Beilak N. Aliev Comparison of Forecast Accuracy for Classic and Alternative Price Bars in IT Companies

DOI: 10.38050/2078-3809-2025-17-1-22-38

Abstract

The article deals with the urgent problem of improving the accuracy of forecasting the price movements of shares of companies from the information technology sector, which is due to the increased interest of investors and traders in this sector in recent years. The aim of the study is to compare the accuracy of forecasts based on classical and non-standard price bars and evaluate their impact on the effectiveness of trading strategies.

Modern statistical methods and machine learning were used as the main research method to analyze and evaluate the predictive abilities of different types of price bars. In the course of the work, software functionality was developed to generate non-standard price bars, such as bars based on the price of gold, and various trading strategies based on moving averages and AutoML models were tested.

The author's results showed that the use of non-standard price bars improves the predictive properties of the models, which leads to an increase in the efficiency of trading strategies. The practical significance of the obtained results lies in providing recommendations to traders and investors on the selection of optimal types of price bars to improve the accuracy of forecasts. Theoretical significance consists in confirming the hypothesis of higher efficiency of non-standard price bars in trading systems focused on IT companies.

Keywords: price forecasting, price bars, IT companies, trading strategies, machine learning, financial markets, AutoML.

JEL: C63, C81, C87.

For citation: Aliev, B.N.  (2025) Comparison of Forecast Accuracy for Classic and Alternative Price Bars in IT Companies. Scientific Research of Faculty of Economics. Electronic Journal, vol. 17, no. 1, pp. 22-38. DOI: 10.38050/2078-3809-2025-17-1-22-38

Download