摘要
随着人民币国际化进程的逐步推进,国际外汇市场的风险性和相依性研究尤为重要.Copula函数作为一个将多元分布建模分解成边缘分布和联合分布的相依结构模型,可解决分布模型难以确定的问题.基于五种货币(即美元、日元、欧元、英镑和港币)日收益数据构建基于AR(2)-GARCH(1,1)-t模型的各个外汇日收益率的边际分布,利用蒙特卡洛方法度量正态Copula函数和t-Copula函数下的投资组合Va R值,为解决外汇市场资产风险问题提供良好的操作与管理决策.
With the acceleration of RMB＇s globalization, it is particularly important to research the risks and the interdependence of the international foreign exchange market. Copulas can be used to decompose multivariate joint distribution into marginal distribution and correlation structure. In this paper we model the actual and joint distribution of daily exchange rate returns of five major currencies,i.e. USD, JPY, EUR, GBP and HKD against RMB mainly based on AR（2）-GARCH（1,1）-t model. The advantage is applied to compute the Va R of a portfolio by Normal and Student＇t Copula function.Furthermore,a new model for portfolio choice based on copulas is proposed by Monte Carlo simulation,and empirical analysis is operated in the hope of helping make a strategic decision in the foreign exchange market.
出处
《苏州市职业大学学报》
2016年第1期40-45,共6页
Journal of Suzhou Vocational University
基金
江苏省高校哲学社会科学研究基金资助项目(2014SJD607)
苏州市软科学研究资助项目(2014JX54)
