Beyond ESG: A Screening Spectrum Investors Can Actually Verify

Unscreened, values-aligned and sustainability-objective holdings need to function as three distinct, auditable categories — mapped against SFDR Article 6, 8 and 9 disclosure tiers and IRIS+ metric sets. Blending them into a single label is the most common source of investor mistrust in this space.

The complaint about ESG is usually framed as a debate about whether it works. That is the wrong argument. The more immediate problem is that the term describes at least three materially different things, and the industry has been content to let one word cover all of them because the ambiguity is commercially convenient.

An investor asking "is this fund sustainable?" is generally asking one of three questions. Does it avoid things I object to? Is it deliberately aligned with outcomes I want? Or does it exist to produce an outcome that would not otherwise have occurred? Those are different mandates with different costs, different reporting obligations and different honest answers.

Three categories, not one spectrum of enthusiasm

Unscreened

No sustainability objective is asserted. The mandate is financial, the universe is unrestricted, and no impact claim attaches to the holdings. Under SFDR this corresponds to Article 6. The category is not a failure state — for many mandates it is the correct and honest classification, and describing it accurately is more useful than implying a screen that is not operating.

Values-aligned

The mandate applies exclusions or tilts that reflect the holder's stated preferences. It restricts a universe; it does not claim to cause an outcome. This maps to Article 8: sustainability characteristics are promoted, but no sustainable investment objective is claimed.

The critical discipline here is not overstating what an exclusion achieves. Declining to hold a security in the secondary market changes who owns it. Whether it changes anything else is a separate question, and one that a values-aligned mandate is not obliged to answer — provided it does not claim to.

Sustainability-objective

The mandate exists to produce a specified outcome, measured against a defined metric set, and reports against it. Article 9 in SFDR terms, with IRIS+ providing the metric vocabulary. This is the only one of the three categories entitled to make a causal claim, and it is the only one that has to survive the additionality test.

The additionality test

The question is simple to state and uncomfortable to apply: would this outcome have occurred without the capital?

In primary markets — a project financing, a growth round, a concessional tranche in a blended structure — the answer can be genuinely yes, and can be evidenced. The capital changed what was built, or when, or whether it cleared a financing hurdle at all.

In liquid secondary markets the honest answer is usually more modest. Buying a listed security from another holder transfers ownership. It does not, in itself, finance anything new. There are second-order arguments — cost of capital, index inclusion, engagement rights — and some are defensible, but they are arguments about influence rather than direct causation, and they should be described that way.

A screen is only meaningful if an investor can verify it. Everything else is a preference expressed at the point of sale.

Why blending the categories destroys trust

The failure mode is predictable. A fund applies a values-aligned exclusion list, reports against an Article 9-flavoured narrative, and describes the result in language that implies additionality. Nothing in that sequence is necessarily a lie. The composite is nonetheless misleading, because the claim being heard is stronger than the claim being made.

When an investor later discovers the distinction — usually through a journalist, a regulator or a redemption conversation — the damage is not confined to the fund. It attaches to the category. Each episode makes the next honest Article 9 mandate harder to sell, because the vocabulary has been spent.

What verifiable looks like in practice

  1. Publish the spectrum, not just the label. State which of the three categories each sleeve occupies, and what specifically each level excludes or targets.
  2. Name the metric set and the verification route. IRIS+ metrics, who collects the data, at what frequency, and who independently checks it. A methodology nobody audits is a description, not a measurement.
  3. State the data provenance. Estimated, reported, or independently assured — for every metric that carries a claim.
  4. Disclose the cost of the screen. What the mandate gives up on the indicators it does not target. A screen with no disclosed trade-off is almost always an undisclosed one.

None of this makes screening easier to sell. It makes it possible to defend, which over a long horizon is the more valuable property.