基于VaR-GARCH模型的中国商品期货动态保证金水平的设定

本文通过对中国大连商品交易所以及上海期货交易所黄豆、铜两种主要期货品种收益率的研究,论证了其时间序列存在ARCH效应;借助VaR风险价值法,运用GARCH模型,建立VaR-GARCH模型,计算出两合约每个交易日VaR风险预测值;计算黄豆0809合约和铜0809合约日涨跌停的绝对值,将其与VaR预测值进行比较,设定保证金水平,确定保证金率并与交易所规定的保证金率进行比较,为交易所制定更合理的期货保证金率提供依据。

China commodity future margin level's dynamic set basing on VaR-CARCH mode

By studying the yields of soybeans futures in China Dalian Commodity Exchange and copper futures in Shanghai Futures Exchange,this paper demonstrated the existence of ARCH effect on the time series of yields;With the VaR method and GARCH model,the paper also established a VaR-GARCH model to calculate the VaR forecast value of the two contracts on each trading day;Further more,after comparing the absolute values of daily risedecline range of the 0809 soybean contracts and 0809 copper contract and VaR values,the paper set the lever of contract margin and determined the rate compared with that in the Exchange to provide a basis when developing a more reasonable margin rate.