基于高阶矩波动和Copula函数的相依性模型及其应用  

Dependence Model and Its Application Based on Higher Moment Volatility and Copula

作　　者：易文德[1] 

机构地区：[1]重庆文理学院数学与统计学院,重庆402160

出　　处：《管理评论》2012年第1期58-66,共9页Management Review

基　　金：国家自然科学基金项目(71071131);教育部人文社会科学研究项目(11XJC790004);重庆市教委科学技术研究项目(KJ111211)

摘　　要：本文提出了基于高阶矩波动的相依结构模型:Copula-NAGARCHSK-M模型。考虑资产的时变条件方差风险、条件偏度风险和条件峰度风险对边缘分布的影响,应用模型研究了上证综指和深证成指对数收益率之间、条件方差之间、条件偏度之间和条件峰度之间的相依结构。发现两股票市场的指数对数收益率之间、条件方差之间和条件峰度之间有相似的相依结构,而条件偏度之间的相依结构则是负方向的相似。A dependence structure model based on higher moment volatility, Copula-NAGARCHSK-M model, is proposed in this paper. The model is used to investigate the dependence structure of logarithmic returns, conditional variances, conditional skewnesses and conditional kurtosises between Shanghai and Shenzhen stock markets based on the consideration that the time-varying conditional variance risk, skewness risk and kurtosis impact on the marginal distributions. The evidences show that the dependence structures of returns, variances and kurtosises between both stock markets are similar, but the dependence structure between skewnesses is similar on the negative directional dependence.

关 键 词：高阶矩 COPULA函数 Copula-NAGARCHSK—M模型 高阶矩相依 

分 类 号：F224[经济管理—国民经济] F832.51