Coherent Risk Management and Its Portfolio Optimization Model

In market risk measurement field, the return-loss distribution exist the severe phenomenon of excess kurtosis and heavy tail. In the meanwhile, method of value at risk itself cannot correspond with coherent risk management system, because VaR can not obey with its sub-additivity rule, which make local optimal be the whole optimal when selecting the optimal portfolio. In order to resolve these problems, proceeding from the theory for coherent risk measurement, we put forward a new technique of risk measure-coherent value at risk (CVaR)-to measure market risk of portfolio, on which we build portfolio optimization model of CVaR and select the optimal portfolio with linear programming. In our empirical research, we use random data of Shanghai Security market from Jan. 1st, 2007 to May 30^th, 2008. We conclude that the result based on optimal model of CVaR is better than that of on optimal model of value at risk.