Per Berg, Koelbel & Rigobon (2020) on ESG-rating divergence: when methodology drifts silently, comparison becomes impossible and trust collapses. Aequara prevents this with a SHA-256 hash of every active methodology, published with every quarterly cycle. Operational amendments do not rotate the hash. Methodology changes do — and require declaring a new index family with parallel publication during the deprecation horizon.
Published 2026-04-01. Current as of 2026-05-10. Unchanged through the v2.1 release (2026-05-06) — that release was an operational amendment (governance, publication cadence, constituent-band tightening), not a methodology change. The hash you should expect to see in the Q3 2026 publication is identical.
Two columns matter: type (operational amendment vs methodology change) and hash rotation. Operational amendments are versioned without rotating the hash; methodology changes require a new index family and parallel publication.
Affects how the methodology is operated, but not what it measures. Examples: governance changes, publication cadence formalization, constituent-band tightening, holiday-adjustment rules, steward rotation policies. Approved by the editorial steward panel. Hash does not rotate. Versioned (v1.0.1 · v2.1.0 · etc.) for tracking.
Affects what the methodology measures or how the measurement is mathematically defined. Examples: weight-axis additions, scoring-formula changes, definition shifts, signal additions. Requires declaring a new index family (NPX-200, NPX-100v2, etc.) and a parallel publication cycle during a deprecation horizon. Hash rotates with the new family.
The point is not stasis — methodology improves, and improvement requires evolution. The point is that users always know which methodology they're looking at. They can compare 2026Q1 NPX-100 to 2027Q1 NPX-100 and trust they're comparing like to like. They can choose to migrate to NPX-100v2 if and when it ships, with a documented deprecation horizon for the original.
Methodology changes (the kind that rotate the hash) require co-signatures from at least three Tier III Custodians in addition to the editorial steward panel. This is structural: it prevents Aequara from quietly drifting its own methodology under commercial pressure. The Custodian community is small, named, and accountable — and they are the firewall against ourselves.