中文摘要：

描述了金融时间序列的一般特性，从收益的波动性与分布出发，组建起计算时变风险价值的VaR-GARCH模型族，并应用该模型族在3种分布假设下对我国铜期货的市场风险进行了实证分析．研究结果表明，基于广义误差分布的VaR-EGARCH模型能很好地刻画期铜收益的尖峰厚尾性与市场风险．同时，对期铜市场风险的变动趋势进行了详细剖析．

英文摘要：

This paper describes the general characteristics of the financial time series and constructs the VaR-GARCH models for calculating time-varying value at risk based on the volatility and distributions of returns. The market risk of copper futures in China is measured empirically by these models on three distributions. The results show that the VaR-EGARCH model based on the general error distribution （GED） can accurately describe high peaks and fat tails of returns and market risk in Chinese copper futures. In addition, the market risk trend of copper futures is analyzed in detail.