Portfolio VaR Estimating And Portfolio Optimizing Based On MRS Copula-ARJI-GARCH Model

As known to all, diversified investment has been the main means to avoid risk, soportfolio risk management has been valued more and became a hot issue under therealistic background such as international financial structure changing and potential riskgradually increasing. While the measurement of portfolio risk, the construction of jointdistribution and the discription of dependency structure between the assets are the coreelements of that hot issue. In view of the deficiencies which appeared in the threeelements setting in the available researches, synthetically considering asymmetricdependence between assets and jump dynamics in asset returns, this paper develops aMRS Copula-ARJI-GARCH model, and conducts a study on portfolio VaR estimatingand portfolio optimization based on this model.Firstly, this paper analyzes the advantage of VaR method in risk measuring, MRSCopula in describing dependency structure between assets and ARJI-GARCH model infitting the marginal distribution. Secondly, a MRS Copula-ARJI-GARCH model isconstructed and used to get VaR of industrial stock index portfolio through Monte Carlomethod. Then comparing its estimating performance with MRS Copula-GARCH-tmodel, MRS Copula-GARCH-n model, dynamic Copula-GARCH-t model and dynamicCopula-GARCH-n model based on risk control chart, violation test and intervalprediction test, which qualitatively and quantitatively explain the model we constructcan improve the accuracy of VaR estimating. Finally, mean-VaR portfolio models areconstructed and used to optimize portfolio based on MRS Copula-ARJI-GARCH modelrespectively under risk minimization strategy, income maximization strategy and utilitymaximization strategy.The empirical research shows that: MRS Copula-ARJI-GARCH model cancomprehensively reflect the probability of extreme returns, which not only makes itoptimal in estimating portfolio VaR, but also helps investors to create cautious decisionsin asset allocation, including realizing risk minimization when given the yieldsdemanded, realizing revenue maximization when crises are definite and achievingutility maximization when consider the benefits and risks together.