基于Hedonic分位数模型对城市住宅价格差异性的研究——以深圳市为例

The study on the difference of urban housing price based on Hedonic quantile regression model: Taking Shenzhen as an example

城市住宅是一种典型的异质性商品,住宅价格由其特征带给人们的效用所决定.对于异质性商品,本文基于2016年深圳市商品房成交的样本数据,以Hedonic模型的相关理论为基础,分别应用逐步回归算法、Lasso惩罚构建对数住宅价格估计模型进行研究,结果表明,对深圳住宅价格影响最大的因素为住宅的绿化率;最后分别以对数住宅价格的5%、25%、50%、75%、95%分位点构建Hedonic分位数回归模型,发现在不同的住宅价格分位点上,显著特征变量的系数发生较大的变化,体现出在住宅不同价位上人们对住宅所具有的特征偏好不同.

Urban housing is a typical heterogeneous commodity,whose price is determined by the utility of all characteristics.For heterogeneous commodities,this paper is based on the sample data of commercial housing transactions in Shenzhen in 2016.Based on the related theory of Hedonic model,we use stepwise regression algorithm,Lasso penalty algorithm respectively to build housing price model.The results show that the most influential factor on the housing price of Shenzhen is the greening rate.Finally,the housing price quantile regression model is constructed with the 5%,25%,50%,75%and 95%points of the housing price in Shenzhen.It is found that the coefficients of significant characteristic variables vary greatly at different housing price points,which is reflected in the different preferences of people on residential properties at different housing prices.