基于R Vine Copula的VaR模型期货市场的风险度量  

Risk measurement of VaR mode futures market based on R-Vine Copula

作　　者：陈金图 刘武强 CHEN Jintu;LIU Wuqiang(Minnan Science and Technology College,Quanzhou 362332 China)

机构地区：[1]闽南科技学院商学院,福建泉州362332

出　　处：《新余学院学报》2024年第4期42-51,共10页Journal of Xinyu University

基　　金：福建省教育厅中青年科研项目“基于数字技术的绿色金融创新推动福建‘双碳’目标实现路径研究”(JAS22221)。

摘　　要：按照传统投资组合的观点,投资者通过投资于低相关性的不同资产可以起到分散风险的作用,资产与资产之间的关系一般以线性相关性来衡量。然而,不同资产之间的关系并非纯粹的线性相关,现实中不同资产之间具有不同的线性相关,但又往往在同一个时间点发生极端的风险损失。构建基于R Vine Copula的VaR风险度量模型,采用Copula模型获得了不同期货资产收益率的相依结构,测算出不同期货资产之间的尾部相依系数,并计算出这些期货资产之间的联合分布和条件分布,在此基础上对各资产在条件相依结构下的VaR进行估计,最后对这些期货资产的VaR进行Kupiec检验,检验结果验证了该模型估计风险的准确性和有效性。According to the traditional view of investment portfolios,investors can diversify their risk by investing in different assets with low correlation,and the relationship between assets is generally measured by linear correlation.However,the relationship between different assets is not purely linear.In reality,different assets have different linear correlations,but extreme risk losses often occur at the same time point.This paper constructs a VaR risk measurement model based on R-Vine Copula.The Copula model is used to obtain the dependent structure of the rate of return of different futures assets,calculate the tail dependency coefficient between different futures assets and the joint distribution and conditional distribution between these futures assets.On this basis,the VaR of each asset under the conditional dependent structure is estimated.Finally,this paper conducts Kupiec test on the VaR of these futures assets,and the test results verify the accuracy and effectiveness of the risk estimation model.

关 键 词：Vine Copula VAR模型 期货市场 返回检验 风险度量 

分 类 号：F724.5[经济管理—产业经济] F224