A two‐stage Bayesian network model for corporate bankruptcy prediction

We develop a Bayesian network (LASSO-BN) model for firm bankruptcy prediction. We select fnancial 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 th...

Full description

Bibliographic Details
Main Authors: Cao, Yi, Liu, Xiaoquan, Zhai, Jia, Hua, Shan
Format: Article
Language:English
Published: John Wiley and Sons Ltd 2020
Subjects:
Online Access:https://eprints.nottingham.ac.uk/61457/
_version_ 1848799878993412096
author Cao, Yi
Liu, Xiaoquan
Zhai, Jia
Hua, Shan
author_facet Cao, Yi
Liu, Xiaoquan
Zhai, Jia
Hua, Shan
author_sort Cao, Yi
building Nottingham Research Data Repository
collection Online Access
description We develop a Bayesian network (LASSO-BN) model for firm bankruptcy prediction. We select fnancial 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.
first_indexed 2025-11-14T20:42:40Z
format Article
id nottingham-61457
institution University of Nottingham Malaysia Campus
institution_category Local University
language English
last_indexed 2025-11-14T20:42:40Z
publishDate 2020
publisher John Wiley and Sons Ltd
recordtype eprints
repository_type Digital Repository
spelling nottingham-614572020-08-28T01:19:25Z https://eprints.nottingham.ac.uk/61457/ A two‐stage Bayesian network model for corporate bankruptcy prediction Cao, Yi Liu, Xiaoquan Zhai, Jia Hua, Shan We develop a Bayesian network (LASSO-BN) model for firm bankruptcy prediction. We select fnancial 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. John Wiley and Sons Ltd 2020-08-10 Article PeerReviewed application/pdf en cc_by https://eprints.nottingham.ac.uk/61457/1/A%20two-stage%20Bayesian%20network%20model%20for%20corporate%20bankruptcy%20prediction.pdf Cao, Yi, Liu, Xiaoquan, Zhai, Jia and Hua, Shan (2020) A two‐stage Bayesian network model for corporate bankruptcy prediction. International Journal of Finance & Economics . ISSN 1076-9307 Bayesian network; LASSO; Accounting ratios; Sensitivity analysis; Interpretability analysis http://dx.doi.org/10.1002/ijfe.2162 doi:10.1002/ijfe.2162 doi:10.1002/ijfe.2162
spellingShingle Bayesian network; LASSO; Accounting ratios; Sensitivity analysis; Interpretability analysis
Cao, Yi
Liu, Xiaoquan
Zhai, Jia
Hua, Shan
A two‐stage Bayesian network model for corporate bankruptcy prediction
title A two‐stage Bayesian network model for corporate bankruptcy prediction
title_full A two‐stage Bayesian network model for corporate bankruptcy prediction
title_fullStr A two‐stage Bayesian network model for corporate bankruptcy prediction
title_full_unstemmed A two‐stage Bayesian network model for corporate bankruptcy prediction
title_short A two‐stage Bayesian network model for corporate bankruptcy prediction
title_sort two‐stage bayesian network model for corporate bankruptcy prediction
topic Bayesian network; LASSO; Accounting ratios; Sensitivity analysis; Interpretability analysis
url https://eprints.nottingham.ac.uk/61457/
https://eprints.nottingham.ac.uk/61457/
https://eprints.nottingham.ac.uk/61457/