Applying artificial intelligence to forecast the market capitalization of global companies based on ESG and financial indicators

This study investigates the effectiveness of artificial intelligence in forecasting the market capitalization of publicly listed firms by integrating Environmental, Social, and Governance (ESG) indicators with traditional financial variables. Grounded in Stakeholder Theory and Signaling Theory, the research evaluates whether ESG information enhances predictive performance beyond conventional financial metrics. The analysis employs a global panel dataset comprising more than 11,000 firm-year observations from approximately 1,000 companies across multiple industries and regions during 2015–2025. Several machine learning models, including Random Forest, CatBoost, Extreme Gradient Boosting, Light Gradient Boosting Machine, Extra Trees, and Linear Regression, have been developed using log-transformed financial variables and dimension-reduced ESG components derived through principal component analysis. Time series validation is applied to ensure temporal robustness. The findings indicate that tree-based ensemble models significantly outperform linear regression, with Random Forest explaining 84.53% of the variation in future market capitalization. Financial indicators, particularly revenue and profit margin, remain dominant predictors, while ESG factors contribute limited short-term incremental value. The results highlight the complementary role of ESG reporting in supporting long-term transparency and sustainable competitiveness.