24th Conference of the International Federation of Operational Research Societies
Abstract Submission

265. Evaluating the Robustness and Practical Utility of XAI-Driven Early Warning Systems in firm distress prediction

Invited abstract in session Business Management in Dynamic Emerging Markets, stream Selected Aspects of International Finance and OR.

Authors (first author is the speaker)

1. Trustlord Marecha
School of Business Management, University of Johannesburg
2. Helper Zhou
School of Accounting, Economics and Finance, University of KwaZulu Natal

Abstract

South African SMMEs face >70% failure rates in the first five years due to opaque risk signals. This study evaluates the real-world applicability and robustness of an XAI-driven Early Warning System (EWS) for firm distress prediction, moving beyond predictive accuracy to assess actionability, temporal resilience, and performance under data scarcity. Using 28,400 SMMEs (2016–2024), SHAP/LIME-augmented Gradient Boosting/LightGBM models provide actionable insights. The XAI-EWS emerges as a deployable tool for lenders and agencies.

Keywords

Status: accepted


Back to the list of papers