We examine the performance of global active real estate mutual fund (REMF) industry as a whole and at individual fund level, relative to the global real estate and stock markets, and both before and net of expenses. We have used the cross-sectional bootstrap to separate genuine skills from luck. We also evaluate whether global active REMF, at aggregate industry and individual fund level, can gain significant benefit from investing internationally, by comparing them with U.S. domestic real estate and stock market. A series of robustness tests are implemented, especially conditional model, timing model, subperiod test, recursive estimates, and Kalman filter, to examine whether performance and global diversification benefit are sensitive to the impact of macroeconomic background. We find that actively-managed global REMF industry, as a whole, fails to beat the global real estate market and stock market, even before the consideration of expenses. It also cannot beat the U.S. domestic real estate and stock market either, implying no significant benefit from globalization. Same inferences hold when the combined benchmark models are employed. At the individual global REMF level, we find limited evidence for the existence of skills and diversification benefits for any global REMFs. These results hold after a series of robustness checks are implemented. Implementing cross-sectional bootstrap provides new insight to the conventional tests based on asymptotic assumptions. Based on simulated empirical cross-sectional distribution, we find fewer fund managers are truly skilled, and most of the outperformed funds are merely lucky.