Journal Article

A Hybrid Early Warning System for Corporate Financial Distress

Yacoub, Mostafa Mohamed, Abo Alkomsan, Asmaa Ahmed

2025

Overview

Financial distress prediction has become a focal research area in corporate finance, driven by the imperative of risk management and informed decision-making. This study proposes a hybrid modeling framework that integrates the Altman Z-score-a longstanding traditional financial metric-with four machine learning algorithms (Random Forest, Support Vector Machine, Logistic Regression, and XGBoost). The framework was empirically tested on companies listed on the Egyptian Exchange, as applied on a sample of 48 companies across five sectors, according to the nature of each of these sectors in terms of type of industry, as follows: food, beverages and tobacco sector, manufacturing sector, health care & pharmaceuticals sector, real estate sector, and services sector over the period from 2019 to 2023. By assigning determined weights to each component, the model describes both linear and non-linear patterns in corporate data. Empirical analysis, spanning multiple sectors, demonstrates the framework's predictive accuracy, achieving a 97.91% success rate, and highlights significant sector-specific distress patterns. These findings underscore the potential of combining established financial theory with modern computational techniques, offering robust and interpretable early warning systems for investors, financial institutions, and regulators.

• Publisher: جامعة المنصورة - كلية التجارة
• Subject: الأزمات المالية / البورصة المصرية / التعليم الآلي