Analysis And Prediction Of International Crude Oil Price Based On Wavelet And Regression Method

Crude oil, also known as black gold, at present, it is the most important basic energy and industrial raw materials, and it is also essential strategic materials, it has a strong supporting role of economic development and national security for oil-producing countries and consuming countries. Therefore, the fluctuations of international crude oil price have played a pivotal role of the international and domestic political and economic situation.The analysis and forecasting of international crude oil price is an important issue to maintain oil production enterprise, consuming business and national interest. Prediction of oil price, is the prerequisite and foundation for the corresponding enterprise of production and consumption business to make decisions. The survival and development of an enterprise depends on the correct decisions, and the correct decision is dependent on the accuracy of pre-informations. Therefore, the forecasting of international crude oil price is of great significance to the maintenance the interests for producer and consumer of oil.Based on the complexity of international oil market, not only considering the self-developing laws of the time series of oil price, and adding major factors that influence oil price in the model, this article adopts the prediction method combining wavelet analysis and multiple regression. Firstly, the sequence of oil price and factors affecting oil price are decomposed by the multi-resolution and single-scale wavelet analysis, achieving low-frequency profile of the sequences reflecting the trend and the high-frequency details reflecting the short-term volatilities. As the decomposition of the factors can reflect the changing law of corresponding decomposed sequence of oil price more sufficiently, this article establishes the forecasting model of profile sequence by multiple regression, and then, similarly, establish the forecasting model of detail sequence. Finally, reconstruct the results of two models using wavelet analysis method to get the final forecasting result of oil price. And the calculation example of the spot price of West Texas intermediate crude oil indicates that the method has high prediction accuracy, which illustrates the feasibility and effectiveness of this proposed method.This paper describes the background, significance and domestic and foreign related theoretical models of the analysis and forecasts of international crude oil price and then analysis the factors affecting international oil price. Chapters 4 and 5 are the core of this paper. The fourth chapter introduces the research methods used in this paper, and proposes the forecasting model of international crude oil price based on wavelet analysis and regression. Chapter 5 conducts an empirical analysis of the monthly WTI crude oil price, validating this proposed method to forecast oil price. Finally, Chapter 6 of this paper provides the concluding remarks, including the research work, and the need for further researches.