| Abstract [eng] |
In this work, methods for modelling and forecasting oil prices are used: regression analysis, exponential smoothing, vector autoregression, integrated model of the moving average and generalized autoregressive conditional heteroskedastiveness models. These models are described in the literature as the most accurately modelling and predictive of crude oil prices. For analysis selected the four oil price times series: WTI (petroleum, widely used as a benchmark in the North American oil market), BRENT (oil produced in the North Sea and used as a benchmark in European, African, Middle Eastern oil markets), Dubai (this is mainly oil produced in the Middle East) and OPEC (Organization of the Petroleum Exporting Countries). For prices analysis have been selected these factors: inflation, the dollar and the euro rate and gross domestic product. |
