The Empirical Research On The Factors Of China’s Real Estate Prices Based On Bayesian Model Averaging Method

In recent years, China’s real estate market has experienced an unprecedented rapid development stage and real estate has really become an important industry in the national economy. While in the course of the continued prosperity of the real estate market, real estate prices increased, to some extent, beyond the affordability of ordinary families. High prices have brought a number of adverse effects on the healthy development of China’s economy and the stability of people’s life. Real estate prices’ continued rising has become the focus of academics and practitioners. What factors affect real estate prices by exactly is not only relevant, but also very urgent. However, domestic scholars used different methods of research and the reason why they got different findings was that these empirical studies were relying on a single model, which was easy to be interfered by the model set and the selecting of variables, then ignored the problem of model uncertainty and this would make us not able to recognize the importance of each factor on real estate prices comprehensively, accurately and effectively. Presently, there is a major ignorance in the problem of model uncertainty of the research on the factors of China’s real estate prices and this dissertation tries to contribute to this area.The object of this paper is to provide an empirical research on factors of China’s real estate prices based on the model uncertainty. In order to understand the selected variables and selected econometric method in empirical research, this paper at first reviews the theoretical and empirical literature examining the factors of China’s real estate prices and also combs the related research literatures of bayesian model averaging method. Then, to highlight the necessity of empirical research, the status of China’s real estate prices have been analyzed. At last, this paper abandons the tradition of artificial pre-set model and variables in empirical research on factors of China’s real estate prices and it uses BMA method that can solve the problem of model uncertainty, adopts thirty provinces(autonomous regions) panel data from 2002 to 2013, classifies and sorts many factors that may affect our prices according to their importance, then this paper proves the effectiveness of the BMA in solving the problem of model uncertainty by two ways. Also, several possible interpretations for these effective empirical results have been provided.The empirical results of the study show that the use of BMA method solves the problem of model uncertainty effectively and make the conclusions scientifical and reliable.It has been found that in the ninteen factors of five levels that may have an impact on China’s real estate prices, there are fourteen variables’ s posterior probability is more than 20%, which are population density, unemployment rate, consumer income levels, the income gap of demand level, completion of housing area, real estate investment, land prices, construction costs of supply level, price levels, the rationalization of the industrial structure of economic level, credit policy, tax policy, land policy of policy level and psychological expectations of other levels. They are the important factors of China’s real estate prices at the present stage. There are five variables’ s posterior probability is lower than 20%, which are economic growth, population aging, upgrading the industrial structure, urbanization rate and interest rate policy, indicating that they still can not be the dominant factors of China’s real estate prices at present. The findings are also compatible with different or even opposite conclusions drawn from numerous previous studies of Chinese scholars who used single model. Moreover, although China is under the background of population aging and urbanization progress, this paper does not find evidence that population aging and urbanization have an important impact on China’s housing prices.