A two-stage Bayesian network model for corporate bankruptcy prediction

Cao, Y;Liu, XQ;Zhai, J;Hua, S

[Cao, Yi] Univ Edinburgh, Management Sci, Sch Business, Edinburgh, Midlothian, Scotland.
[Liu, Xiaoquan] Univ Nottingham Ningbo, Ningbo, Zhejiang, Peoples R China.
[Zhai, Jia] Xian Jiaotong Liverpool Univ, Sch Business, Suzhou, Jiangsu, Peoples R China.
[Hua, Shan] Univ Surrey, Surrey Business Sch, Guildford, Surrey, England.

INTERNATIONAL JOURNAL OF FINANCE & ECONOMICS

DOI:10.1002/ijfe.2162

Publication Year:2020

JCR:Q2

ESI Discipline:ECONOMICS & BUSINESS

Latest Impact Factor:3.07

Document Type:Others

Identifier:http://hdl.handle.net/20.500.12791/005090

Abstract

We develop a Bayesian network (LASSO-BN) model for firm bankruptcy prediction. We select financial ratios via the Least Absolute Shrinkage Selection Operator (LASSO), establish the BN topology, and estimate model parameters. Our empirical results, based on 32,344 US firms from 1961-2018, show that the LASSO-BN model outperforms most alternative methods except the deep neural network. Crucially, the model provides a clear interpretation of its internal functionality by describing the logic of how conditional default probabilities are obtained from selected variables. Thus our model represents a major step towards interpretable machine learning models with strong performance and is relevant to investors and policymakers.

Keywords

Bayesian network accounting ratios sensitivity analysis LASSO interpretability analysis

Copyright 2006-2020 © Xi'an Jiaotong-Liverpool University 苏ICP备07016150号-1 京公网安备 11010102002019号