Predicting Global Banking Stability: A Machine Learning Approach Using the Camels Framework

Abstract

This study assesses the performance of the 33 largest banks in the top 13 GDP countries using the CAMELS framework. Beyond traditional CAMELS ratings, the research explores predictability through 25 machine learning algorithms. Analyzing data on key CAMELS variables—capital adequacy, asset quality, management efficiency, earnings quality, liquidity, and market risk sensitivity—the study aims to predict future bank performance. Machine learning algorithms are employed to build predictive models, evaluating accuracy, precision, and generalization capabilities. The findings offer insights crucial for risk management, regulatory decisions, and proactive measures to prevent banking crises.

Keywords: Banking performance, CAMELS framework, machine learning, Early Warning Systems, Classification