FORECASTING CORPORATE FINANCIAL PERFORMANCE USING MACHINE LEARNING METHODS AND TIME SERIES ANALYSIS

Ayagoz Imansakipova, Daria Moldash, Umit Payiz

2026

Full text Open access: Gold Metadata (CSL-JSON)
DOI 10.56132/2791-3368-2026-2-66-204-214
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

The article discusses the problem of forecasting financial performance of companies using machine learning methods and time series analysis. The relevance of the research is due to the need to improve the accuracy of financial forecasting in conditions of high market uncertainty and digitalization of the economy. The purpose of the study is a comparative analysis of forecasting models of quarterly revenue of companies based on financial and macroeconomic data. The study is based on quarterly data from 21 public companies for the period 2008-2025 using macroeconomic indicators, including the consumer price index, U.S. government bond yields, the price of Brent crude oil, and the NASDAQ Composite and S&P 500 stock indexes. ARIMA, SARIMA, Prophet, linear regression, Support Vector Regression, Random Forest, XGBoost, and Long Short-Term Memory models were used for forecasting. Preliminary data processing, formation of lag signs and integration of macroeconomic factors are performed. The quality of the models was assessed using RMSE, MAE, MAPE, and sMAPE metrics. The results showed that ensemble machine learning algorithms provide higher prediction accuracy compared to classical time series models. Random Forest has shown the most consistent results. It is established that the choice of the optimal model depends on the industry specifics and the specifics of the company's time series. The developed approach can be used for financial planning, investment analysis and management decision support.