摘要
目的在金融市场多元化的背景下,研究各市场间的相关性对风险控制及投资组合问题有着巨大的帮助,构建GARCH Pair-Copula模型研究标普500指数、道琼斯工业指数、恒生指数、日经225指数、沪深300指数之间的相依关系。方法选取5个股指的日收益率序列,利用AR(1)-GARCH(1,1)-偏斜t模型和AR(1)-GJR(1,1)-偏斜t模型拟合其分布,用IFM方法估计参数,利用AIC和BIC准则选取最优的模型。结果在C藤和D藤结构下,选取Clayton Pair-Copula,t Pair-Copula,SJC Pair-Copula刻画市场间的相依关系。结论基于C藤的t Pair-Copula能较好地刻画市场间尾部对称的相依结构。
Purposes—To study the dependence of S&P 500 index,Dow Jones industrial index,Hang Seng index,Nikkei 225 index and Shanghai-Shenzhen 300 index by constructing GARCH Pair-Copula model because the study on the correlation between different markets under the background of diversification of financial markets is helpful for risk control and portfolio problems.Methods—The AR(1)-GARCH(1,1)-skewed t model and AR(1)-GJR(1,1)-skewed t model were used to fit the distribution of the selected daily return series of five indexes and the parameters were estimated with IFM method,meanwhile,AIC and BIC criteria were employed to select the optimal model.Result—Under the structure of C-vine and D-vine,Clayton Pair-Copula,t Pair-Copula,SJC Pair-Copula are selected to describe the dependent relationship between the markets.Conclusion—The results show that t Pair-Copula based on C-vine can depict the symmetric tail dependence structure between markets.
作者
张转转
卢俊香
ZHANG Zhuan-zhuan;LU Jun-xiang(School of Science, Xi’an Polytechnic University, Xi’an 710061, Shaanxi, China;Faculty of Economics and Management, Xi’an University of Technology, Xi’an 710048, Shaanxi, China)
出处
《宝鸡文理学院学报（自然科学版）》
CAS
2020年第1期7-13,共7页
Journal of Baoji University of Arts and Sciences(Natural Science Edition)
基金
国家自然科学基金项目(11601410)
陕西省自然科学基金项目(2017JM1007)
中国博士后科学基金(2017M613169)。
