Research On Nonlinear Dynamics Characteristics And Predictability Of China's Stock Market

As a "barometer" of a country's macro economy,the stock market occupies a dominant position in the national economy.Since the 21 st century,China's stock market has made great progress,but it also faces many problems that are difficult to solve.To this end,it is particularly important to correctly understand the Chinese stock market and explore its laws of movement.With the traditional linear capital market theory falling into the dilemma of not explaining a large number of "viscosities" in the stock market,the nonlinear science represented by chaos and fractals has gradually developed,which has opened up new ideas for the study of the law of stock market movement.Based on this background,this paper studies the nonlinear dynamics and predictability of China's stock market from the theoretical perspective of nonlinear science.First of all,from the three perspectives of investor behavior,policy rules and external environment,the main dynamic factors affecting stock price changes are systematically sorted out.Through analysis,it is clear that all kinetic factors are nonlinear,they are related to each other,interact with each other,and are twisted together to determine the nonlinear motion of the stock market.Secondly,using the feedback principle,the nonlinear feedback mechanism of the dynamic factors acting on the stock price movement is qualitatively analyzed from the whole.It is found that the stock price movement often shows positive feedback,and it also shows a certain negative feedback in a long time.The existence of feedback mechanism makes the stock price movement have memory and trend in time,showing dynamic characteristics similar to fractal and chaos.Thirdly,using BDS test,R/S analysis and chaotic parameter calculation,the dynamic characteristics of stock index sequences in Shanghai and Shenzhen stock markets are quantitatively analyzed,which proves the existence of nonlinear,chaotic and fractal features in China's stock market.Finally,on the basis of understanding the law of stock price movement,the ARIMA model under the linear thinking paradigm and the local prediction method under the nonlinear thinking paradigm(zero-order local prediction method,weighted zero-orderlocal prediction method)Refers to the prediction attempt and the discussion of predictability.