Enhancing Probability of Default Prediction: Non-Linear Modeling in Turbulent Economic Times
This thesis investigates the application of spline regression models to predict the Probability of Default (PD) under varying macroeconomic conditions, exploring whether these models can enhance predictive accuracy over traditional linear models and compare favorably to XGBoost. The study analyses the non-linear dynamics between PD and key macroeconomic indicators within a Swedish small-sized corp
