This study provide empirical evidence whether bias in the standard errors of Jensen’s alpha explains conflicting results in the extant literature in real estate funds. Significant alphas in real estate mutual funds and REITs are compared with heteroskedasticity consistent covariance matrix estimators (HC1, HC2 and HC3), Newey-West standard errors, a robust regression tempering the effect of high leverage points, a GARCH model, and a HC3 adjusted wild bootstrap. In the analysis of real estate mutual funds and a separate sample set of REITs, the HCCME had a minimal impact attenuating the number of firms with excess returns. Contrary to expectations the differences from HC1 to HC2 to HC3 were also negligible. The Newey-West standard error provided highly variable results when compared with the OLS results particularly in the REIT sample. Of the techniques to adjust for bias in the standard error, the wild bootstrap with HC3 adjustment to the standard error provided the most conservative result to the number of real estate mutual funds and REITs with significant alphas. The co-movement of real estate funds suggests common exogenous influences. Including state variables such as the changes in unexpected inflation, term spread, default spread, market skewness and industrial production growth in a multi-factor model is used to identify systemic economic factors in significant alphas. The significant alphas varied with the inclusion of these variables, the time period and the bias adjustment.
Identifer | oai:union.ndltd.org:unt.edu/info:ark/67531/metadc699987 |
Date | 08 1900 |
Creators | Rogers, Nina |
Contributors | Karafiath, Imre, 1955-, Tieslau, Margie A., Winson-Geideman, Kimberly |
Publisher | University of North Texas |
Source Sets | University of North Texas |
Language | English |
Detected Language | English |
Type | Thesis or Dissertation |
Format | vi, 134 pages : color illustrations, Text |
Rights | Public, Rogers, Nina, Copyright, Copyright is held by the author, unless otherwise noted. All rights reserved. |
Page generated in 0.0014 seconds