Five years to $100B: the private market now finishes compounding before the IPO bell

The number Forge does not publish is the one that would settle its own argument. The August update says a "meaningful share" of late-stage appreciation now happens while companies are still private, and names Anthropic, OpenAI, Databricks and Stripe as the evidence ^[8]. Meaningful is not a percentage, and the four names are not a sample.

What is checkable is the clock. Sixteen years on average for the pre-2011 cohort including SpaceX, Stripe and Waymo ^[2]; roughly five for Anthropic, OpenAI and xAI ^[3]. That gap of eleven years ^[4] is longer than the investment period of most venture funds. Appreciation that used to be split between a long private hold and the first years of public trading now completes inside one fund's life, funded by sovereign wealth, private equity, corporate and crossover money competing for the same handful of names ^[6].

Forge's July index split is where the argument gets awkward for Forge. The private-constituent FPMI rose 9.3% while the FAPMI, which carries public post-IPO exposure, fell 4.2% ^[10][c10b], a spread of 13.5 points in one month ^[15]. The drag was SpaceX, down 36.6% after listing ^[11]. Forge's framing is that the IPO is becoming a liquidity milestone rather than a value creation event ^[9]. In its own benchmark, one newly public name was able to overwhelm every positive private contribution in a cap-weighted index. That is the listing setting the price, not confirming it.

The private side of that divergence is thinner than the headline number suggests. SambaNova rose 142.9% and added 6.5 points to the equal-weighted FPMI after a July 8 announcement of a first close ^[12], which is about 70% of the month's 9.3% gain from a single position ^[13]. A financing headline moved the index; breadth did not.

That matters because the price discovery Forge credits for private markets behaving more like public ones comes from tender offers and secondary transactions ^[7], the same channels that produce the marks. More frequent marking is not the same as a clearing price with a float behind it. If the five-year cohort lists at valuations already set by that pool, public buyers are underwriting an exit rather than an entry. Forge itself leaves open whether the next wave of $100 billion-plus listings breaks the pattern or repeats it ^[14].