The Cvar Risk Measure In Portfolio Optimization In Applied Research

The aims of Portfolios theory is to attain the Portfolios of the maximum of the investment's return with the given value of the risk of Portfolios or of the minimum of investment's risk with the given level of the investment's return. Therefore, the measurement of the risk and return is the core. Compared with the measurement of return, the measurement of the risk is more important and complicated.At first, based on the Portfolios theory, expatiate on The definitions, calculation and characters of CVaR(Conditional-Value-at-Risk). At the same time, with the standards of the coherent risk measure, this paper compares the method of CVaR with the method of VaR and variance, and makes a conclusion that CVaR is better than others.Then, this paper establishes mean-CVaR model, analyses it's efficient frontier and compares it with mean- variance model to make a conclusion that both of them have the same result under normal distribution of the return condition.And then, this paper analyses the effect of the parameters and constraints to the mean-CVaR model, including, believe degree and transaction cost. This paper also, establishes a simple dynamic CVaR model to study the decisions of the long time investment.At last, this paper chooses a portfolio including 8 stocks from china's stocks market to make empirical research. The statistical character is that each of their return is not a normal distribution. The results of empirical research show that the results of mean-CVaR model and mean-variance model are different. The condition of normal distribution is not necessary for mean-CVaR model. So it is very fit for China's stocks market. The results of empirical research support the importance of believe degree and the fact that transaction cost makes lower returns.