期刊文献+

基于Copula房地产与金融行业的股票相关性研究 被引量:40

Investigating the Dependence Structure between Real Estate and Finance Industries Based on the Copula Function
下载PDF
导出
摘要 房地产市场和金融市场的关系多数文献从相互影响的角度研究,而本文以房地产与金融行业的股票收益率数据,引入Copula方法定量刻画两个行业股票之间的相关关系。实证结果表明,双参数结构的Copula拟合度普遍高于单参数结构的Copula;如果仅考虑单参数或双参数Copula函数结构,可能会得出错误的结论。本文同时选取单参数和双参数的Copula拟合房地产与金融行业的股票收益率之间相关性结构,通过AIC、BIC最小原则应为BB3 Copula,说明两个行业股票在市场低迷时期的尾部相关性高于活跃时期的尾部相关性。因此,在投资股票时,不能通过对这两个行业股票的组合投资降低风险。 Real estate markets and financial markets play essential roles for the continuous growth of developing economies.The development of these two markets has an interrelated influence on one other.Recent studies show that real estate and financial markets are linearly correlated with each other.However,the nonlinear Graner causality between these two markets does not exist.Most of these studies are qualitative and have paid little attention to using stock data to analyze the dependence structure between the real estate and financial industries.This paper applies the Copulas theory to investigate the dependence structure between real estate and financial industries.The Copula theory can capture nonlinear,asymmetric and tail dependence.Our analysis is based on the stock database of Tsinghua Financial Data from July 2,2001 to August 5,2009.We are able to use 1,953 pairs of data.Univariate distribution method is based on a symmeparametric extreme value model.EV Copulas,Archimedean Copulas and Archimax Copulas simulate the correlation between real estate and financial markets.The first part of this study estimates the parameters of the Copula function.We adopt the margin inference method by estimating the parameters for the univariate marginal distribution and for the Copula function.The computation of the Kurtosis method shows that the threshold values are the same for both the univariate distribution and the Generalized Pareto distribution(GPD).Diagnostic plots of the GPD fit are Excess Distribution,Tail of Underlying Distribution,Scatter Plot of Residuals and Record Development.These plots indicate that GPD appears to fit the distribution of threshold excess fairly well.The second part discusses the rationale of selecting the Coupla function.Based on AIC and BIC minimum theories,the Gumbel Copula function shows that a correlation between these two markets exists in only upper tail for single parameter Copulas.However,the BB3 Copula has a higher correlation in the lower tail than the upper tail for a variety of parameters used in the Copula function.These findings illustrate that two different stock returns are more likely to correlate with each other during market downturns than upturns.These findings need careful interpretations if selected Copulas with one or two parameters need to be simulated simultaneously.In summary,the dependence patterns between the financial time series vary with the Copula functions.The GPD model needs to estimate the threshold values in order to exactly fit the margin distribution of Copula functions.In addition,creating an investment portfolio of two different stocks won't necessarily help reduce investment risk.
出处 《管理工程学报》 CSSCI 北大核心 2011年第1期165-169,164,共6页 Journal of Industrial Engineering and Engineering Management
基金 国家杰出青年科学基金资助项目(70525005)
关键词 COPULA 峰度法 GPD 尾部相关系数 copulas kurtosis method GPD coefficient of tail dependence
  • 相关文献

参考文献21

二级参考文献50

共引文献339

同被引文献359

引证文献40

二级引证文献253

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部