应用Copula理论的人民币汇率风险测度  

Measuring the Risks of Chinese Exchange Rate Portfolio Based on Copular Method

作　　者：郑冀[1] 何宏[1] 蒋超 杨振东[1] 

机构地区：[1]后勤工程学院,重庆401311 [2]猪八戒威玛科技有限公司,重庆401120

出　　处：《重庆理工大学学报（自然科学）》2013年第7期134-139,共6页Journal of Chongqing University of Technology：Natural Science

摘　　要：借助Copula理论,同时配合GARCH类模型,建立了一个能最为理想地描述汇率组合之间相关结构的Gumbel-Copula-PARCH(1,1)-t模型。借助Monte Carlo仿真技术,于不同置信度下求出能够度量汇率组合损失情况的VaR值。通过研究"失败率"来评价应用MonteCarlo仿真技术的Gumbel-Copula-PARCH(1,1)-t模型所模拟计算出的VaR值的实际效果。This paper has introduced the theory of Copula and GARCH Models to establish a GumbelCopula-PARCH （1,1）-t model which is capable to describle the dependence structure of exchange rate portfolio the most perfectly. Then, it applied the emulation technique of Monte Carlo to calculate the VaR of exchange rate portfolio under different confidence levels. At last, we evaluate the actual effect of VaR, which is calculated by the Gumbel-Copula-PARCH （1,1）-t model with applying the emulation technique of Monte Carlo, by researching the so-called ＂the failure rate＂.

关 键 词：阿基米德COPULA GARCH类模型 VAR 

分 类 号：O22[理学—运筹学与控制论]