基于Copula-GARCH模型的外汇风险研究  

Research on foreign exchange risk based on Copula-GARCH model

作　　者：党聪 卢俊香[1,2] DANG Cong;LU Jun-xiang(School of Science, Xi'an Polytechnic University, Xi'an 710048, Shaanxi, China;School of Economics and Management, Xi'an University of Technology, Xi'an 710048, Shaanxi, China)

机构地区：[1]西安工程大学理学院,陕西西安710048 [2]西安理工大学经济与管理学院,陕西西安710048

出　　处：《宝鸡文理学院学报（自然科学版）》2020年第4期10-14,19,共6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);陕西省自然科学基础研究项目(2017JM1007);中国博士后科学基金(2017M613169)。

摘　　要：目的构建Copula-GARCH模型,研究外汇市场之间的风险。方法选取美元和欧元兑人民币收益率序列,利用GARCH模型拟合其边缘分布,选择合适的Copula函数刻画两市场间的相关性,通过蒙特卡洛模拟计算VaR值。结果两者收益率之间不存在单一的正负相关性,且尾部渐进独立,VaR序列基本可以覆盖实际损失值。结论通过Copula函数拟合的VaR序列对金融风险有一定的指导意义。Purposes—To construct a Copula-GARCH model for the study of the risks between foreign exchange markets.Methods—The sequence of the returns of RMB exchange rates against the euro and the US dollar are selected and their marginal distributions are fitted by using the GARCH model,the appropriate Copula function is chosen to characterize the correlation between the two markets and VaR value is calculated by Monte Carlo.Result—The results show that there is no single positive and negative correlation between the two returns,the tail is gradually independent and the VaR sequence can basically cover the actual loss value.Conclusion—The VaR sequence fitted by Copula function has some guiding significance for financial risk.

关 键 词：外汇 二元Copula函数 GARCH模型 VAR值 

分 类 号：F832.6[经济管理—金融学]