical methodologies or have incorporated limited time series methodologies such as multiple regression with lags. The conflicting results of prior research may be due in part to the statistical methodology used. The contribution of this research is to apply a multiple time series methodology to establish whether the income and capital return components of the RN Index returns and EREITs' returns are correlated. Using a multiple time series methodology avoids potential spurious correlation, identifies both leading and lagging relationships, and removes autocorrelation from the series. A sample EREIT portfolio of 17 EREITs is formed and from this sample portfolio a Cash Flow Index and a Share Price Index are generated (independent variables). The RN Index is used to generate a Cash Flow Index and a Yield Rate Index (dependent variables). Two models are then developed to test the research hypotheses. The first model tests the relationship between the cash flows. The second model tests the relationship between the RN Index Yield Rates and the EREITs' Share Price Index. Data is analyzed on a quarterly basis for the period 1980 to 1992. The results indicate that the cash flows produced by the RN Index portfolio are significantly correlated with the dividends produced by the sample EREIT portfolio. The results also indicate that there is no significant correlation between the RN Index Yield Rates and the EREIT's Share Price Index. The results do not support EREITs' returns as an operationalization of the same segment of the real estate market returns as the RN Index returns measure.