中文摘要：

文章以华夏沪深300ETF联接为例,对开放式基金的风险度量进行了研究。选取2009年至2015年的每日单位净值,以对数收益率为样本数据。该基金的收益率具有尖峰厚尾性、波动聚集性。建立能消除ARCH效应的GARCH模型,引入适用于厚尾性的t分布和GED分布对样本数据拟合。预测各模型的条件均值和条件方差,最后代入到VaR公式中计算华夏沪深300ETF联接在不同分布下的风险值。应用Kupiec失败率检验法对基于各模型计算的VaR值的准确性进行检验,结果显示t分布下的GARCH-M模型能较好地拟合该基金并有效地度量风险。

英文摘要：

300ETF is taken as an example to study the risk measurement of Open-end Fund in this article. Selecting the dai- ly price from 2009 to 2015 and takeing its yields as sample data, the 300ETF yields has the feature of fat tails, volatility clustering. So, GARCH models which could eliminate the ARCH effect and introduce t-distribution and GED-distribution which fitting the fat tails is chosen as the model. Forecasting the conditional variance and conditional mean of each model and then use the VaR formula to calculate the 300ETF values at risk in different distribution. Finally, testing the accuracy of VaR value on GARCH models by Kupiec test method. The results indicate that the calculated risk value by GARCH-M model on t-distribution can re- flect market risk better and evaluate the risk correctly.