An Inquiry into the Nature of Predictive Model Classes and the Forecasting of Corporate Insolvency

March 1, 2026Open Access

View Full Paper

Ask AI

Bookmark

Share

Key Points

This research aims to compare the predictive accuracy and interpretability of various machine learning models for estimating corporate insolvency risk.
Systematic comparison of logistic regression, Random Forest, XGBoost, and neural networks.
Analysis based on data from annual reports of Austrian non-financial firms.
Evaluation using ROC-AUC, PIT approach, and interpretability methods like SHAP values and PDPs.
XGBoost shows the highest predictive accuracy, outperforming logistic regression and other models.
Modern machine learning models demonstrate better explainability with the application of interpretability techniques.
Calibration techniques improve alignment between estimated and observed failure rates, particularly benefiting complex models.

Answers are grounded in this paper and cite their sources.

Compares predictive models to forecast corporate insolvency, indicating the effectiveness of modern algorithms.

No abstract yet. The journal hasn't published one we can show.

No takes yet. Share an insight, caveat, or question.

Mahmoud Abdussalem Sakka (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Share