中文摘要：

分阶段定量研究国际石油价格并作出预测，为我国应对油价波动以确保石油安全及采购提供借鉴性指导．借助Eviews计量经济学软件，选取Brent油价为被解释变量，道琼斯商品指数，美国商业原油库存，美元指数，基金总未平仓合约和突发风险事件为解释变量，将研究范围按油价走势特点分6个阶段，运用协整理论建立协整关系，得出各阶段各变量对Brent油价的贡献比例；建立VARM和VECM，对2017年Brent油价进行预测．结果表明，供需基本面起主导作用，特别是2008年金融危机以后供需基本面影响占比为50-60％，定量分析与实际国际石油市场及油价走势特点相匹配；VARM和VECM对2017年Brent油价预测均值分别为65．63和59．37美元／桶，VECM更接近路透社的58．01美元／桶．2017年，石油市场或由供需过剩转为缺口，推动油价水平整体上升，预计全年Brent油价均价为52—60美元／桶，VARM和VECM预测趋势与其基本一致．综合来看，VECM预测效果更优，解释性更强．

英文摘要：

Studing quantitatively in stages, and making Prediction the international oil price can provide reference guidence for China dealingwith oil price fluctuations to ensure oil Se- curity and procurement. By using the Eviews Econometrics Software, thus paper select Brent oil price as explained variable and five variables are ultimately selected as the explanatory variables including Dow Jones-AIG Commodity Index, Commercial crude oil inventories in the US, US dollar index, the total fund open interest, and the risk emergency factor. The research range is divided into 6 stages according to the oil price trend. Use the cointegration theory to establish the cointegrationn relationship and obtain the contribution ratio of each variable to the Brent oil price. Then, set up the VARM and VECM to forecast the Brent oil price monthly in 2017. The results show that the fundamentals of supply and demand play a leading role, the influence of whiCh accounts for 50-60% especially after the 2008 financial crisis. The quantitative analysis matches the actual international oil market and oil price trend. The average values respectively in VARM and VECM are 65.63 and 59.375/bbl. The prediction value of VECM is more close to that of Thomson Reuters 58.015/bbl. In 2017, the international oil market will transfer from the excess of supply and demand into the insuffi- ciency so that the overall level of Brent oil price will rise. Brent oil annual price in 2017 will be expected to 52-605/bbl. The prediction trend of the VARM and VECM is consistent with this period, Overall, the predict effect of VECM is better and the explanation of it is stronger than VARM.