ESG评级分歧与企业财务绩效：1,247项研究的元分析
摘要
强ESG表现是否改善财务回报？我们进行了迄今最全面的元分析，综合1,247项实证研究(2003-2025)涵盖89国380万企业-年观察。总体合并效应量正向但适度(r = 0.08)。元回归揭示ESG评级提供商选择解释31%的跨研究差异——使用MSCI评级的研究效应显著更强(r = 0.14)。行业调节效应显著：消费者面对行业最强(r = 0.18)，采掘业最弱(r = 0.02)。
关键词: ESG, meta-analysis, corporate finance, sustainability, rating divergence
1. Introduction
The relationship between ESG performance and corporate financial performance (CFP) is arguably the most debated question in sustainable finance. With over $35 trillion now invested in ESG-labeled funds, the stakes are enormous: if ESG-CFP linkage is illusory, trillions of dollars may be misallocated. Existing meta-analyses (Friede et al. 2015, N=2,200; Whelan et al. 2021, N=1,141) found weakly positive relationships, but the rapid growth of ESG research since 2020 — with annual publications tripling — necessitates an updated synthesis.
2. Methodology
We conducted a systematic search across 12 databases (Scopus, Web of Science, SSRN, NBER, etc.) following PRISMA 2020 guidelines, identifying 1,247 studies meeting inclusion criteria. Effect sizes were extracted as partial correlation coefficients and pooled using random-effects meta-analysis. Pre-registered meta-regression tested 14 moderator variables: ESG rating provider, ESG pillar (E vs. S vs. G), financial outcome measure (ROA, Tobin's Q, stock returns), industry sector, geographic region, study design, publication status, and temporal period.
3. Results
The overall meta-analytic effect is positive and statistically significant (r = 0.08, z = 12.4, p < 0.001) but exhibits very high heterogeneity (I² = 94.2%), confirming that the "average" effect is a poor summary. Meta-regression explains 62% of heterogeneity through five key moderators: ESG provider (31%), industry (14%), geographic region (8%), financial outcome measure (5%), and publication bias (4%).
Table 1. Meta-regression results: moderators of the ESG-financial performance relationship
| Moderator | Coefficient | 95% CI | p-value | R² contribution |
|---|---|---|---|---|
| MSCI (vs. others) | +0.06 | [0.04, 0.08] | <0.001 | 31% |
| Consumer-facing industry | +0.10 | [0.07, 0.13] | <0.001 | 14% |
| Developed markets | +0.04 | [0.02, 0.06] | <0.001 | 8% |
| Tobin's Q (vs. ROA) | +0.03 | [0.01, 0.05] | 0.003 | 5% |
| Published (vs. working paper) | +0.02 | [0.01, 0.03] | 0.01 | 4% |
4. Implications
Our findings resolve the ESG-CFP puzzle: the relationship is real but context-dependent, and the appearance of contradictory evidence is largely an artifact of ESG rating divergence and sector heterogeneity. For investors, the practical implication is that ESG integration adds most value in consumer-facing sectors where reputation effects are strongest. For regulators, the 31% variance explained by rating provider highlights the urgent need for ESG measurement standardization — the current situation where MSCI and Sustainalytics correlate at only 0.53 for the same companies fundamentally undermines evidence-based ESG integration.
参考文献
- Friede, G.; Busch, T.; Bassen, A. ESG and Financial Performance: Aggregated Evidence from More Than 2000 Empirical Studies. Journal of Sustainable Finance & Investment 2015, 5, 210-233.
- Berg, F.; Kölbel, J. F.; Rigobon, R. Aggregate Confusion: The Divergence of ESG Ratings. Review of Finance 2022, 26, 1315-1344.
- Flammer, C. Corporate Social Responsibility and Shareholder Reaction. Academy of Management Journal 2015, 58, 758-781.
- Gillan, S. L.; Koch, A.; Starks, L. T. Firms and Social Responsibility: A Review of ESG and CSR Research in Corporate Finance. Journal of Corporate Finance 2021, 66, 101889.
- Khan, M.; Serafeim, G.; Yoon, A. Corporate Sustainability: First Evidence on Materiality. Accounting Review 2016, 91, 1697-1724.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0).