An econometric analysis of the Hong Kong real estate market : hedonic price study and index number construction

Abstract

The Hong Kong housing market is an indispensable component in Hong Kong financial and social systems. Correctly gauging property price movements is crucial for the decision-making of investors, homebuyers and policymakers. The goal of this thesis is to improve hedonic price modelling and index number construction in the Hong Kong housing market using econometrics approaches. This objective has been achieved by addressing the omitted variable and the model misspecification problems of hedonic price modelling. Three approaches are adopted, namely, the Stein variance Double k-class estimator, the hybrid model of hedonic price and repeat sales methods, and the smoothing spline semiparametric procedure. The Stein variance Double k-class estimator ameliorates the omitted variable problems by incoporating researchers’ experience and expertise in hedonic price modelling. Compared with traditional OLS approach, Double k-class estimators have more accurate predictive mean squared error terms and more precise parameter estimates. The hybrid model combines repeat sales and hedonic price methods. Both numerical and empirical findings suggest that the proposed hybrid model effectively improves hedonic price modelling when important variables are excluded from the model, at least for the cases that have been considered in the Monte Carlo experiment design. The smoothing spline semiparametric technique is found to be much less sensitive to omitted variable problems than the OLS method. Significant evidence of a non-linear relationship between property price and certain housing traits such as floor level and gross area is observed based on the coefficient estimates of the smoothing spline semiparametric method. An overview of the Hong Kong residential property market, including discussions on the land supply system, public housing policy, residential property market, and an important property price index in Hong Kong, is also given in this thesis, as all the empirical evidence in this study is obtained using data from the Hong Kong real estate market.