A Comparative Study on Stock Price Forecasting after the COVID-19 Pandemic

Abstract

Nowadays, stock price forecasting is still a difficult problem plaguing people, due to the uncertainty of the market. We contrast the effectiveness of three models in this research.: ARIMA, SVM, and BPNN model. To get the outperformed model in predicting after the COVID-19 pandemic, the data set is the stock price of Citic Securities (600030) on the Shanghai Stock Exchange from 2022.4.1 to 2022.12.30. Experiment results show that all three models predict outcomes with very little error from the actual values. And BPNN model has the best accuracy because of its great ability to portray the non-linear characteristic of our data.