Relational Graph Modeling for Credit Default Prediction: Heterogeneous GNNs and Hybrid Ensemble Learning

TL;DR

This paper constructs a large heterogeneous graph for credit default prediction, evaluates GNNs and hybrid models, and finds that combining GNN embeddings with tabular models improves predictive performance and provides insights into relational influences.

Contribution

It introduces a massive-scale heterogeneous graph for credit risk modeling and demonstrates that hybrid ensemble models outperform standalone GNNs and tabular models.

Findings

Hybrid ensemble models outperform standalone GNNs and tabular models.

Contrastive pretraining stabilizes training but offers limited downstream benefits.

Relational signals influence subgroup behavior and screening outcomes.

Abstract

Credit default risk arises from complex interactions among borrowers, financial institutions, and transaction-level behaviors. While strong tabular models remain highly competitive in credit scoring, they may fail to explicitly capture cross-entity dependencies embedded in multi-table financial histories. In this work, we construct a massive-scale heterogeneous graph containing over 31 million nodes and more than 50 million edges, integrating borrower attributes with granular transaction-level entities such as installment payments, POS cash balances, and credit card histories. We evaluate heterogeneous graph neural networks (GNNs), including heterogeneous GraphSAGE and a relation-aware attentive heterogeneous GNN, against strong tabular baselines. We find that standalone GNNs provide limited lift over a competitive gradient-boosted tree baseline, while a hybrid ensemble that augments tabular features with GNN-derived customer embeddings achieves the best overall performance, improving both ROC-AUC and PR-AUC. We further observe that contrastive pretraining can improve optimization stability but yields limited downstream gains under generic graph augmentations. Finally, we conduct structured explainability and fairness analyses to characterize how relational signals affect subgroup behavior and screening-oriented outcomes.