Sharing the upside

The house’s answer to the only economics question that matters: how do you get paid like an underwriter without becoming one? The answer is an instrument stack that scales with premium while the risk stays on someone else’s paper — plus the discipline (three kills, a tripwire, a neutrality rule) that keeps the house from trading its standard for a book.

The instrument stack

Three instruments, papered together, in escalating order of commitment: GROUNDED · Glacis

Instrument	What it pays	Why the counterparty signs
Per-attested-insured fee	A fee per policy on the attested book — oracle economics, recurring for the life of the program	The cover is unwritable, or unpriceable, without the independent signal
Profit commission	A contingent share of underwriting profit on the book the evidence improves	Pay-for-performance: the carrier shares upside only if the delta is real
Pre-negotiated participation option	The right, papered in advance, to take risk via quota share or sidecar once the loss delta is demonstrated	Costs the carrier nothing today; commits nobody to anything but a structure

The label on the deal matters less than the instrument inside it. And note who pays for what: the vendor pays a subscription for attestation; the carrier shares underwriting economics. The insurer is a channel — never a billed customer for the attestation itself. Billing the channel is how plumbing turns into competition. GROUNDED · Glacis

Quota share and sidecars, precisely

• Quota share — proportional reinsurance: a fixed percentage of every premium dollar and every loss dollar on a defined book. The simplest way to hold a slice of performance. STABLE
• Sidecar — a ring-fenced vehicle, often funded by third-party capital, that takes a defined share of a specific book for a defined period. Bounded, inspectable, unwindable. STABLE

The house’s use of both is deliberately deferred: the option is negotiated now and exercised only on evidence. That is how the economics of risk participation get captured without the balance sheet, the claims shop, or the standards war arriving early. GROUNDED · Glacis

Parametric triggers and oracle economics

The parametric product (parametric triggers) is the wedge for affirmative cover because it needs no loss history: it prices the probability of a defined, attested signal crossing a threshold — computed from the house’s own telemetry — and deletes the causation problem at claim time. A cover priced on a signal is structurally unwritable without an independent oracle to produce the signal, and the oracle is paid per policy, forever. GROUNDED · Glacis

The parametric lane runs through Testudine Syndicate Services, whose caveats the house treats as load-bearing rather than as objections: GROUNDED · Glacis

• insurance is annual — a signal that only helps mid-term is helping at the wrong moment;
• a pre-binding signal (informs selection and pricing at submission) is worth more than a post-binding one (monitors a risk already on the books);
• parametric is a blunt tool — basis risk is where parametric products go to die (20-00);
• Testudine is not a carrier: capacity assembled through syndicates at the Exchange is slow.

The layer cake

The house’s ordering doctrine for the whole commercial arc: GROUNDED · Glacis

1. Own the evidence layer and the standard — the receipts, and OVERT.
2. Co-own the pricing vernacular — one case study plus one rating plan, built with a carrier, so the market prices AI risk in the house’s units.
3. Rent the risk capital — the participation option above, exercised late if at all.

Never invert the order. Renting capital before the vernacular exists makes a small MGA with no moat; building vernacular before the evidence layer exists makes a consultancy. The triangulation is the standards-laboratory-and-data-bureau position, not the inspector-who-became-an-insurer position: history’s inspection houses earned the right to hold risk on decades of proprietary loss data, and the modern MGAs that launched holding risk did it with veteran insurance benches and reinsurance paper from day one. The house has neither — yet — which is the point of the next section. GROUNDED · Glacis

The three kills

Why not become the carrier or MGA now? Three independent kills; any one suffices. GROUNDED · Glacis

1. No proven loss-ratio delta. The capacity conversation opens with show me the delta, and the honest answer today is that the measurement clock has only just started. An MGA without a loss-ratio delta is a broker with extra steps (page 00).
2. Taking risk forfeits the standard. OVERT’s entire architecture is independence. The moment the house holds a book, every other carrier treats receipts as a competitor’s proprietary format; Grossmünster Re, Aegle Assurance, and the Meridian Rating Bureau are handed a standards war they would happily fight; and a seat at the regulator’s table becomes untenable for a market participant. Carrier-enforced distribution exists only while carriers see the house as plumbing, not competition.
3. It is a second company. Capital, fronting relationships, a claims operation, and an insurance bench the house does not have. The precedents that look encouraging are precisely the ones that prove the kill — see the layer cake above.

The tripwire — and the two kill conditions

The MGA option is not dead; it is parked behind a tripwire: GROUNDED · Glacis

• evidence thresholds — a large attested-inference base, a demonstrated catch rate, and a carrier profit-share statement showing the delta in a counterparty’s own numbers;
• an insurance operator hired — the bench kill answered before the option can matter;
• and the trigger: the option is exercised only if a carrier refuses to share demonstrated economics. It is a trigger, not a capability checklist — reaching the thresholds is not a reason to fire; being refused the economics is.

The structure is pre-papered, so exercising is a matter of weeks rather than a year of build-out. GROUNDED · Glacis

Two kill conditions bound the whole insurance layer honestly: GROUNDED · Glacis

1. Carriers adopt the signals but refuse the economics for 12–18 months → the market is saying take the risk yourself — and by then the accumulated loss data has made capacity cheap to rent.
2. The loss delta never materializes → there is no underwriting business to share in, and the insurance layer collapses back to compliance tooling. Painful, honest, and priced in from the start.

The neutrality rule

The cap-table doctrine, absolute: capacity partners and standards governance must never share a cap table. GROUNDED · Glacis

The live case: Aegle Assurance — a capacity-side player whose warranty-style product is unwritable at scale without an independent oracle, and who may also want to invest. Acceptable shapes:

• design partner at commercial per-policy rates — the best version: they pay like a customer and validate the oracle;
• a small, non-lead check in a round not led by insurance money, with zero exclusivity and OVERT governance visibly multi-stakeholder.

Not acceptable: a lead check, an exclusive data license, any seat that lets one rating vernacular claim the receipts lean its way. Two vernaculars — Aegle’s and the Meridian Rating Bureau’s — bidding on neutral receipts beats taking sides in a war the house would lose by winning. GROUNDED · Glacis

Next

→ 16 VERIFY hub: the checklist

Drill this page →24 bank questions stand behind what you just read. Check it while it’s warm.