ESG Reputation Risk Matters: An Event Study Based on Social Media Data

[Controversies, reputation, and firm value 4/4]

Can a spike in negative tweets knock down a company's share price?

Maxime Nicolas, Adrien Desroziers, Fabio Caccioli, and Tomaso Aste analyse 114 million tweets about S&P 100 firms posted between 2016 and 2022 in "ESG Reputation Risk Matters: An Event Study Based on Social Media Data", published in Finance Research Letters.

They classify them with a purpose-built sustainability lexicon of ten MSCI-inspired categories, and define reputation-risk events as abnormal spikes in negative posting activity.

By excluding any event within five trading days of an earnings release or a controversy news item, they strip out the tangible costs of incidents and isolate the purely reputational channel.

Their main conclusions include:

• On the day of a sustainability-related reputation event on social media, abnormal returns drop by a statistically significant 0.29% on average, with no physical cost involved.
• The reaction is concentrated in the Social and Governance pillars. Over the two days around the event, social-risk and governance-risk events reduce stock prices by 0.36% and 0.33% respectively.
• Environmental-risk events show no significant overall effect, which suggests shareholders respond to the tangible costs of environmental incidents rather than to negative chatter about them.
• One environmental exception stands out: negative discussions about "Environmental Opportunities", covering clean tech, green building, and renewables, reduce stock prices by 0.34%.
• Within the pillars, "Product Liability" events cut prices by 0.32% and "Corporate Governance" events by 0.33%.
• Governance dominates the sample, with 495 of the 665 detected events, against 253 environmental and 79 social.

This article shows controversy feeds built on news providers carry a distinct, priced reputational signal, particularly for consumer-facing large caps.

It also adds to how sustainability affects firm value: as these events carry no direct financial cost, the measured reaction is evidence that markets price sustainability reputation as such, not only its physical consequences.

The sample however only covers S&P 100 larger caps with data ending in January 2022, before many recent changes at Twitter and it becoming X. Sector-level and cross-country evidence would usefully complete the picture.