MSCI, S&P Global, and Sustainalytics — the three agencies that shape how institutional investors assess environmental and social performance — have each made meaningful updates to their scoring frameworks over the past year or two. Some changes were announced. Most weren't widely tracked by the sustainability teams whose numbers they directly affect.
What's opening up is a gap between how companies believe they're performing and how they're actually being scored. And unlike a reputational gap, this one shows up in financing costs, investor access, and index inclusion.
MSCI updated its ESG Ratings methodology to place greater emphasis on industry-specific key issue weighting. What gets measured, and how heavily it counts, now varies more significantly by sector. A company that was scoring well on a broad environmental metric may find itself exposed on a sector-specific issue it never prioritized, with no change to actual operations.
S&P Global revised its ESG Scores to give increased weight to controversy assessments. That means how a company appears in third-party media, regulatory databases, and NGO reporting now carries more weight, not just what it self-reports. One enforcement action, one credible critical report, one sustained NGO flag can move a score in ways that even strong internal data won't offset.
Morningstar Sustainalytics has continued tightening its ESG Risk Ratings model, giving progressively less weight to company self-reported figures where independent verification is absent.
According to Morningstar Sustainalytics' methodology, where company disclosure is absent or insufficient, unverified data is supplemented by agency estimates based on scientific models, particularly for greenhouse gas (GHG) emissions. While these models are designed to fill data gaps, they may not align with a company’s own internal view of its performance.
Across all three, the direction is consistent: less weight on what you tell them, more weight on what they can confirm from somewhere else.
This is where many sustainability teams run into trouble. Improving performance doesn't automatically improve scores, and the assumption that it does creates blind spots.
Rating agencies pull from a wide range of sources beyond company disclosures: regulatory filings, government emissions registries, news databases, NGO publications, and proprietary data collection. Where data is missing or unassured, they apply estimation models that may or may not align with what a company has internally tracked.
If your emissions data isn't flowing through the channels agencies actually monitor, or hasn't been third-party verified, it may not be factoring into your score the way you'd expect. And it works the other way too: a single controversy flag in a widely used media database can suppress a score even when underlying performance is genuinely strong.
Sustainability teams that focus only on improving operations, without understanding how those improvements are being measured and by whom, are optimizing for a scorecard they're not actually reading.
Score-based filtering determines which investor mandates your securities can access. Green bond pricing is increasingly tied to third-party ESG ratings. Several major ESG indices use agency scores as primary criteria for inclusion, meaning a meaningful score drop can trigger forced selling by funds that track those indices. In sectors where large buyers have made Scope 3 commitments, ESG scores are also being embedded in supplier qualification and contract renewal processes — so the pressure isn't only coming from the investor side.
These aren't soft consequences. They show up in interest rates, investor conversations, and purchasing decisions.
Most sustainability teams can't answer precisely how each major agency is scoring their company, which data those agencies are using, or where the gaps are. That's the first problem to solve.
All three agencies publish methodology documentation. MSCI and Sustainalytics offer company-facing portals where you can see what data they hold and initiate a correction or engagement process. Working through that, understanding what's actually pulling your score down, is foundational work that a surprising number of organizations haven't done.
Third-party assurance on emissions data has also shifted from a credibility signal to a functional scoring input. Without it, agencies discount self-reported numbers or replace them with their own estimates.
And this conversation belongs in finance, not just sustainability. If ESG scores are affecting financing costs, index inclusion, or investor access, that's a financial risk question. Sustainability teams can own the data quality work. But the business implications need a broader table.
