The sudden outbreak of the COVID-19 pandemic has brought about an undeniable impact on the national economy. In order to delve into how to more efficiently prevent the risks associated with major public health emergencies in the future, this paper focuses on the stock market’s return rate. By integrating a multiple linear regression model with an LTSM model, a comparative assessment of the stock market’s impact before and after the pandemic is conducted. The research findings indicate that there is no significant difference in stock market return rates between the two periods. Furthermore, a comparison is made between the improved model and the Purely Linear Regression Model, ARIMA, Prophet, and ESM, revealing that the improved model provides more accurate predictions of stock market return rates. This discovery holds the potential to significantly reduce the latent risks brought about by major events and crises, enabling governments to make relevant policy adjustments based on the independent variables mentioned in this paper, thus mitigating potential economic losses.