Bankruptcy prediction for SMEs using transactional data and two-stage multiobjective feature selection

https://doi.org/10.1016/j.dss.2020.113429Get rights and content
Under a Creative Commons license
open access

Highlights

  • A bankruptcy prediction model for SMEs that uses transactional data under a scenario where no accounting data are required.

  • Offline and online test results both confirm that transactional data–based variables improve SME bankruptcy prediction.

  • A two-stage multiobjective feature-selection method and compare it with other benchmark methods.

Abstract

Many bankruptcy prediction models for small and medium-sized enterprises (SMEs) are built using accounting-based financial ratios. This study proposes a bankruptcy prediction model for SMEs that uses transactional data and payment network–based variables under a scenario where no financial (accounting) data are required. Offline and online test results both confirmed the predictive capability and economic benefit of transactional data–based variables. However, incorporating those features in predictive models produces high dimensional problems, which deteriorates model interpretability and increases feature acquisition costs. Thus, we propose a two-stage multiobjective feature-selection method that optimizes the number of features as well as model classification performance. The results showed that the proposed model achieved similar classification performance while greatly reducing the cardinality of the feature subset. Finally, the feature importance evaluation for features in the optimal subset confirmed the importance of transactional data and payment network-based variables for bankruptcy prediction.

Keywords

Bankruptcy prediction
Payment and transactional data
Expected maximum profit
Data imbalance
Feature selection

Cited by (0)

Gang Kou is a Distinguished Professor of Chang Jiang Scholars Program in Southwestern University of Finance and Economics, managing editor of International Journal of Information Technology & Decision Making (SCI) and managing editor-in-chief of Financial Innovation (SSCI). He is also editors for the following journals: European Journal of Operational Research, and Decision Support Systems. Previously, he was a professor of School of Management and Economics, University of Electronic Science and Technology of China, and a research scientist in Thomson Co., R&D. He received his Ph.D. in Information Technology from the College of Information Science & Technology, Univ. of Nebraska at Omaha; Master degree in Dept of Computer Science, Univ. of Nebraska at Omaha; and B.S. degree in Department of Physics, Tsinghua University, China. He has published more than 100 papers in various peer-reviewed journals.

Yong Xu is currently a Ph.D. candidate in School of Business Administration, Southwestern University of Finance and Economics. His research interests include credit scoring and machine learning, particularly cost-sensitive learning.

Yi Peng is a Chang Jiang Scholar Chair Professor of School of Management and Economics, University of Electronic Science and Technology of China. Previously, she worked as Senior Analyst for West Co., USA. Dr. Peng received her Ph.D. in Information Technology from the College of Information Science & Technology, Univ. of Nebraska at Omaha and got her Master degree in Dept of Info. Science & Quality Assurance, Univ. of Nebraska at Omaha and B.S. degree in Department of Management Information Systems, Sichuan University, China. Dr. Peng's research interests cover Knowledge Discover in Database and data mining, multi-criteria decision making, data mining methods and modeling, knowledge discovery in real-life applications. She has published more than eighty papers, which have been cited for several thousand times, in various peer-reviewed journals such as Omega. Yi Peng is listed as the Highly Cited Researcher by Clarivate Analytics (Web of Science).

Feng Shen is a professor of School of Finance, Southwestern University of Finance and Economics. He received his Ph.D. from the School of Business, Sichuan University in 2015. His current research interests include financial risk management, credit scoring, machine learning and data mining.

Yang Chen is a Professor in School of Business Administration, Southwestern University of Finance and Economics in China. He received his Ph.D. in Finance and Decision Sciences from Hong Kong Baptist University. His current research interests include the dark side of social media, human resource management, and so on. He has published research papers in journals such as Decision Sciences, Decision Support Systems, European Journal of Information Systems, Human Resource Management, Information & Management, Journal of Business Ethics, Journal of Information Technology, and Journal of Product Innovation Management. He is currently serving as the Associate Editor for Internet Research.

Kun Chang, General Manager of Department of Data Application, Shandong City Commercial Banks Alliance Co., LTD, Jinan, 250104, China.

Shaomin Kou, Assistant General Manager of Department of Data Application, Shandong City Commercial Banks Alliance Co., LTD, Jinan, 250104, China.