Wti Crude Oil Futures Price Fluctuation Analysis

Oil is the indispensable resource of modern industrial economy,it is the blood of industry.With the development of our country's economy,the increase of the proportion of industrial structure in industrialization constantly,and domestic oil supply shortage,countries still needs a large number of imported oil,especially for crude oil dependence on OPEC countries,also contributed to the phenomenon of "Asian premium".Due to the increasing size of China's oil consumption market,the Asian premium has more and more influence on China.It can be seen that the fluctuation of oil price has an important influence on China's industrial economy,so the accurate prediction of oil price is of great significance to relevant departments and investors.In this paper,the WTI oil futures price of NYMEX(New York mercantile exchange)is selected as the research object,and the fluctuation of oil price is predicted with the advantage of multi-resolution of wavelet in time frequency.Based on this feature,firstly,based on WTI oil phase price,we draw the contour line of the morlet wavelet coefficient and the continuous wavelet power spectrum,and analyze the periodic fluctuation characteristics of WTI.Again using wavelet coherence and granger causality test analyzes the WTI oil prices and China's industrial production growth rate of resonance behavior and influence each other relations,shows that the oil price fluctuation period is the granger cause of industrial added value growth rate,with unidirectional conduction function.Finally,based on the time series model and wavelet multi-resolution analysis theory,the ARIMA model of stochastic time series and the combined model of wavelet are established to study the crude oil price.In empirical study,selects the db5 wavelet function for crude oil futures prices five layer decomposition reconstruction,according to the data characteristics of each component to establish appropriate ARMA(p,q)prediction model,each component layer of the combination of the fitting results for synthetic crude oil futures prices model fitting results compared with single model predicted results,results show that the ARMA-wavelet combination prediction model has better prediction ability,accuracy is better than a simple time series model,which proves the feasibility and effectiveness of this method.