Research On The Stock Market Model Based On Delayed Information Impact

Finding the key factor and possible “Newton’s laws” in financial markets has remained the central issue in this research domain.However,with the development of information and communication technologies,financial models are becoming more realistic but complex by incorporating more individual characteristics,network relationships and evolutionary rules,which contradicts the objective law “Greatest Truths Are the Simplest.” The power-law distribution suggests that large effects in complex systems are often triggered by a minority of key factors,and it is not necessary to describe individual traits in detail in economic modeling but only to find models that are simple enough to include key factors that can be fully explained;and also can generate stylized facts.Therefore,this paper attempts to find the most critical factor in the financial market and presents an evolutionary model that is independent of micro features.Now,the percolation model,the Ising model,and the network topology model have been widely studied.Most of them share a common feature double probability forms.Inspired by it,a stock market model of delayed information impact is proposed in the paper.In the model,information is the only key factor independent of the internal characteristics of the market,and stock price fluctuations are the emergence of the collective intelligence of investors’ behavior.Considering that significant events have a sustained impact on the stock market,we assume that the information influence will sustain,but the impact strength of the information will decay over time.So,the stock price model of decayed information impact includes two components(external and internal)–external: the generation and delay of market information,internal: the fluctuation of collective decision-making in the given information.We estimate the descriptive statistics of simulated returns,K-S test,power-law fit,as well as characterize the nonlinear behavior of returns by correlation dimension analysis,calculating Max Lyapunov Exponent,sample entropy,and Hurst exponent.By verifying with comparing to the real market,showing that simulated returns have almost the same macro-statistical laws and chaotic property as the real market returns.On the other hand,the different stock price series generated by the model with the random behavior of the population with the same historical information can be analogous to stocks of the same industry in the real market.The statistical results show that the two have similar single-peaked distribution.Besides,since each set of simulated data is generated under the same sequence of historical information,this can explain the probability that the stock market will also evolve similarly to the previous ones when similar historical information is available.The paper demonstrates that the model can recapture the universal characteristics of stocks and the chaotic characteristics of the real stock market,capture the key factors in the financial markets.It also can be effectively used in stock correlation research and history recurrence.Our work opens a new way to selecting rational portfolios,complements current industry correlation research methods.It provides a useful framework for understanding stock price evolution from the emergence of collective intelligence.Also,it opens up new ideas and provides theoretical support for studying the core structure or critical factor of financial market.