城市住房价格指数编制方法的拓展设计及实证检验  

An Extension in the Construction of Urban Housing Price Indexes and Its Application

作　　者：石薇[1] 李强[2] 王洪卫[1] 

机构地区：[1]上海金融学院 [2]新加坡国立大学房地产系及房地产研究中心

出　　处：《数量经济技术经济研究》2014年第12期127-141,共15页Journal of Quantitative & Technological Economics

基　　金：国家自然科学基金项目"快速城市化进程与住房公共政策:交互性与协调性发展研究"(NSF71173045)的资助

摘　　要：科学监测城市房价走势,在当前环境下尤为重要。为拓展国际通行方法编制国内单一城市房价指数的适用性,引入样本匹配重复交易法构建房价指数,以提高样本容量与可比性。基于上海数据的实证结果表明,相较于传统重复交易法和特征价格法,样本匹配重复交易法能更准确地反映住房价格变动,结果异常波动性更小,噪声影响程度更低,在克服样本代表性误差和变量缺失误差方面效果更显著,对编制国内城市房价指数具有较好的应用价值。Scientific monitoring of real estate prices in urban areas is essential, especially under current market conditions. In order to apply internationally accepted method- ology to the construction of domestic urban housing price index, this paper introduces the matching repeat sales methodology that improves comparability of sample sizes and sample selection. Our empirical results, based on Shanghai housing transaction data, show that the housing price index derived from the matching repeat sales methodology can reflect the housing price fluctuation more accurately compared with traditional repeat sales and hedonic methodologies. Specifically, the matching repeat sales price index has smaller abnormal volatility and less noise. It overcomes sample representative error and omitted variable problem. It can be widely applied to the construction of domestic urban housing price indexes for both new and second-hand homes.

关 键 词：住房价格 样本匹配 特征价格法 重复交易法 

分 类 号：C813[社会学—统计学]