This paper conducts a PRISMA-guided systematic review of the empirical literature on environmental, social, and governance (ESG) risk integration in European banking. Using evidence systematically retrieved from Scopus, ScienceDirect, IDEAS/RePEc, and SSRN, the review synthesizes 51 peer-reviewed and working studies published between 2020 and 2025, reflecting the recent and rapidly evolving nature of this research field. The analysis classifies the literature into three domains—pricing and allocation, monitoring and stress testing, and governance and management control systems—and evaluates whether ESG variables operate as first-order drivers within production credit-risk models. The results indicate that while ESG signals are increasingly incorporated into pricing decisions, stress-testing exercises, and governance frameworks, no study provides verifiable evidence of full model-level integration within Probability of Default (PD) or Loss Given Default (LGD) models. The contribution of this review lies in systematically identifying the structural, data-related, and supervisory constraints that sustain this integration gap and in proposing a roadmap that distinguishes incremental ESG sensitivity from genuine prudential model embedding. Overall, the findings clarify that ESG responsiveness in European banking is substantial, yet integration into core risk models remains limited.
Zervoudi et al. (Tue,) studied this question.