ESG-Stability Factor Performance : Constructing and Testing ESG Stability and Momentum in Sector-Neutral Portfolios

This thesis tests whether the time‐series behaviour of firms’ environmental, social, and governance (ESG) scores contains priced information in U.S. large-cap equities. Using a monthly panel of S\&P 500 constituents, I form sector-neutral, value-weighted long–short portfolios based on two signals: ESG momentum (recent improvement in an ESG score) and ESG stability (low variability of that score over time). To reduce mechanical tilts, scores are aligned for a risk-rating polarity change, winsorised cross-sectionally, and rank-neutralised by size and sector. Performance is evaluated with standard asset-pricing tests—Fama–French five-factor regressions and models augmented with price momentum and a defensive/low-volatility factor—using Newey–West inference. Model adequacy is further assessed with Gibbons–Ross–Shanken (GRS) tests. The evidence points to a robust premium for ESG stability: portfolios long stable scorers and short unstable scorers earn positive, statistically significant abnormal returns that are not explained by standard factors. By contrast, ESG momentum does not deliver positive abnormal performance; at intermediate horizons it underperforms. GRS tests indicate that adding the stability factor improves the joint pricing of test portfolios relative to FF5 alone. Overall, the results suggest that firms with smoother, less noisy ESG trajectories outperform those with volatile trajectories, while recent ESG score improvements do not translate into excess returns. The findings are robust across reasonable window choices and weighting schemes and highlight stability as an incremental source of cross-sectional variation in returns.