Modelling Sovereign Credit Ratings: Evaluating the Accuracy and Driving Factors using Machine Learning Techniques

TL;DR

This study compares machine learning models for predicting sovereign credit ratings, finding that Multilayer Perceptron performs best and identifying key economic and institutional factors influencing ratings.

Contribution

It evaluates multiple machine learning techniques for sovereign rating prediction and highlights the most effective model along with important predictive factors.

Findings

MLP achieves 68% accuracy in rating prediction

Regulatory quality and GDP per capita are key influential variables

MLP outperforms CART, SVM, NB, and OL models

Abstract

Sovereign credit ratings summarize the creditworthiness of countries. These ratings have a large influence on the economy and the yields at which governments can issue new debt. This paper investigates the use of a Multilayer Perceptron (MLP), Classification and Regression Trees (CART), Support Vector Machines (SVM), Na\"ive Bayes (NB), and an Ordered Logit (OL) model for the prediction of sovereign credit ratings. We show that MLP is best suited for predicting sovereign credit ratings, with a random cross-validated accuracy of 68%, followed by CART (59%), SVM (41%), NB (38%), and OL (33%). Investigation of the determining factors shows that there is some heterogeneity in the important variables across the models. However, the two models with the highest out-of-sample predictive accuracy, MLP and CART, show a lot of similarities in the influential variables, with regulatory quality, and GDP per capita as common important variables. Consistent with economic theory, a higher regulatory quality and/or GDP per capita are associated with a higher credit rating.