Using ML to analyze ESG ratings to support the investors’ decision-making process

As sustainable development is becoming an important decision-making factor in financial markets, investors increasingly rely on environmental, social, and governance (ESG) ratings. Unfortunately, they must face persistent challenges concerning ESG data quality, transparency, and comparability. This study explores how machine learning (ML) can address these gaps. To answer this question, a scoping review of recent research was conducted. The findings indicated four key applications of ML in ESG rating analysis: predicting ESG scores, detecting controversies, identifying drivers of ESG performance, and reconstructing opaque rating mechanisms. Advanced ML models such as XGBoost, Random Forest, and ELM demonstrate the ability to enhance precision and reduce subjectivity through data-driven insights. These findings position ML as a strategic tool supporting responsible investing and offering a competitive edge in navigating the complexities of ESG issues and integrating them with investors’ decision-making processes.