VaR And CVaR Methods For Risk Evaluation Of Portfolio And Empirical Study

Financial risk management is an essential problem in practice, academia and supervise department. VaR and CVaR have become standardized means of financial risk evaluation currently. In this paper, VaR and CVaR for portfolio were researched by Historical simulation, Monte Carlo simulation and optimization methods.The introduct in this paper generalized the background of VaR and CVaR. In the second part, it was expatiated that the methods of calculating VaR for portfolios including option by Historical simulation and Monte Carlo simulation and two empirical examples were given. In the third part, three problems were researched: Firstly, the Mean-CVaR model was studied on the assumption that the interests of risk securities obeyed normal distribution and was compared with classical Mean- variance model. Efficient frontier and economic meanings were given. Secondly, risk-free asset was added to Mean-CVaR model under the normal distribution condition. It was found that efficient frontier for Mean-CVaR model with risk-free asset. Thirdly, the Mean-CVaR model was researched on the assumption that the interests of risk securities obeyed Laplace distribution. Efficient frontier and economic meanings were also given. At last, three empirical examples were given.