Price Prediction in terms of Linear Regression For F, BYDDY and POAHY

Abstract

In the environment surrounding the capital markets, investors are keen on predicting the stock market. Many companies and research institutions are following stock price movements every day. As more and more models and algorithms are proposed, the accuracy of forecasting is also greatly improved. This paper mainly focuses on adopting linear regression models to predict the future trends of three stocks, BYDDY from A-shares, POAHY from European stocks, and F from US stocks. By building the linear regression model and calculating the R-squared, the accuracy of the model is above 90% and the MAE between the predicted value and the true value is very small. The correlation analysis shows that the variables Open, High, Low, and Adj Close have a positive correlation with the predicted values. In addition, there is a correlation between the three stocks. In addition to successfully predicting the trends of the three stocks, the areas of stock price fluctuations were also obtained. Forecasting using linear regression models is a proven way to make it easier to understand and calculate. These results shed light on verifying the feasibility of the linear regression model to predict stock prices.