Front Matter

Copyright Page

© 2022 International Monetary Fund

WP/22/153

IMF Working Paper

Strategy, Policy and Review Department

Understanding and Predicting Systemic Corporate Distress: A Machine-Learning Approach

Prepared by Burcu Hacibedel and Ritong Qu1*

Authorized for distribution by Martin Cihak and Daria Zakharova

July 2022

IMF Working Papers describe research in progress by the author(s) and are published to elicit comments and to encourage debate. The views expressed in IMF Working Papers are those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.

ABSTRACT:

In this paper, we study systemic non-financial corporate sector distress using firm-level probabilities of default (PD), covering 55 economies, and spanning the last three decades. Systemic corporate distress is identified by elevated PDs across a large portion of the firms in an economy. A machine-learning based early warning system is constructed to predict the onset of distress in one year’s time. Our results show that credit expansion, monetary policy tightening, overvalued stock prices, and debt-linked balance-sheet weaknesses predict corporate distress. We also find that systemic corporate distress events are associated with contractions in GDP and credit growth in advanced and emerging markets at different degrees and milder than financial crises.

article image

Contents

  • 1 Introduction

  • 2 Data

    • 2.1 Constructing Economy-level PD Indices

    • 2.2 Predictors of Systemic Corporate Distress

  • 3 Identifying Corporate Sector Distress

    • 3.1 A Markov-Switching Model for PD Indices

    • 3.2 Model Estimates and Periods of High Corporate Sector Distress

  • 4 An Early-warning System for Corporate Distress

    • 4.1 Model Combination

    • 4.2 H-block Cross-validation

    • 4.3 Model Performance

    • 4.4 Interpreting Model Predictions with Shapley Values

      • 4.4.1 An Application: Corporate Distress Risk Index for Emerging Markets

  • 5 Macroeconomic Implications of Systemic Corporate Distress: Initial Findings

  • 6 Conclusion

  • Tables and Figures

  • Appendix A MCMC Algorithm to Identify Corporate Distress

  • Appendix B Constructing Predictors from Compustat Global

  • Appendix C Machine Learning Models and Hyperparameter Selection

    • C.1 Logistic Regression with Regularization

    • C.2 Random Forest

    • C.3 Support Vector Machine

    • C.4 Linear Discriminant Analysis

    • C.5 Extreme Gradient Boosting Tree

*

The views expressed in this paper are those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management. We would want to thank Bruno Albuquerque, Jorge Antonio Chan-Lau, Sophia Chen, Salih Fendoglu, Marco Gross, Weining Xing and seminar participants at the IMF for helpful comments; Chuqiao Bi, Ruofei Hu, Roshan Iyer and Jose Marzluf for excellent research assistance. Special thanks to Bruno Albuquerque for helping us understand Compustat data. All the errors are our own.

Understanding and Predicting Systemic Corporate Distress: A Machine-Learning Approach
Author: Ms. Burcu Hacibedel and Ritong Qu