Machine Learning Approaches For Predicting Corporate Financial Stability Across Industries
Abstract
In this paper, the application of machine learning (ML) models in predicting corporate solvency within the Energy, Process Industries, and Transportation sectors was examined. ML was presented as an innovative alternative to traditional financial analysis, which can be time-consuming. For this purpose, a dataset of 262 public companies was analyzed. Key financial ratios were focused on to classify firms as either solvent or at risk. Next, Logistic Regression, Random Forest, and Support Vector Machines were employed to assess predictive accuracy and interpretability. It was revealed that predictive performance is strictly industry-dependent. Linear models were suggested for capital-intensive industries to ensure high recall, while Random Forest was recommended for industries with operational volatility to maximize precision. Based on the results, a dual-layer risk assessment architecture was proposed to enhance decision-making processes for analysts. Finally, future research directions were recommended to explore additional metrics for improved solvency forecasting.