Driving Forces Analysis for Residential Housing Price in Beijing

AbstractPrevious research showed that various factors could influence the housing market. In this paper, hedonic pricing method was employed to analyze the effects of structural variables, including land transaction price, the distance to downtown area, central business district, railway station and hospital, floor area ratio (FAR), number of bus lines nearby and dichotomous variables, including nearness to rail transit, recreational facilities and parks which reflects the accessibility and living conditions, on housing transaction price. Hedonic pricing models including linear and semi-logarithm regression model were constructed. Results showed that the semi-logarithm model had relatively stronger explanatory power than linear model. The main determinants of housing transaction price in Beijing city were land transaction price, FAR and the distance between housing to downtown area. Among which, transaction price of located land had notably raised housing transaction price, contributing 98.8% to the selling price. FAR and distance from housing to the downtown area were the main negative driving forces for housing transaction price. Compared with structural variables, though correlation analysis indicated that nearness to rail transit and existence of recreational facilities had significant positive correlation with housing transaction price, it was not demonstrated in the regression results. In this study, wavelet-based denoising method was tentatively employed in pretreating data for semi-logarithmic models, and result suggested that the explanatory power of semi-logarithm regression was enhanced