SpaceX, OpenAI, Anthropic: One Vintage's Entire Power Law, Three Names

Three companies. ~$1.4T in combined private-market valuation. A Forbes analysis published today says their IPO proceeds will collectively dwarf every other VC cash return since 2016. What that means for your fund's DPI clock, carry economics, and next raise is the subject of this brief.

The Setup

On May 20, 2026, SpaceX filed its S-1 with the SEC, designating ticker SPCX on the Nasdaq Global Select Market. That filing is the formal clock-starter for the largest single-vintage liquidity event in the history of venture capital. Six days later, Forbes reported that the IPOs of SpaceX, OpenAI, and Anthropic are "on track to outstrip the cash VCs raked in on every other startup exit since 2016 — combined." The combined private marks across the three names stand at roughly $1.435T: SpaceX at ~$400B (with public-market speculation exceeding $1T), OpenAI at $852B post-money targeting a Q4 2026 IPO, and Anthropic at ~$183B with no announced IPO date. The power law — the statistical fact that a small number of investments produce nearly all venture returns — has never been this visibly concentrated in three names inside a single 18-month window.

The Read

The Power Law Has a Name, a Ticker, and a Filing Date

Every investment in a fund must, at the moment the check is written, have a credible path to returning the entire fund. That discipline is what separates a portfolio of 60 hedge-shaped bets from a portfolio of high-conviction positions with genuine power-law tails. The Forbes analysis confirms what any GP with exposure to these three names already knows in theory but has never been forced to operationalize in practice: the power-law tail for an entire decade of U.S. venture investing is sitting in three cap tables. The combined private marks of ~$1.435T represent a gross equity supply event that has no historical precedent — not in the dot-com window, not in the 2012–2014 consumer-internet run, not in the 2020–2021 SPAC and growth surge.

SpaceX's 24-year private holding period — founded 2002, S-1 filed 2026 — is itself a data point that should rewrite how LP agreements are drafted for the next cycle. The standard 10-year fund life was never designed to carry a position this long. Funds that invested in SpaceX in 2008 have been carrying the position for 18 years. That is not a failure of fund management; it is evidence that the most durable monopolies — those built on proprietary technology that is demonstrably superior on a meaningful dimension, combined with network effects and scale economics that compound with each new customer — take longer to convert than any fund structure assumes. The lesson is not that 10-year funds are wrong. The lesson is that the positions that actually return the fund are precisely the ones that outrun the fund clock.

TVPI and DPI Are Not the Same Event — and Conflating Them Is a Material Error

The most important communication task for any GP with exposure to SPCX, OpenAI, or Anthropic is one that most LP letters will handle badly. When SPCX prices its IPO, TVPI — Total Value to Paid-In, the current mark-to-market of unrealized positions — spikes immediately. DPI — Distributions to Paid-In, the cash actually returned to LPs — does not move at all on that day. The standard SEC-governed 180-day post-IPO lockup means that DPI from the SpaceX IPO cannot begin before roughly November 2026 at the earliest. Employee share unlocks under 10b5-1 plans typically stagger over 12–24 months post-IPO, pushing the full distribution event for most SPCX investors into 2027–2028 — and further still for OpenAI and Anthropic holders, whose IPO timelines trail SPCX by at least two quarters.

LPs who conflate the mark event with the distribution event will mismodel their own liquidity, their re-up capacity for the GP's next fund, and their portfolio allocation. GPs who allow that conflation to persist in their quarterly letters are failing at the most basic obligation of investor communication. The IPO date and the cash date are separated by a minimum of 180 days and, realistically, 12–24 months of staggered unlock.

Sequencing Risk and the Late-Issuer Discount

Public-market absorption capacity is finite. The sequencing of ~$1.435T of new equity into a compressed 12–18-month window creates a structural discount risk for the second and third issuers. OpenAI targets Q4 2026 — pricing after SPCX has absorbed the first and largest tranche of new supply. Anthropic has no announced IPO date, making it the most timing-uncertain of the three. For GPs holding OpenAI and Anthropic, the TVPI marks anchored to $852B and ~$183B respectively should be stress-tested against SPCX's first-60-day public trading range before those numbers appear in LP materials as validated baselines. Secondary-market prices on all three names are trading now and represent the most accurate forward signal for where each IPO will clear.

