Predicting the price of a stock using regression analysis
At the moment, there is a high tendency towards investing financial resources for additional earnings on stock exchanges, their speculations, etc. A large number of people work on trading platforms, buy or sell stocks, which causes their price to change in different periods. On the basis of their prices, charts are formed, according to which technical and fundamental analysis is performed to build a strategy for earning and forecast not even the price of a stock, but the direction of movement of its value in the direction of growth or decrease. To obtain an answer to this question, they resort to various methods of studying the historical data of stocks, studying their qualitative indicators, as well as mathematical forecasting methods. This article examines an example of how the algorithm for predicting the stock price at the close of the trading day on the market. The method, which will provide information about the stock price, is based on the use of linear regression analysis. This method is best suited for research, because stock price charts are linear. The method of regression analysis allows you to take into account such factors as the cyclical recurrence of trends and tendencies of increase or decrease in the value of a stock, the length of the trading period, during which historical data is obtained and studied, etc. The authors also studied the work of the forecasting algorithm based on regression analysis and studied factors affecting accuracy – the size of the training sample, the amount of historical data, the number of days for which the price will be predicted in the future. In the course of the study, information was obtained on the accuracy of forecasting stock prices, which is 96 % depending on the currency in which the stock is traded and the company that issued it. It should be noted that the accuracy of the forecast is also affected by the volume of historical data and the financial position of the corporation in the market. The materials are of practical importance for the development of decision support systems in economic sectors. The developed model can be used as a basis for helping to make decisions for trading on the stock exchange.