Master’s research / Financial markets
When climate policy reverses, how do markets respond?
Lower compliance costs might benefit firms. Policy instability might also increase risk. This research examines how equity markets reacted across four stages of US environmental deregulation.
The research in numbers
- 929 firms in the event study: 501 US, 333 European and 95 Asian.
- 899 firms and 3,589 firm-event observations in the regression sample after exclusions.
- Four policy events from January 2025 to February 2026.
- Master’s thesis grade: 1.0. Related manuscript prepared for ZfU.
From a policy question to a dataset
The manuscript uses adjusted closing-price extracts and a firm-characteristics snapshot. It aligns each security with the URTH market benchmark on their common trading dates and converts market capitalisation into a consistent USD measure. Aligning calendars avoids treating a market closure as a zero-return observation.
Measuring an unusual market reaction
A market model estimates each firm’s usual relationship with the benchmark using 250 common trading days and a ten-day gap before the event. The difference between observed and model-expected returns is the abnormal return. Summing those differences gives cumulative abnormal return, or CAR.
The analysis compares three event windows: one, three and five common trading days either side of the event. A pooled regression then examines event and regional differences, including event-by-region interactions, sector effects and firm characteristics, with standard errors clustered by firm.
What the manuscript reports
The proposed-rule stage provides the clearest evidence of negative and heterogeneous repricing under the primary formal-publication date. The broader pattern challenges a uniform regulatory-relief interpretation. It is consistent with policy instability being relevant to valuation, but does not establish the transmission mechanism causally.
Regional attribution needs care: using the public-announcement date instead of the formal-publication date changes the US and European estimates. The positive Asian interaction over the wider window is a relative regression result, not a claim that every Asian firm gained.
Testing the interpretation
- Compare announcement and formal-publication dates.
- Re-estimate across short and wider event windows.
- Compare firm-clustered and HC3 standard errors.
- Check firm fixed effects and estimation windows that exclude earlier study events.
- Diagnose why aggregate event-level economic policy uncertainty cannot be separately estimated alongside event indicators.
What the evidence cannot settle
Four events provide limited scope to separate policy news from other simultaneous news. Regional samples are uneven, and firm-level clustering does not remove all dependence across firms facing the same event. The fundamentals snapshot also post-dates the events, so those controls describe firms rather than establish predetermined causal effects.
Why this matters for analytical work
The research connects data preparation, statistical modelling and careful communication. Calendar alignment, sample definitions, model identification and sensitivity checks determine what an analyst can responsibly conclude from an apparently simple chart.