Search

This paper analyzes the prediction performance of human resources (HR) variables in corporate failure modeling. We define corporate failure as a two‐phase process from financial distress to bankruptcy, so that we can determine the prediction power of HR variables along a firm's phase in the financial deterioration process. We demonstrate the use of HR variables and their application to a two‐phase corporate failure model, providing first evidence for the predictive power of HR variables. The experimental results, based on real‐world datasets from Belgium, show that HR variables used in conjugation with accounting‐based information improve the accuracy of prediction modeling. However, the predictive power of HR variables varies in different phases of corporate failure with better prediction accuracy during the initial symptoms of corporate failure (i.e., financial distress). Findings show that our proposed model predicted financial distress with 84.1%, whereas the accuracy decreased to 83.3% when predicting bankruptcy. Besides, they also show that, on average, the inclusion of HR variables improves the global accuracy of the prediction models of 3.8% and allows to decrease Type I error of 5%.

Previous studies of bankruptcy prediction in imbalanced datasets analyze either the loss of prediction due to data imbalance issues or treatment methods for dealing with this issue. The current article presents a combined investigation of the degree of imbalance, loss of performance, and treatment methods. It determines which imbalanced class distributions jeopardize the performance of bankruptcy prediction methods and identifies the recovery capacities of treatment methods. The results show that an imbalanced distribution, in which the minority class represents 20%, significantly disturbs prediction performance. Furthermore, the support vector machine method is less sensitive than other prediction methods to imbalanced distributions, and sampling methods can recover a satisfactory portion of performance losses. Accordingly, this study provides a better understanding of the data imbalance issue in the field of corporate failure and serves as a methodological guide for designing bankruptcy prediction methods in imbalanced datasets.

This paper proposes a methodology for identifying data samples that are likely to be mislabeled in a c-class classification problem (dataset). The methodology relies on an assumption that the generalization error of a model learned from the data decreases if a label of some mislabeled sample is changed to its correct class. A general classification model used in the paper is OP-ELM; it also provides a fast way to estimate the generalization error by PRESS Leave-One-Out. It is tested on two toy datasets, as well as on real life datasets for one of which expert knowledge about the identified potential mislabels has been sought.

We examine the impact of two types of earnings management, accruals and real activities manipulation, on bankruptcy prediction models. Consider that the form of engaging on earnings management might characterize firms’ financial condition, we hypothesize and find that the special information extracted from earnings management models, which encompass potential manipulations, serves to improve the accuracy of bankruptcy prediction models. We further provide evidences that when potential accruals manipulation is measured, it allows an improvement of bankrupt firms, while real activities measures, especially sales manipulation, enhance the prediction improvement of non-bankrupt firms. Our finding suggests that failed firms are more likely to engage on accruals manipulation, while healthy firms do it on real activities manipulation. As a result, this study contributes to the literature by analyzing a novel explanatory variable, earnings management variable, to predict bankruptcy that can overcome financial alteration and can be applied to any type of firms.