Research on the Construction of a Financial Crisis Early Warning Model for Listed Companies from the Perspective of ESG: Taking the Lithium Battery Industry as an Example

Read full text
Authors: Yucong Liu
Cited by: 0

Abstract

Traditional financial crisis early warning models for listed companies rely predominantly on conventional financial indicators and often overlook ESG risk and its impact on corporate performance. This limitation becomes particularly evident in policy- and regulation-intensive industries such as lithium battery manufacturing, where existing models demonstrate insufficient accuracy. Our study develops a novel dynamic financial crisis early warning model that incorporates ESG factors to increase both precision and predictive power. Using 66 representative lithium battery companies from China’s securities market as research subjects, we select 21 financial metrics across six dimensions: debt repayment capacity, operational efficiency, growth potential, cash flow generation, and ESG performance. Through KMO test validation, 19 indicators were identified as model factors for entropy-weighted logit regression analysis. The findings reveal a significant negative correlation between ESG performance and financial risk. Validation through 30 randomly selected companies with similar ESG ratings demonstrates the model’s reliability, providing actionable insights for lithium battery industry risk management. This research explores theoretical and practical pathways for integrating ESG principles into financial crisis early warning systems, establishing a composite model that combines traditional financial signals with nonfinancial ESG indicators. The proposed framework offers investors, regulators, and corporate managers more forward-looking and comprehensive decision-making support.

Fields of Study & Subjects

Financial Distress and Bankruptcy Prediction Credit Risk and Financial Regulations Working Capital and Financial Performance

Publication details

Journal / source
Exploring Science Academic Conference Series
Volume
10
Pages
101–115
Publication date
6 January 2026