Prediction of Financial Soundness Index for Viability in the European Banking System Using Machine Learning

Abstract

This study aims to predict financial soundness for sustainability of European banking system using machine learning techniques. A sustainable banking system ensures the efficient use of economic resources, manages risks in a balanced manner, and contributes to long-term financial stability and economic continuity both at the sectoral and national levels. In this context, this study constructs a banking soundness index based on the financial soundness indicators determined by the IMF. The study sample consists of liquidity, asset quality, capital adequacy, and, profitability data of banking system of 51 European countries between 2008 and 2023. However, due to lack of data, financial soundness indicators of 28 European countries were included in the analysis and the data are processed using min-max normalization methods. ANN, MLR, and DTR methods were applied and, ANN machine learning method provided the highest success (R2 = .996, MAE = 0.007, RMSE = 0.011). The findings of the study demonstrate the effectiveness of the model in providing early warnings regarding the financial soundness of these institutions, contributing significantly to the academic literature by evaluating the long term sustainability of financial system. It is essential, however, to recognize that while a unified index allows for consistent cross-country comparisons and risk assessments, it may not fully encapsulate the unique characteristics of individual banking systems. Therefore, complementary country-specific analyses will be crucial for comprehensive risk assessments.