SWT-ARMA Modeling of Shenzhen A-Share Highest Composite Stock Price Index

Abstract

Simulation and prediction of stock trading price time series have always been one of the important research contents in the financial investment industry. Based on the data from the national stock trading statistics table, SWT and ARMA steady-state modeling, the multiscale simulation and prediction model SWT-ARMA is constructed on the time series of the highest composite stock price index Shenzhen A-share, which quantitative simulation analyzes and embedded predicts the evolutionary trend of the sequence and its effectiveness. The results shows that the fitting ability and extrapolation prediction accuracy of the multiscale SWT-ARMA model are higher than the accuracy of the ARMA model, and it is adaptive to the smoothing preprocessing and modeling of the stock data. Therefore, this method can effectively model and predict the time series of stock trading prices.