中文摘要：

文章以扰动项为正态分布的GARCH模型研究了Brent、Dubai和Minas原油市场的价格波动风险,以Kupiec的失败频率检验法检验模型的有效性,结果表明,模型能够刻画考察期的原油收益率波动特征,也能较好地度量上述3个原油市场的价格风险。由于中国成品油定价过程中参照Brent、Dubai和Minas原油市场,文章对3个原油市场的历史数据拟合结果进行了相关性分析,通过组合波动最小原理得出3地市场的权重比为0.238 9∶0.575 9∶0.185 2时,中国成品油用油成本的波动率达到最小值。通过蒙特卡罗模拟验证了这一结果,实证表明在75美元/桶的国际油价水平下,置信水平取95%,使用最优权重能够减少用油成本波动0.16~0.17美元/桶。

英文摘要：

The price risk of Brent,Dubai and Minas Crude Oil market with GARCH model at Normal distribution is analyzed in this paper,and the results are examined by using Kupiec test.The price volatility is significantly described;price risks are measured as well.Because China sets its refined oil price according to Brent,Dubai and Minas Crude oil market,the correlation between the three markets are analyzed based on their historical data.The results present that the weights over the three markets at 0.2389： 0.5759： 0.1852 would be minimized oil cost volatility in China＇s refined oil market.The results are tested by Monte Carlo simulation,too.Giving the international oil price at 75 dollars/barrel and taking 95% confidence level,China＇s oil cost volatility could be reduced 0.16-0.17 dollars/barrel in the method mentioned in this paper compared with common empirical methods.