新建住宅价格指数编制模型的优良性检验  

A Goodness Test of the Model for Compiling New Residential Price Index

作　　者：易佳 金升平[1] YI Jia;JIN Sheng-ping(College of Science,Wuhan University of Technology,Wuhan 430070,China)

机构地区：[1]武汉理工大学理学院,湖北武汉430070

出　　处：《数学的实践与认识》2022年第3期60-69,共10页Mathematics in Practice and Theory

摘　　要：研究房价指数编制模型的优良性检验方法可以指导房价指数编制工作,有助于在实际应用中选择更为科学准确的房价指数.从统计指数偏误的角度来评价房价指数的优良性,并给出了检验房价指数是否有型偏误的具体方法.对BMN重复交易模型和标普Case-Shiller模型从理论推导角度进行时间互换检验,发现前者没有型偏误而后者有型偏误.从模型计算角度对两个新建住宅价格指数编制模型进行实证分析,发现对数匹配模型没有型偏误,贴现匹配模型有型偏误.The study of the goodness of the house price index compilation model can guide the compilation of house price indexes and help to select more scientific and accurate house price indexes in practical applications.This paper evaluates the goodness of house price indices from the perspective of statistical index bias,and gives specific methods to test whether house price indices have type bias.By performing the time reversal test on the BMN repeated sales model and the S&P Case-Shiller model from the perspective of theoretical derivation,we find that the former has no type bias while the latter has type bias.By empirically analyzing the two new residential price indexing models from the perspective of model computation,we find that the logit-matched model has no type bias and the discounted-matched model has type bias.

关 键 词：房价指数 指数偏误 时间互换检验 优良性检验 

分 类 号：F299.23[经济管理—国民经济] F224