This thesis aims to enhance corporate risk assessment through studying ESG (Environmental, Social, and Governance) significance in modeling bankruptcy. Through using artificial intelligence’ natural language processing (NLP), we develop a proxy for ESG scoring based on companies annual reporting retrieved through EDGAR. We integrate the derived ESG score with traditional financial ratios used to calculate Altman’s Z-score (1968) in predicting bankruptcy. Through S&P’s Compustat, we obtain a sample of 108 healthy & bankrupt firms -spanning 14 years of fiscal observations- and match them through time and industry. We use stepwise GLM regression to estimate bankruptcy probability observing that ESG and its interaction terms proved significant in bankruptcy prediction, and that the average marginal effect of ESG is negatively correlated with bankruptcy indicating ESG’s positive impact on firm performance. Our research observes ESG metrics for a balanced sample of healthy and bankrupt firms, which despite restricting sample size protects our study against survivorship bias unlike current literature that considers large datasets without factoring ESG for failed firms. We recommend future research to increase bankrupt firms' data points, consider their ESG metrics, and account for the impact of economic downturns to further understand ESG’s trajectory on firm performance.


School of Business


Management Department

Degree Name

MS in Finance

Graduation Date

Winter 1-31-2025

Submission Date


First Advisor

Rim Cherif

Second Advisor

Malek Ben Abdellatif

Committee Member 1

Tarek Eldomiaty

Committee Member 2

Malek Ben Abdellatif

Committee Member 3

Rim Cherif


50 p.

Document Type

Master's Thesis

Institutional Review Board (IRB) Approval

Not necessary for this item