Multifractal Analysis Of Metal Futures,Foreign Exchange And Stock Markets

With the deepening of global economic integration,the economic links and interdependence among the countries in the world have been strengthened,the domestic economic rules of countries have been converging,and the effect of information dissemination and risk transmission among financial markets has been significantly enhanced.As important parts of the financial markets,the internal linkages among the futures,foreign exchange,stock and currency markets have become more and more obvious.At peesent,China is in the critical period of the reform and innovation of financial institutions,RMB internationalization and the control of COVID-19,and the complexity and risk of the financial markets are increasing.Therefore,the influence of fluctuation between different markets are closely concerned by securities investors,market regulators and policy makers.Firstly,this dissertation uses multifractal analysis methods to explore the autocorrelation and cross-correlation of the return rate series of China's metal futures,foreign exchange and stock markets,and reveals the nonlinear fluctuation characteristics and dynamic complexity of these three markets.The results show that the autocorrelation and cross-correlation of metal futures,foreign exchange and stock markets all have obvious multifractal characteristics,and the multifractal strength of the cross-correlation relationship between different markets is less that of the autocorrelation of the market itself.At different time scales,the cross-correlation between the markets shows different multifractal characteristics.In addition,we also find that the Hurst surface between the stock market and other markets is the most complex,and that the stock market is the most sensitive to external factors.Secondly,the overlapping smoothing window technology is introduced to improve the existing MV-MFDCCA fractal analysis method which eliminates the discontinuity of sequence fitting and the pseudo-fluctuation generated by the fluctuation function,and the improved method is applied to study the overall cross-correlation between Chinese and American stock markets.Finally,from the perspective of system theory,the improved multifractal analysis method is used to reveal the time-varying characteristics of the multiple interactions between Chinese and American stock markets before and after the outbreak of COVID-19,and to analyze the market that plays a leading role after the interaction of Chinese and American stock systems.The results show that before and after the outbreak of COVID-19,the autocorrelations of both the two stock market return series and their corresponding component series have multifractal characteristics.The efficiency of Chinese and American stock markets has decreased,and the internal complexity of the markets have increased.In addition,there are long-range power-law cross relationship and multifractal characteristics between Chinese and American stock markets before and after the outbreak of COVID-19,and the influence of American stock market on Chinese stock market is dominant regardless of small or large fluctuations.Besides,the Hurst surface is used to visualize the time-varying cross relationship between Chinese and American stock markets,and finding that the outbreak of COVID-19 enhances the multifractal characteristics between Chinese and American stock markets,and increases the risk complexity between Chinese and American stock markets.