|dc.description.abstract||Real estate investment trusts (REITs) is a useful way to channel and structure the capital flow to the real estate market. Historical statistics indicate that the market capitalization of the REITs industry amounted to $600 billion at the end of 2012, representing an increase of 260 times since 1972. REITs have emerged as means for both institutional and small investors to hold, and invest in, diverse property assets after 1992. Like any other publically traded stocks, REITs are listed and traded on major stock exchanges, and most listed on NYSE.
The study consists of three empirical essays. This fist essay provides a cross-sectional and time-series investigation of conditional and unconditional expected returns of real REITs index momentum portfolios against real estate property, big-cap stock small-cap stock, and bond index. The results indicate that REITs returns exhibit a higher correlation with up move of financial market, but a lower correlation in market downturns. REITs may possibly provide diversification benefits to multi-asset investment portfolios. The results also show that the returns of momentum portfolios are different from the NAREIT index, and display asymmetric volatility as well. The results of regressions also indicate that REITs return exhibit the greater sensitivity to large- and small-cap stock index, and less closely with those of bond and real estate index.
The second essay explores the time-varying relationship between the return and risk for the portfolios in the U.S. real estate investment trust market. Three categories of REITs portfolios are formed based on the features of size, momentum and book-to-market portfolios. Under the conditional capital asset pricing model this essay uses the generalized autoregressive conditional heteroskedasticity in mean model by pooling the time-series and cross-sectional effects to estimate the conditional covariance with the market portfolio and test whether the dynamic conditional covariation predicts the time-varying expected. When restricting the slope to be the same within the REITs portfolios, the empirical results place a positive tradeoff between the return and predictable covariation in all REITs portfolios. On the other hand, relaxing to different slopes across the portfolio also conclude a positive return-risk tradeoff, besides the Winner portfolio in momentum portfolio shows a higher return at a lower covariance level. Furthermore, this paper examines the significance of intertemporal hedging demand in each REITs portfolio by extending the intertemporal capital asset pricing model with a set of prevailing macroeconomic variables and financial market indicators. This essay successfully examines the predictable movements that in return could be attributed to changes in covariances with innovations in macroeconomic variables and financial market indicators. Overall, the conclusive results show that innovations in inflation rate, de-trended short-term interest rate, Fama-French momentum factor, S&P/Case-Shiller home price index, and Barclay Capital long-term government/corporate bond index play a crucial role in hedging demand for the REITs portfolios.
The third essay gives empirical evidence of time-series predictive capability of average variance and average correlation on the return of U.S. REITs market. Using an approach based on quantile regression, this essay explores the ability of average variance and average correlation to predict the distribution of REITs returns. The results indicate that lower average variance is significantly related to large future loss, whereas higher average variance is positively related to large gain. Finally, this essay studies investment trading strategies that condition on average variance and average correlation, and shows that the implementation of these strategies, by focusing on the tails of the return distribution of REITs returns, would generate a better performance than the benchmark return.||