Search for a command to run...
Purpose This study aims to develop an AI-based predictive model to predict the Leverage variable as a crucial factor in financial performance and risk management within the banking sector. Design/methodology/approach The research is based on a data set extracted from the annual disclosure reports of 15 banks listed on the Amman Stock Exchange over 13 years (2010–2022). The data set consists 19 variables, and rigorous preprocessing techniques, such as outlier detection, variable scaling and variable engineering, were applied to ensure data quality. Four AI models were developed to identify complex nonlinear structures in the data, supporting precise leverage prediction. The developed models include Gradient Boosting Regressor (GBR), Extreme Gradient Boosting (XGBoost), Random Forest (RF) and Decision Tree (DT) represent a range of ensemble methods that are well-suited for financial data sets. Findings After developing and applying the four AI models—GBR, XGBoost, Random Forest and Decision Tree—a comprehensive performance evaluation and hyperparameters optimization were conducted using four metrics: Coefficient of Determination (R²), Mean Absolute Error (MAE), Mean Squared Error (MSE) and Root Mean Squared Error (RMSE). The results demonstrated that the Gradient Boosting Regressor (GBR) consistently outperformed the other models across all metrics, confirming its reliability in predicting financial Leverage. The model’s superior ability to capture nonlinear relationships, combined with robust hyperparameter optimization, highlights its potential to enhance strategic decision-making and improve financial risk management in the banking sector. Originality/value To the best of the authors’ knowledge, this study is the first attempt to provide a structured comparison of several ensemble learning models, such as GBR, XGBoost, Random Forest and Decision Tree, in predicting financial leverage within the banking sector. By doing so, the research fills an important gap in the existing literature. It also offers a clearer understanding of how AI-based models can support financial stability and contribute to more informed decision-making.