Each of the 11 energy-sector assets carries an annual volatility σ, expected return μ, and an ethics score (0–100, built from emissions intensity, community/labour impact and energy-access benefit — see the article's sections on energy justice, ethical investment and corporate responsibility).
Raising the ethics floor removes low-scoring assets (coal, oil & gas, unabated gas peakers) from the investable set. The engine draws 6000 random long-only weight combinations over whatever remains, computes each portfolio's
σp² = wᵀ Σ w μp = wᵀ μ
using a correlation matrix (fossil assets move together, renewables move together, the two groups are weakly linked), then reads off the best achievable return at your chosen target risk from the upper edge of that cloud — the efficient frontier.
- X axis — portfolio risk (annual volatility, %)
- Y axis — portfolio expected return (%/yr)
- Z axis / colour — portfolio's weighted-average ethics score
The gap between the unscreened frontier and the screened one is the real, quantified "cost of ethics" — and how small or large it is as the floor rises is the actual tradeoff behind ESG investing.