SHAP Stability in Credit Risk Management: A Case Study in Credit Card Default Model

TL;DR

This paper assesses the stability of SHAP, an explainable AI method, in credit card default prediction models, highlighting its dependence on variable importance and offering practical insights for credit risk management.

Contribution

It provides an empirical evaluation of SHAP stability in credit risk models, revealing factors influencing consistency and guiding its application in financial risk management.

Findings

SHAP stability correlates with variable importance levels.

Consistency of SHAP explanations varies across features.

Practical recommendations for using SHAP in credit risk assessment.

Abstract

The increasing development in the consumer credit card market brings substantial regulatory and risk management challenges. The advanced machine learning models applications bring concerns about model transparency and fairness for both financial institutions and regulatory departments. In this study, we evaluate the consistency of one commonly used Explainable AI (XAI) technology, SHAP, for variable explanation in credit card probability of default models via a case study about credit card default prediction. The study shows the consistency is related to the variable importance level and hence provides practical recommendation for credit risk management