The Top 10

IThe ranking

Ranked by the most recently disclosed private valuation, the cohort is steeply concentrated at the top. SpaceX sits clear of the field at roughly $1.25tn following its February merger with xAI; on 28 May, Anthropic’s $65bn round lifted it to $965bn, marking the first time it has been marked above OpenAI. Below the trillion-dollar tier, the distribution falls away sharply: the tenth name, Kalshi, is marked at $22bn — roughly one-fiftieth of the leader.

IIWhere the weight sits

Grouped by sector, the aggregate is lopsided. Artificial intelligence and data account for close to half of the total; aerospace & defence, almost entirely SpaceX, takes another third. A single consumer-internet name, ByteDance, carries more weight than the three fintech names combined. The much-discussed breadth of the private-market boom is, at the top, a story about two industries.

IIIWhat the marks rest on

Where revenue is publicly reported on a roughly comparable basis, the implied multiple tells the clearer story. ByteDance — profitable, scaled, growing — is marked near three-and-a-half times reported revenue, a figure a public investor would recognise. At the other end, OpenAI is marked in the mid-thirties on an annualised revenue base it has only recently reached. The cohort is, in effect, two businesses wearing one label: scaled incumbents priced on cash flow, and frontier bets priced on the size of the prize.

IVWho is setting the price

The most consequential shift is not in the levels but in where they are struck. For most of this cohort, the last meaningful mark came not from a primary round but from a tender offer or secondary sale — the mechanism by which employees and early holders find liquidity without a listing. Secondary transaction volume reached a reported record of roughly $226bn in 2025. The private market has, quietly, built its own price-discovery apparatus.

That apparatus is now extending into derivatives. On 29 May the CFTC approved the first perpetual-futures contract listed on a US-regulated exchange. The contract was bitcoin; the structure is the point. Offshore, synthetic pre-IPO perpetuals tracking SpaceX and other private names are already trading on order books, with one venue reportedly pricing a recent listing within a few percent of its public open. Price discovery for companies in this very table is beginning to happen continuously, in public view, before an S-1 is ever effective.

VWhat the desk is watching

Three observations follow, offered as research rather than recommendation. First, the gap between the last private mark and reported listing targets is, for some names, wide — Revolut was marked at $75bn in November and reporting points to listing ambitions far above it; SpaceX’s reported IPO range sits well above its last standalone tender. Gaps of that size are claims about the future, and the public market will test them.

Second, the durability of these marks rests on the methodology behind them. Quarterly marks are set by GP valuation committees per each firm’s policy; independent specialists are typically engaged for the hardest-to-mark positions; third-party auditors test those marks at year-end. A secondary print is a transaction, not an audited mark — a useful signal, not a settled value. Where public comparables compress, the question is how quickly private marks follow.

Third, concentration cuts both ways. A cohort whose aggregate is four-fifths two sectors is a cohort whose aggregate moves with two narratives. We observe the dependency; we draw no conclusion the reader has not earned for themselves.