The Performance Evaluation Of Our Country’s Open-End Funds Based On Var And Cvar

With the development of the open-end funds, open-end funds have become the mainstream investment. Investing the open-end funds has the advantage of tax avoidance, so more and more investors like them. Funds have a vide range of products and our country has hundreds of open-end funds at present. In the face of the various open-end funds, performance evaluation of funds is important to make a reasonable decision for investors. Many investors think that funds are deceptious, because there is still some deviation on the understanding of the funds. They think that investing funds is a risk-free investment, but any financial product is risky, just the size of risk different. Otherwise, our country’s fund performance evaluation indexes have an excessive emphasis on fund’s net value. They pay attention to the returns of fund and ignore the risk. This article introduces the performance evaluation model—RAROC model, which evolues by the traditional Sharpe ratio. This model has such a comprehensive consideration of returns and risk that fund evaluation becomes more objective and reasonable. At the same time, this article improves the method of risk measurement in the RAROC model so as to make the model more precise.Firstly, this article reviews the main theory and research achievements about fund’s performance evaluation at home and abroad, puts forward replacing VaR with CVaR as the risk measurement in the RAROC model and makes the fund’s performance evaluation based on that. Secondly, this article introduces the theory of VaR and CVaR, based on it, choosing8equity funds,8hybrid funds,8bond funds to make an empirical analysis. All the sample funds use the parameter model GARCH, EGARCH, TARCH which make an assumption for the T distribution and GED distribution, and extreme value model POT which is based on kurtosis method to estimate the value of VaR and CVaR.The empirical results show that the fitting effect of GARCH model at the assumption of GED distribution is the best and the POT model based on kurtosis method can well fit the tail distribution. Through the comparison of loss function raised by Lopez(1998), GARCH model at the assumption of GED distribution is superior to POT model when estimating the value of VaR, while POT model is superior to GARCH-GED model when estimating the value of CVaR through the comparison of ELC statistic. In the end, after the model to compare, choose the best result of risk measurement to make the performance evaluation.