The independence thesis

What OVERT is established that the evidence must be verifiable by strangers. This page is about the harder discipline behind that: the standard only stays credible if Glacis refuses to profit from the risk the evidence prices.

The sentence to memorize

From the draft brochure, verbatim: GROUNDED · Brochure

“Glacis does not hold risk and does not aspire to become an insurer. OVERT’s value as an open standard depends entirely on Glacis remaining independent of the underwriting economics it enables. The moment Glacis holds a book, every carrier treats attestation receipts as a competitor’s proprietary format — and the standard collapses.”

Sit with the mechanism, because it is game-theoretic, not moral. A carrier will happily consume evidence from a neutral standard. The same carrier will never build its book on a data format owned by a party competing for that book — adopting it would hand a rival both the pricing signal and the customer relationship. Neutrality is not a virtue Glacis performs; it is the adoption condition for the entire market. STABLE

The role trio

The structural answer is a three-way separation, stated in the draft in three lines: GROUNDED · Brochure

Party	Role	What it earns	What it must never do
The Syndicate	Writes the paper — holds the risk	The underwriting profit	Depend on evidence only the evidence vendor can interpret
Glacis	Produces the evidence at the inference boundary	Earns per attested inference population	Hold a book, or aspire to
OVERT	Stays open and neutral	Nothing — that is the point	Belong to anyone

Each row is load-bearing. The syndicate keeps the profit because it carries the loss — Volume 1’s oldest rule (see delegated authority). Glacis is paid on the denominator — the attested inference population — so its incentive is to grow and harden the evidence base, not to select against it. And OVERT earns nothing because a standard with a P&L is a product with a lobby. GROUNDED · Brochure The “earns per attested inference population” model is a draft-application business design, not a signed commercial arrangement — hedge it as “the draft proposes” when selling.

There is precedent for every seat at this table: exchanges do not trade against their members, and the credibility of a rating agency rests on not owning the bonds it rates. The Glacis knowledge base carries the long-horizon framing as “Moody’s for AI risk” — an internal aspiration, so attribute it (“the team’s long-term framing is…”) rather than asserting it. GROUNDED · KB

Yes, this contradicts Volume 1 — on purpose

If you studied the three revenue streams, you learned a stream 2 in which Glacis itself becomes an MGA — sells and underwrites a Tech E&O policy on a carrier’s paper, which is why this academy spent a volume on producer licensing. The Inference Re draft explicitly renounces that: no book, no underwriting ambition, evidence economics only.

This is strategy evolution, not error, and you should be able to narrate it honestly: STABLE

The Volume 1 position came from the earlier one-pager era: license → MGA → underwriting profit, with the loss ratio as the flywheel hub.
The Volume 2 position is the brochure’s: the flywheel still turns on the loss ratio, but the syndicate owns that hub; Glacis powers it and gets paid on attested volume. Independence buys standard adoption, which is worth more than one MGA’s book.
The licensing knowledge is not wasted — you sell to underwriters now, and section 22-02 teaches how the reconciliation plays in a live conversation.

What independence buys, concretely

Trigger admissibility. A parametric clause pays on OVERT events. In a dispute, evidence under an open standard survives the argument “your vendor graded its own homework” (see parametric triggers). STABLE
Cross-carrier pooling. The accumulation problem is only solvable if competing syndicates feed one schema — which they will only do if none of them, and no risk-holding rival, owns it. GROUNDED · Brochure
The actuarial future. An experience base “an actuary can project from” requires a denominator that is “independently verified, not vendor-reported” — the ILS arc inherits its credibility from this page. GROUNDED · Brochure

The next page descends from why the standard must be independent to how a stranger actually verifies a record without asking Glacis anything.

→ 02 Chain of custody and audit access

Drill this page →18 bank questions stand behind what you just read. Check it while it’s warm.