银行业如何影响房地产业?Copula分位数回归及预测方法  

How Does Banking Industry Affect Real Estate? Copula Quantile Regression and Predicting Approach

作　　者：田茂茜[1,2] 虞克明[2,3] 

机构地区：[1]西安交通大学金禾经济研究中心,陕西西安710049 [2]石河子大学统计与金融系,新疆五家渠831300 [3]英国布鲁奈尔大学统计系

出　　处：《数理统计与管理》2015年第1期150-161,共12页Journal of Applied Statistics and Management

基　　金：全国统计科学研究计划项目:基于众数统计量的收入分配差距测度及其影响机制研究(2013LY022);国家自然科学基金资助项目:贝叶斯离散分位数回归模型:理论;方法及应用(11261048);石河子大学中青年科研骨干培育基金资助项目:基于Copula分位数回归模型的金融风险测度及防范研究(RWSK11-Y08)

摘　　要：众所周知,房地产业与银行业是高度相关的,如何确定银行业股票收益率对房地产业股票收益率的影响以及如何根据银行业股票收益率预测房地产业股票收益率的波动是非常重要的问题。本文首先使用Copula分位数回归建立了银行业股票收益率对房地产业股票收益率的回归模型,并且给出了Copula分位数回归基础上的CopuIa选择新标准,即分位数损失函数距离意义下的Copula函数选择准则,依据该准则我们选取Clayton Copula分位数回归模型刻画了低迷时期银行业股票收益率如何影响房地产业股票收益率。It is well-known that Real Estate and Banking Industry are highly relevant.How to determine the effect of the Banking Industry stock returns on the Real Estate stock returns and how to predict the fluctuations of the Real Estate stock returns by the Banking Industry stock returns are very important issues.In this article,we firstly use Copula quantile regression model to describe the relationship between these two stock returns,and then propose a distance criterion based on the quantile loss function for Copula selection.In order to investigate the dependent structure of these two stock returns,the appropriate Clayton Copula quantile regression model is selected by the new distance criterion.Finally,we predict the volatility of the Real Estate stock returns by the Banking Industry stock returns based on the Clayton Copula quantile regression model.

关 键 词：COPULA 非对称LAPLACE分布 分位数回归 分位数损失函数 

分 类 号：O212[理学—概率论与数理统计]