ESG and systemic risk

ABSTRACT

How do changes in Environmental, Social and Governance (ESG) scores influence banks’ systemic risk contribution? Using a dynamic panel model, we document a beneficial impact of the ESG Combined Score and Governance pillar on banks’ contribution to system-wide distress analysing a panel of 367 publicly listed banks from 47 countries over the period 2007–2020. Stakeholder theory and theory relating social performance to expected returns in which enhanced investments in corporate social responsibility mitigate bank-specific risks explain our findings. However, only better corporate governance represents a tool in reducing bank interconnectedness and maintaining financial stability. The results are robust to alternative measures of systemic risk, both contribution and exposure, as well as when estimating a static model. Our findings stress the importance of integrating banks’ ESG disclosure into regulatory authorities’ supervisory mechanisms as qualitative information.

KEYWORDS:

• Systemic risk
• financial stability
• Corporate Social Responsibility (CSR)
• environmental
• Social and Governance (ESG) scores

JEL CLASSIFICATION:

• G01
• G21
• M14

Acknowledgements

Andrieș and Sprincean acknowledge financial support from the Romanian National Authority for Scientific Research and Innovation, CNCS – UEFISCDI – PN-III-P4-ID-PCE2020-0929. Ongena acknowledges financial support from ERC ADG 2016 - GA 740272 lending.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ ESG scores are used in the empirical literature as proxies for Corporate Social Responsibility (hereafter CSR) performance (see, e.g. Sassen, Hinze, and Hardeck 2016; Chiaramonte et al. 2021). However, Gillan, Koch, and Starks (2021) point-out that ESG is a more expansive term that CSR, the main distinction between the two being that ESG incorporates governance aspects explicitly, whereas CSR includes governance in an indirect manner.

² More recently, Lööf, Sahamkhadam, and Stephan (2022) show that ESG ratings are associated with lower downside risk, but also with lower upside return potential.

³ However, as argued by Peloza (2006), insurance value from sustainable practices varies across industries in which companies operate, and also depends on their age.

⁴ Ticker LX4GLBK$.

⁵ Recent surveys are provided by Benoit et al. (2017) and Silva, Kimura, and Sobreiro (2017).

⁶ Xu, Hu, and Das (2019) employ separately regional and global state variables and find the results to be very similar.

⁷ Brownlees and Engle (2017) show that the findings are substantially stable when changing prudential capital ratio to different values.

⁸ Given the multitude of providers of ESG ratings, Berg et al. (2021) propose a noise-correction procedure and confirm the positive impact of sustainable activities on stock returns, the results being even stronger than previously estimated in the literature.

⁹ If an indicator is irrelevant for a particular sector, then it is excluded from the calculation.:

¹⁰ The weights are the following: CSR strategy −0.13; Management −0.67; Shareholders −0.20.

¹¹ When the ESG Controversies Score is higher than the ESG Score, then the ESG Score is equal to the ESG Combined Score.

¹² In other industries, such as fossil fuel firms, ESG activities are measured especially through carbon risk, proxied for instance by total fossil fuel reserves (Delis et al. 2021).

¹³ The index is constructed as equally weighted average of the six dimensions of governance indicators.

¹⁴ We thank a referee for pointing-out this issue.

¹⁵ Crisis periods are defined as 2008–2009 for non-European countries (Bouslah, Kryzanowski, and M’Zali 2018), and according to European Systemic Risk Board methodology for European states.

¹⁶ Small and large banks are defined based on the median size (total assets) of the sample in 2020, with the cut-off value being USD 55.55 billion.

Additional information

Funding

The work was supported by the H2020 European Research Council [ERC ADG 2016 - GA 740272 lending]; Romanian National Authority for Scientific Research and Innovation [CNCS – UEFISCDI – PN-III-P4-ID-PCE2020-0929].