Combining Intra-Risk and Contagion Risk for Enterprise Bankruptcy Prediction Using Graph Neural Networks

TL;DR

This paper tackles SME bankruptcy prediction by jointly modeling intra-risk and contagion risk. It introduces an intra-risk encoder built from statistically significant enterprise indicators and a contagion-risk encoder that combines Hyper-GNNs and Heter-GNNs to capture both high-order and pairwise diffusion effects within an enterprise knowledge graph. A new real-world dataset, SMEsD, is released to benchmark this task, and the proposed ComRisk framework demonstrates superior performance over twelve baselines across accuracy, F1, and AUC, validating the benefit of integrating both risk channels. The work advances practical financial risk analysis by providing a public dataset and a scalable, dual-encoder approach for forecasting bankruptcy with improved reliability and interpretability in risk diffusion contexts.

Abstract

Predicting the bankruptcy risk of small and medium-sized enterprises (SMEs) is an important step for financial institutions when making decisions about loans. Existing studies in both finance and AI research fields, however, tend to only consider either the intra-risk or contagion risk of enterprises, ignoring their interactions and combinatorial effects. This study for the first time considers both types of risk and their joint effects in bankruptcy prediction. Specifically, we first propose an enterprise intra-risk encoder based on statistically significant enterprise risk indicators for its intra-risk learning. Then, we propose an enterprise contagion risk encoder based on enterprise relation information from an enterprise knowledge graph for its contagion risk embedding. In particular, the contagion risk encoder includes both the newly proposed Hyper-Graph Neural Networks and Heterogeneous Graph Neural Networks, which can model contagion risk in two different aspects, i.e. common risk factors based on hyperedges and direct diffusion risk from neighbors, respectively. To evaluate the model, we collect real-world multi-sources data on SMEs and build a novel benchmark dataset called SMEsD. We provide open access to the dataset, which is expected to further promote research on financial risk analysis. Experiments on SMEsD against twelve state-of-the-art baselines demonstrate the effectiveness of the proposed model for bankruptcy prediction.

Figures (8)

• Figure 1: A toy example of enterprise knowledge graph which is extracted from the newly constructed dataset SMEsD.
• Figure 2: The overall architecture of the proposed method. (I) Enterprise Intra-Risk Encoder using the enterprise statistically significant features in Table \ref{['tab:bankruptcy-analysis']}. (II) Enterprise Contagion-Risk Encoder is equipped with two sub-models: (a) Hyper-Graph Neural Networks using enterprise hypergraph, (b) Heterogeneous Graph Neural Networks using enterprise heterogeneous graph, and (c) Combining intra- and contagion risk. (III) Enterprise Bankruptcy Prediction.
• Figure 3: Enterprise Intra-Risk Encoder.
• Figure 4: Hyper-Graph Neural Networks.
• Figure 5: Heterogeneous Graph Neural Networks.

Theorems & Definitions (3)

• Definition 1
• Definition 2
• Definition 3