Semiparametric Hedonic Price Index In The Preparation Of Cell Phone Applications

With the development of mobile telephone market in china in recent years, mobile telephone plays an increasingly important role in people's daily life. In most references, the price quantify effect analyses are very few. Almost person does not work out the mobile telephone price index. Since the mobile telephone is a heterogeneous goods, we can not use the traditional method of compiling the price index for mobile telephone. According to the hedonic price theory of western country, I make use of the hedonic price model to analyze characteristics of the various features of variable effects on prices. At the same time, in accordance with the principle of optimal, we make use of semiparametric regression to select the best model and compile the price index.First of all, we review the references of hedonic price theory and semiparametric regression of domestic and foreign. So we understand characteristics of hedonic price method and semiparametric regression method of the status quo. Secondly, we introduce the characteristics of hedonic price and its theoretical basis -Lancaster's theory of hedonic consumption and Rosen's theory of supply and demand equilibrium model in detail. Immediately we introduce estimation methods of nonparametric regression and semiparametric regression, the choice of kernel function and bandwidth, and so on. An Empirical Analysis of the paper lay a solid theoretical basis.We select a total of 388 samples of data of variable and price information of mobile telephone from December 2007 to April 2008. We establish general linear regression model and semiparametric regression model, and understand semiparametric regression model's fitted effect is far superior to the general linear regression model, through hausman testing and comparison of R~2. According to the semiparametric regression which be selected, we analyze the various features of mobile phones on the effect of the price and find that most of the variables selected on the impact of price changes are significant.According to semiparametric regression model, we produced the mobile phone hedonic price index of combined data and find that the phone is the price index declined month by month.