Carry Is Not Guaranteed by a Big Name

Founders Fund holds stakes in all three companies. Sequoia, Andreessen Horowitz, Greenoaks, and Lightspeed hold stakes in at least two of the three. The IPO sequence will be a transparency event as much as a wealth event — it will reveal which funds were genuinely positioned across their full portfolio versus which ones have a single generational name propping a mediocre distribution of results. A fund that underperformed on its non-SpaceX, non-OpenAI, non-Anthropic positions may not hit the hurdle rate required to unlock carry on the winning names. LPs evaluating GP track records after these IPOs should demand full fund-level carry models, not name-level return tables.

The Action

Build the DPI waterfall now. Map your fund's SpaceX, OpenAI, and Anthropic stakes against the 180-day lockup expiry and the 10b5-1 stagger schedule. The date on which paper TVPI becomes LP cash is 12–24 months after the IPO date, not the IPO date itself. Model three scenarios for each name: base, delay-12, and delay-24.
Demand IPO-event vs. cash-event clarity from your GPs in the next quarterly report. If your GP's Q2 2026 letter conflates a TVPI mark with a distribution, that is a material communication failure. Flag it. A secondary-LP sale market around these three names is already developing for LPs who need to access value ahead of the distribution clock.
Watch secondary-market prices on all three names as the leading IPO-valuation indicator. Secondary trades are the most accurate forward mark available before the S-1 prices. A sustained secondary discount of 20% or more to the last round price is a signal to revise TVPI assumptions downward — not to wait for the prospectus to tell you.
If you are a GP preparing to raise a new fund, time your formal launch to the six-month window after the first of the three IPOs prices well in public markets. Visible cash returns to LPs — even anticipated DPI — compress fundraising timelines materially. The SPCX lockup expiry in approximately November 2026 is the earliest credible signal date.
Stress-test your fund for a 12–24 month IPO delay on Anthropic specifically. Founded 2021 with no announced IPO date, Anthropic carries the least timing certainty of the three. If your fund's end-clock requires exits before 2030, model the fund-extension cost and communicate the probability to your LP advisory committee now, not when the delay is confirmed.

The Counter

The strongest structural objection is also the most obvious: all three IPOs could face material delays. OpenAI's governance restructuring is ongoing; Anthropic has announced no IPO timeline; SpaceX's commercial path depends partly on Starship V3 validation. Stretch the sequence 12–24 months and the Forbes magnitude thesis still holds — the gross equity supply does not shrink — but the fund-duration risk becomes acute, particularly for 2021-vintage investors in Anthropic who could face a 10-year hold before liquidity. The framework response is not to dispute the counter but to operationalize it: GPs should be running base, delay-12, and delay-24 scenarios for each of the three names and reporting the fund-extension probability to LPs in the current quarterly cycle, not after a delay is announced. The second objection — that sequencing $1.435T of new equity into a single 12–18-month window will compress multiples for the later issuers — is quantifiable and real. The second and third large-cap IPOs in a compressed supply window have historically traded at additional discounts relative to the lead issuer. That is not a reason to revise the power-law thesis; it is a reason to benchmark OpenAI and Anthropic TVPI assumptions against SPCX's actual public trading range, not against the last private mark. Paper returns are not cash. The distinction has never mattered more than it will in the next 24 months.

Primary Sources

Forbes — Iain Martin: SpaceX, OpenAI, and Anthropic IPOs to Outstrip All Other VC Exits Since 2016 Combined — Forbes, 2026-05-26
SpaceX (SPCX) IPO 2026 — S-1 Filed — QuantLogix Research, 2026-05-20
OpenAI IPO 2026 — $852B Post-Money — QuantLogix Research, 2026-05-21
Anthropic IPO Research — ~$183B Private Valuation — QuantLogix Research, 2026-05-20

Anonymized senior-practitioner discussion of frameworks for educational purposes — not personalized investment advice. QuantLogix is a research platform. Nothing in this article constitutes a recommendation to buy or sell any security. Past performance does not guarantee future results.