Explainable Graph Neural Networks for Interbank Contagion Surveillance: A Regulatory-Aligned Framework for the U.S. Banking Sector

TL;DR

This paper introduces ST-GAT, an explainable graph neural network framework for early detection of bank distress and macro-prudential surveillance in the U.S. banking system, utilizing publicly available data and achieving state-of-the-art performance.

Contribution

The paper presents a novel, explainable GNN framework tailored for interbank contagion surveillance, incorporating temporal dynamics and transparency, with publicly available data and code.

Findings

ST-GAT achieves the highest AUPRC among GNNs (0.939) for bank distress prediction.

Ablation shows the BiLSTM component improves AUPRC by +0.020.

Permutation importance highlights ROA and NPL Ratio as key predictors.

Abstract

The Spatial-Temporal Graph Attention Network (ST-GAT) framework was created to serve as an explainable GNN-based solution for detecting bank distress early warning signs and for conducting macro-prudential surveillance of the interbank system in the United States. The ST-GAT framework models 8,103 FDIC insured institutions across 58 quarterly snapshots (2010Q1-2024Q2). Bilateral exposures were reconstructed from publicly available FDIC Call Reports using maximum entropy estimation to produce a dynamic directed weighted graph. The framework achieves the highest AUPRC among all GNN architectures (0.939 +/- 0.010), trailing only XGBoost (0.944). Ablation analysis confirms the BiLSTM temporal component contributes +0.020 AUPRC; temporal attention weights exhibit a monotonically decreasing pattern consistent with long-run structural vulnerability weighting. Permutation importance identifies ROA (0.309) and NPL Ratio (0.252) as dominant predictors, consistent with post-mortem analyses of the 2023 regional banking crisis. All data are publicly available FDIC Call Reports and FRED series; all code and results are released.