Carbon-Policy Surprises, Fossil Dependence, and Forward Downside Risk: Evidence from International Equity Markets

Abstract

Unexpected carbon-policy information may affect international equity markets unevenly, but a credible test requires outcomes measured after the information arrives and exposures fixed before the event sample. Using a primary sample of 15 national equity markets and 45 EU ETS announcements that contribute complete forward country-event windows, we construct five-trading-day forward downside semideviation from returns on days +1 to +5 and measure fossil dependence with each country’s fixed 2010 fossil-energy share. Country and event fixed effects isolate cross-country heterogeneity. The Surprise × Fossil interaction is positive (β = 0.975, SE = 0.389, p = 0.016) and remains significant under restricted wild-cluster bootstrap inference by event (p = 0.026) and country (p = 0.041). A one-standard-deviation surprise implies about 3.3 basis points more downside semideviation at the 75th than at the 25th percentile of fossil dependence, approximately 5.8% of the baseline mean. The result is also present in an expanded 27-market robustness sample (β = 0.803, p < 0.001). A predetermined listed-market Energy + Materials exposure does not absorb the fossil interaction: in the joint primary-sample model, Surprise × Fossil remains positive (β = 0.959, p = 0.014), whereas Surprise × Listed-Market Exposure is insignificant (β = −0.108, p = 0.801). In 2020–May 2024, the estimated Surprise × Fossil interaction is small and not statistically distinguishable from zero (β = −0.075, p = 0.788). The formal post-2019 slope change is negative (β = −1.050), although statistical support for the decline depends on the inference method (event-cluster wild-bootstrap p = 0.035; country-cluster wild-bootstrap p = 0.125). Overall, the evidence supports a pre-2020 structural-vulnerability gradient in downside severity, with evidence of later attenuation that is sensitive to the inference procedure rather than a uniform destabilizing effect of carbon policy. This distinction is relevant to sustainable-finance assessments of transition risk.