Priced for a Monopoly.

Executive summary

Reported marks for the two largest frontier-AI issuers imply a combined enterprise value of well over a trillion and a half dollars against annualised revenue of roughly fifty billion and operating losses measured in the tens of billions. Allocators are being asked, in effect, to underwrite a monopoly-shaped outcome without monopoly-grade disclosure. The first audited S-1s will not resolve that tension; they will sharpen it.

A cluster without precedent

Two confidential filings have landed within weeks of each other, with a third cohort of infrastructure issuers queued behind them. The compression matters: allocator budgets are finite, underwriter syndicates overlap, and index-inclusion mechanics will compete. The historical pattern in tight windows is that the second deal prices at a discount to the first, or one is delayed. Neither outcome is priced into the current marks.

Why the filing changes the analysis

Until an S-1 is filed and audited, every number quoted for these issuers is a secondary-source estimate — useful, defensible, and not the same thing as a disclosure. The filing converts roughly a dozen of those estimates into a primary record, and forces the issuer to characterise its own related-party economics, its purchase commitments, its concentration exposures, and its governance reserves.

That conversion is asymmetric. Numbers that were comfortable as estimates become uncomfortable as disclosures; numbers that were uncomfortable as estimates sometimes turn out to be conservative once the auditors are through. Either way, the framework for analysis tightens.

Valuation against disclosure

The standing Cabier framework — fundamentals, discounted optionality, named narrative premium — survives the move from estimate to disclosure. What changes is the size of the residual. The honest answer for both issuers is that, on reported marks, the narrative premium is material; the discipline is to name it, not to absorb it into a multiple. The frame "priced for a monopoly" is not an accusation. It is a description of the residual that has to be defended once the fundamentals and the discounted options are accounted for.

What the first audited numbers will anchor

Four numbers will do most of the work. The audited revenue run-rate; the audited operating loss and its trajectory; the magnitude of off-balance-sheet purchase and capacity commitments; and the disclosed scope of the governance reserves held by non-profit parents or long-term benefit trusts. Each of those moves the sum-of-the-parts walk. None of them, on its own, breaks the case for the issuer — but together they constrain how aggressively the residual can be defended.

The institutional takeaway

Allocators should not wait for the S-1 to do the work the framework already permits. Build the sum-of-the-parts walk on reported marks now; mark the governance discount independently; identify the disclosure thresholds at which the residual is no longer defensible. When the audited numbers arrive, the test is whether the walk holds — not whether the headline figure is the same.

A forthcoming Cabier tool will support the disclosure-threshold test directly. Until then, the discipline is the standing one: name each layer, price it separately, and refuse to let the narrative premium be absorbed quietly into a multiple.