Abstract:

Oil price prediction plays an important role in production and consumption planning; affecting operating result of an oil and gas company. There are many prediction methods on long term and short term basis. Out of short-term predition, eg yearly, quarterly, monthly, weekly, prediction on quarterly basis is important in planning and implementing production and consumption process. Any prediction, however, exists error, so that it is needed to choose an estimation method with minimum error. This paper mentions the estimation methods such as ARIMA, ARCH, GARCH, and ANN, ... Thorough calculating comes up with the conclution that ANN method brings out minimal forecasting error. This method can be applied in PetroVietnam, BSR, a company which is affected by oil prices fluctuation, contributing to improving the effectiveness of planning both in domestic and abroad. Quarterly oil prices prediction is an economic problem with high sensitivity, complexity and subject to many factors. So, it’s important to study for better improvement