Journal article 192 views
The role of feature importance in predicting corporate financial distress in pre and post COVID periods: Evidence from China
Shusheng Ding 
,
Tianxiang Cui ,
Anthony Graham Bellotti,
Abedin Abedin,
Brian Lucey
,
Tianxiang Cui ,
Anthony Graham Bellotti,
Abedin Abedin,
Brian Lucey
International Review of Financial Analysis, Volume: 90, Start page: 102851
Swansea University Author: Abedin Abedin
DOI (Published version): 10.1016/j.irfa.2023.102851
Abstract
The prediction of firm financial distress during the COVID-19 crisis episode attracted massive academic attention since economic uncertainty was exacerbated. In this paper, we propose a firm financial distress prediction model based on the Extreme Gradient Boosting-Genetic Programming (XGB-GP) frame...
| Published in: | International Review of Financial Analysis |
|---|---|
| ISSN: | 1057-5219 1873-8079 |
| Published: |
Elsevier BV
2023
|
| Online Access: |
Check full text
|
| URI: | https://cronfa.swan.ac.uk/Record/cronfa64218 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract: |
The prediction of firm financial distress during the COVID-19 crisis episode attracted massive academic attention since economic uncertainty was exacerbated. In this paper, we propose a firm financial distress prediction model based on the Extreme Gradient Boosting-Genetic Programming (XGB-GP) framework by investigating subsamples of pre-COVID and post-COVID periods. The key contribution of our paper is that we explore time-varying prediction features for pre-COVID and post-COVID periods. We illuminate that the earning financial indicator is the dominant feature for financial distress prediction during the pre-COVID period, whereas total financial leverage is the most important factor during the post-COVID period. On this basis, our XGB-GP financial distress prediction model exhibits higher prediction accuracy than the traditional models. As a result, managers can modify the financial leverage level to improve the financial situation of the firm by reducing the debt burden and increasing profitability during the post-COVID period. |
|---|---|
| Keywords: |
Financial distress prediction, Time-varying feature selection, Extreme gradient boosting, Genetic programming, COVID-19 crisis |
| College: |
Faculty of Humanities and Social Sciences |
| Start Page: |
102851 |