This research project examines the investment performance of Hong Kong?s real estate and property stocks over a period of nineteen years from 1984 to 2002. Evaluations using time-varying Jensen index adjust the asset return comparison for possible time-varying investment risk (return variation). The return and risk comparison for direct and indirect real estate investment is further investigated with an alternative GARCH-M model in capturing the possible time-varying risk premium. Unlike previous studies, one of the major contributions of this research is to measure the risk-adjusted real estate returns by taking into account the impact of illiquidity, management and transaction costs, and vacancy risk with actual transaction data. The time-varying models have not been applied for analysing the returns on property investment in Hong Kong so far. The results of this study help resolving the puzzle why real estate offers superior investment performance as stated in the existing literature that is inconsistent with the Efficient Market Hypothesis. With arbitrage, the capital theory predicts that all investments should display similar risk adjusted return in the long run. As a result, it is impossible for direct property to earn abnormal risk adjusted return in the long run. The empirical results show that superiority of real estate property in risk-adjusted return is reduced or even disappeared in the period of 1991 to 2002 when the impact of illiquidity and the 1997 Asian Financial Crisis are taken into account. This conclusion is further strengthened with the favourable empirical evidence obtained when all the unsystematic risks including the management and transaction costs, and vacancy risk are incorporated into the analysis. The superior investment performance for real estate relative to property stocks found in the literature is too good to be true. Its existence is simply due to the omission of fundamentals factors like unsystematic risks in the analysis. The impact of liquidity on return assessment has been under-researched, this study attempts to fill the gap by quantifying the impact of liquidity explicitly in explaining the excess return of real estate investment. The empirical result is consistent with the intuition that investors require compensation in holding illiquid assets.