Multifractality And Cross-Correlation Analysis Of Domestic And International Crude Oil Markets

With the increase of China’s imports of international crude oil, there are a lot of contact between China crude oil prices and international crude oil prices. So, the study of China and international crude oil market structures and the fluctuation of prices have important significances for risk management and the formulation of China’s crude oil market policy.Using the multifractal detrended fluctuation analysis method, we study the multifractal characteristics of PetroChina and Sinopec two listing Corporation stock returns. Combining the multifractal spectrum method, we compare the strength and the size of the risk for the multifractality of two stock returns series. The results show that the stock return series of two companies have obvious multifractal characteristics and the multifractality is stronger and fluctuation complexity is higher for Sinopec return series. In general, Sinopec shares have more profit space, but the risk is also greater in the two stock returns series. We also use the cross-correlation statistics and multifractal analysis method to study the domestic and international crude oil market return series and confirm the existence of cross-correlated behaviors between crude oil markets. Meanwhile, we compare the multifractal strength of cross-correlated returns series and discuss the sources of multifractality. The results show that the cross-correlated behavior of small fluctuations is stronger than the large fluctuation’s. Both the long-range correlation and the fat-tail distribution play important roles in the contributions of multifractality. The main structures of this thesis are organized as follows:Chapter 1 is the introduction. We introduce the background and significances of the study. Furthermore, we explain the research necessity of the domestic and international crude oil markets and summarize the research results of multifractal analysis for domestic and international crude oil markets in recent years.Chapter 2 mainly introduces the multifractal analysis methods of MF-DFA and MF-DCCA.In Chapter 3, we first use MF-DFA method to study the multifractal characteristics of PetroChina and Sinopec two listing Corporation stock returns. The results show that the stock return series of two companies have obvious multifractal characteristics, and the multifractality of Sinopec return series is stronger and its fluctuation complexity is higher. Then, we apply the MF-DCCA method to study WTI/Brent, WTI/Daqing, WTI/Shengli, Daqing/Shengli four price return series and confirm the existence of cross-correlated behaviors between the crude oil markets. Further, we point out that this relationship has multifractal characteristics. Meanwhile, we compare the multifractal strength of cross-correlated returns series. The results show that the cross-correlated behaviors of small fluctuations are persistent and the cross-correlated behaviors of large fluctuations are anti persistent. Comparing the multifractal spectrum widths of return series, we find that the financial risk measurements of the portfolio series are less than the risk measurements of autocorrelation series.In Chapter 4, we study the causes of multifractality of WTI/Brent, WTI/Daqing, WTI/Shengli, Daqing/Shengli four price return series. The results find that both long-range correlations and fat-tail distributions play important roles in the contributions of multifractality. Further, we point out the main reason of the multifractality.Chapter 5 is the summary of results of this thesis and the prospect of further research.