基于贝叶斯模型平均（BMA）方法的中国房地产价格影响因素分析

对可能影响中国房价的诸多因素的重要性问题进行识别和检验，基于模型不确定性的视角，使用中国30个省（区）2002~2013年面板数据，采用贝叶斯模型平均（BMA）方法进行模型设定与分析。研究认为，在可能对中国房价产生影响的19个指标中，信贷政策、心理预期、物价水平、房屋竣工面积和产业结构合理化等5个解释变量的后验概率大于90%，它们是影响现阶段中国房地产价格的决定因素；应通过差别化的信贷政策分区域控制房价，通过新闻媒体公开统计和发布房地产数据正确引导人们的心理预期，通过适宜的货币政策有效控制物价，通过保障性住房建设增加房地产供给，通过合理化的产业结构引导房价调控等，促进中国房地产市场的健康发展。

Analysis of determinants of real estate prices in China based on Bayesian Model Averaging(BMA)

This paper identified and checked the determinants of real estate prices in China, and adopted thirty provinces (autonomous regions) panel data from 2002 to 2013 to do model specification and analysis by use of bayesian model averaging method from the perspective of model uncertainty. The results show that in the nineteen factors of five levels that may have an impact on Chinas real estate prices, the posterior probability of five variables is more than 90%, which are credit policy, psychological expectations, price levels, completion of housing area, and the rationalization of the industrial structure. They are the determinants of China’s real estate prices at the present stage. Therefore, in order to promote the good development of China’s real estate market, differential credit policy should be adopted to control real estate price according to different zones, the real estate data should be collected publicly and published via the media to lead masses’ psychological expectation, appropriate monetary policy should be made to control prices levels effectively, indemnificatory housing should be increased to meet the real estate supply, and industrial structure rationalization should be improved to adjust housing price. 