Research on Stock Market Investment Model Based on Time Series Forecasting and Dynamic Programming
XuDong An, Xuan Yi, BeiQi Zhou, Qinjuan Zhang · Atlantis highlights in economics, business and management/Atlantis Highlights in Economics, Business and Management · 2023
Forecasting the value of a stock population has always been attractive and challenging for shareholders due to its inherent dynamics, nonlinearity and complexity.In this paper, we propose a stock investment model based on time series forecasting and dynamic programming.The time series forecasting model is utilized for next day high-and low-price prediction, combined with the dynamic programming model to formulate stock trading strategies.The study conducted simulated back tests on 200 stocks randomly selected from the Chinese Shanghai and Shenzhen stock markets, and the results show that the scheme proposed in this study can achieve a return of more than 12% after more than a 3-month investment cycle.