企業間取引ネットワークを考慮した信用リスク分析

Read full text Summarise with VIRAT
Authors: Taishin Wakisaka, Hiroshi Tsuda
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

It is well known that inter-firm transaction networks function as channels for transmitting credit risk. Previous studies have shown that rating prediction models incorporating customer–supplier information achieve higher predictive accuracy than those relying solely on financial indicators. This study aims to quantify the intensity of credit risk transmission through inter-firm networks using Graph Neural Networks (GNNs) and to test its statistical significance. The intensity of credit risk transmission refers to the extent to which a supplier is affected when a customer experiences a credit event. Our results demonstrate that the attention weights obtained from estimating suppliers' credit risk using a Graph Attention Network (GAT) effectively capture the strength of credit risk propagation across firms. In conclusion, the GNN-based analysis enables a quantitative evaluation of supplier–customer relationships in terms of their impact on credit risk transmission.

Fields of Study & Subjects

Financial Distress and Bankruptcy Prediction Credit Risk and Financial Regulations Working Capital and Financial Performance

Publication details

Journal / source
Institutional Repositories DataBase (IRDB)
Volume
66
Issue
4
Pages
215–225
Publication date
1 January 2026
Language
JA