Sovereign credit risk has become increasingly significant for policymakers, investors, and financial institutions, particularly during periods of political instability. Traditional risk assessment methods often fail to capture the dynamic and nonlinear relationships in credit markets. This study explores the application of machine learning (ML) and deep learning (DL) techniques for forecasting sovereign credit risk under political crises. Using sovereign credit default swap (CDS) spreads and macroeconomic indicators from Egypt, Saudi Arabia, and Morocco, we implement models including Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and Extreme Gradient Boosting (XGBoost) to forecast default probability and CDS spread movements. Our results demonstrate that DL models, particularly GRU-based architectures, outperform traditional linear and classical statistical models in terms of predictive accuracy. The findings suggest that ML and DL approaches provide robust tools for capturing complex dependencies in sovereign credit risk, offering strategic insights for investors and policymakers facing political uncertainties.

sovereign credit risk, political crisis, machine learning, deep learning

Abid, A., & Abid, F. (2023). A methodology to estimate the optimal debt ratio when asset returns, and default probability follow stochastic processes. Journal of Industrial & Management Optimization, 19(10), 1.

Abid, A., & Abid, F. (2024). Sovereign credit risk in Saudi Arabia, Morocco and Egypt. Journal of Risk and Financial Management, 17(7), 283.

Abid, A., Abid, F., & Kaffel, B. (2020). CDS-based implied probability of default estimation. The Journal of Risk Finance, 21(4), 399–422.

Abid, A., & Suissi, N. (2024). Sovereign CDS spread and term structure forecasting based on neural network. Global Business Review.

Annaert, J., De Ceuster, M., & De Jonghe, F. (2000). Modelling European credit spreads. Deloitte & Touche.

Ao, S. I., & Fayek, H. (2023). Continual deep learning for time series modeling. Sensors, 23(16), 7167.

Badr, O. M., & El khadrawi, A. F. (2016). Macroeconomic variables, government effectiveness and sovereign credit rating: A case of Egypt. Applied Economics and Finance, 3(4), 29–36.

Augustin, P., Sokolovski, V., Subrahmanyam, M. G., & Tomio, D. (2022). How sovereign is sovereign credit risk? Global prices, local quantities. Journal of Monetary Economics, 131, 92–111.

Bai, M., Zhou, Z., Li, J., Chen, Y., Liu, J., Zhao, X., & Yu, D. (2024). Deep graph gated recurrent unit network-based spatial–temporal multi-task learning for intelligent information fusion of multiple sites. Expert Systems with Applications, 240, 122072.