Estimation of chain reaction bankruptcy structure by chance discovery method — With time order method and directed KeyGraph

Abstract

Chain reaction bankruptcy is regarded as common phenomenon and its effect is to be taken into account when credit risk portfolio is analyzed. But consideration and modeling of its effect leave much room for improvement. That is mainly because method for grasping relations among companies with limited data is underdeveloped. In this article, chance discovery method is applied to estimate industrial relations that are to include companies’ relations that transmit chain reaction of bankruptcy. Time order method and directed KeyGraph are newly introduced to distinguish and express the time order among defaults that is essential information for the analysis of chain reaction bankruptcy. The steps for the data analysis are introduced and result of example analysis with default data in Kyushu, Japan, 2005 is presented. The structure estimated by the new method is compared with the structure of actual account receivable holders of bankrupted companies for evaluation.

Keywords

Chance discovery / credit risk / chain reaction / bankruptcy

References

[1] Cooper G., Herskovits E.. A Bayesian methods for the induction of probabilistic networks from data. Machine Learning, 1992, 9: 309-347.

[2] FISC. (1999). Working Report on Risk Management Model

[3] Goda, S. & Ohsawa, Y. (2006). Chance discovery in credit risk management — estimation of chain reaction bankruptcy structure by chance discovery method. In: Proceedings of the 2006 IEEE International Conference on Systems, Man, and Cybernetics (6 Volumes), October 8–11, 2006

[4] Motomura, Y. (1997). Modeling with large scale data by Bayesian networks. JSAI Technical Report SIG-CII-9707-04

[5] Motomura, Y. (1998). Bayesian network that learns conditional probabilities by neural networks. Workshop on Information-Based Induction Sciences (IBIS)

[6] Mathematical Systems (2006). Inc.: BayoNet 4.0 manual, http://www.msi.co.jp/BAYONET/files/manual.pdf

[7] Nakabayashi, A. & Sasaki, M. (1998). Models for credit risk measurement and its application to Japanese bank. Friday Review, 2 (2)

[8] Ohsawa, Y. (2005)a. Discovering risky subduction zones of earthquake. In: Information Technology for Chance Discovery, pp. 314–325. Tokyo-Denki University Press

[9] Ohsawa, Y. (2005)b. Visualizing relations between chances and surrounding events. In: Information Technology for Chance Discovery, pp. 121–153. Tokyo-Denki University Press

[10] Saunders, A. (2001). Credit Risk Measurement (Japanese), Kinyu Zaisei Jijo Press

[11] Shigematsu, K., Ueno, M. & Motomura, Y. (2006). An Introduction to Bayesian Networks, Baifukan

[12] Torii, H. (1998). Portfolio based credit risk management models — its effectiveness and limit. Weekly Kinyu Zaisei Jijo Magazine, June