Market-Implied Sustainability: Insights from Funds' Portfolio Holdings

Battle Count: 76

Highly relevant for quantitative portfolio management. The paper demonstrates that MIS-based portfolio tilting strategies improve risk-adjusted performance (lower VaR, CVaR, EVaR, and higher Sharpe ratios) compared to ESG-based strategies. The portfolio optimization framework (mean-CVaR, mean-EVaR, max-Sharpe) with sustainability constraints is directly applicable to systematic trading and asset allocation. The quarterly rebalancing with transaction costs (15 bps) provides practical implementation guidance. However, the strategies are tilting-based rather than pure alpha-generating, and the signal is derived from fund holdings rather than direct market data.

Moderate complexity. The MIS score computation is straightforward (difference of proportions). The quantile regression (MMQR) requires specialized econometric knowledge. Portfolio optimization with multiple risk measures (CVaR, EVaR) and constraints requires linear/quadratic programming expertise (Mosek solver). The main practical challenge is data acquisition (LSEG licensed data for fund holdings, ESG scores, and SFDR classifications). The bivariate quantile selection for corner strategies adds algorithmic complexity.

Methodology is well-documented with clear formulas and algorithmic steps. However, data is licensed from LSEG (formerly Refinitiv), limiting full reproducibility. The SFDR classification is taken from a single period (2023), and the code is implemented in Matlab 2025b with Mosek 10.1, which are proprietary tools. No public code repository is mentioned.