A Study Of The Relationship Between International Crude Oil Price And Stock Market Based On EEMD Method

As an industrial raw material and energy source for promoting social and economic development,crude oil is an indispensable and important component of the world economy.With the continuous development of global industrialization and integration,the links between different assets are becoming more and more close.In 2017,China imported more crude oil than the United States and became the world’s largest importer of crude oil.In 2018,the dependence on crude oil rose to 70.9%,an increase of 3.5% over the previous year.Demand has increased year by year,so the study of the impact of international crude oil prices for China’s economy is of great significance.This paper takes Brent crude oil price and Shanghai Composite Index as the research object,from the perspective of time domain and frequency domain to dynamically analyzes the correlation between international crude oil price and China’s stock market.Then predicts the international crude oil price and stock market price.Research analyze mainly from the following aspects:Firstly,from the perspective of frequency domain to analyze the structure characteristics of crude oil price and stock market price.EEMD decomposition is performed on two price series.Then,the two price series is divided into low frequency term,high frequency term and trend term according to IMF reconstruction method.The results show that the high frequency component reflects the normal market fluctuation of the crude oil price and the Shanghai Composite Index.The low frequency component reflects the major events on the crude oil price and the Shanghai Composite Index.The impact of the trend item reflects the internal trajectory of the crude oil price and the Shanghai Composite Index.And explaining the market volatility,major event changes and internal trajectories in the corresponding period.Secondly,according to the high-frequency component,low-frequency component and trend term of the crude oil price and the stock market price.Through the rolling window test method to analyze the dynamic relationship of the crude oil price and the stock market.In order to avoid the problem of pseudo-regression in the state space model,first the stability of the variable is tested.And then in order to judge whether the variable meets the requirements of the state space model,the cointegration test is performed.The state space model is constructed to dynamically analyze the linkage relationship with the crude oil price and the stock market price from the time domain perspective.The results show that although there are some major events or market fluctuations in different periods,the impact of the crude oil market on the Shanghai Composite Index is far greater than the impact of the Chinese stock market on the crude oil market.Finally,a multi-scale reconstruction combination model is constructed to predict Brent crude oil price and stock price series.The EEMD method is used to decompose the crude oil price and the stock price original sequence,and then reconstructed into low frequency term,high frequency term and trend term.The SVR models are established for the three subsequences obtained in the previous step,and the predicted results used to training SVR model to get the final predicted value.The results show that compared with the single model and the unreconstructed multi-scale combined model based on EMD and EEMD,the multi-scale reconstruction combined model constructed in this paper is significantly better than other models,and achieves very high precision prediction.