Unveiling the Hidden Symmetries in Financial Markets through Non-linear Analysis: Empirical Evidence from International Markets(مقاله علمی وزارت علوم)
منبع:
Journal of Money and Economy, Vol. ۲۰, No. ۲, Spring ۲۰۲۵
275 - 322
حوزههای تخصصی:
This research delves into the application of Chaos Theory and non-linear analysis to the stock market, utilizing empirical data from international markets to scrutinize the presence of chaotic trends and non-linear processes within the time series of 20 international stock price indices, spanning from January 1984 to January 2024. Through the employment of predictability and non-linearity tests, the study found evidence of a non-linear process in the stock price index. Further, correlation dimension tests were conducted to evaluate the correlation between observations, uncovering a significant correlation between variables. Cumulative periodic tests were subsequently applied to refine the analysis, taking into account the chaotic nature of the stock price index variable, thereby affirming the chaotic nature of this process. Following the validation of the stock price index's predictability, ARFIMA, FIGARCH, LSTAR, and ESTAR models were applied for forecasting future periods. Among the models that incorporate long-term memory in the stock price index variable, the FIGARCH model exhibited superior forecasting power by accounting for both the long-term memory and the variance and changes of the variable. Among the nonlinear models, the ESTAR model demonstrated the highest prediction capabilities. The implications of this study are of considerable significance for investors aiming to comprehend and analyze stock market trends, offering a novel perspective on the stock market's dynamics.