The IPO Window Is Open: Inside the $75 Billion SpaceX Deal Reshaping the S&P 500
A Bloomberg-style panel — moderator Natalia (speaker B), a late-stage growth investor 'Paul' whose fund's single biggest position is SpaceX (speaker A), and Deutsche Bank's Stefan (speaker C, barred from naming live deals) — on the week SpaceX is set to become **'the 5th largest public company in the US on Friday'** via the largest IPO ever (~$75bn, 4× oversubscribed). Paul's bull case: a **~$30tn TAM, three subsidiaries (Starlink, xAI, X) each above $1bn revenue** ('never been seen before in the IPO market'), a 'rule of 60' profile (30% growth + 30% EBITDA margin), 60% of CapEx already going to AI and 60% of revenue from Starlink, flipping to AI-majority by 2028; sell-side underwriters model a **100% revenue CAGR — '$20 billion today... to $1 trillion of revenue in 5 years.'** The most useful half is the sober counter-weight. Stefan reframes the 'will mega-IPOs drain liquidity' fear with two crude ratios — equity raised ÷ S&P market cap was ~1.5% in 2020-21 vs ~1% now ($600-650bn into a $65tn S&P) — concluding absorption is a non-issue *in a healthy market*. Paul names three concrete day-1 risks: hitting an unprecedented growth number, the deep-cap-table divergence (**Facebook compounded ~20% beating the S&P by 500-600bps; Uber ~6%, lagging by 600-700bps**), and simply opening $30-50bn+ of single-name liquidity without the exchange breaking. The throughline for a builder-investor: the IPO pop is mean-reverting (**day-1 ~22-23% in '25 halves to ~11-12% by day 30**), the real action is in private 'atoms + software' compounders (Anduril, Ramp, Deel, Databricks, Hadrian, Cerebras), and the private market now dwarfs the public one — **OpenAI's $120bn private raise was 'the biggest equity capital markets transaction in history,' '9 times' Ant's 2018 $14bn**. Cross-reads to Chamath this week: a regime of abundant liquidity flatters everything; discipline is the screen (30% min growth, path to free cash flow), not the mood.
Key points
- SpaceX prices Friday as the largest IPO ever (~$75bn, 4× oversubscribed), and Nasdaq fast-tracked it into the Nasdaq 100 — Paul argues index providers 'have to change the rules to reflect the market' when a company is the 5th largest in the country on day one, something the S&P/Dow rule-writers never anticipated.
- The bull structure: ~$30tn TAM across three subsidiaries (Starlink, xAI, X) each >$1bn revenue; a 'rule of 60' company (30% revenue growth + 30% EBITDA margin); 60% of CapEx on AI, 60% of revenue from Starlink today, flipping to AI-majority revenue by 2028.
- Sell-side underwriters forecast a ~100% revenue CAGR over five years — '$20 billion today... to $1 trillion of revenue in 5 years' — which Paul frames as the reason for 4× oversubscription, comparing demand to Alibaba's ~20×-covered $25bn IPO a decade ago.
- Liquidity-absorption fear, defused with ratios (Stefan): ~$500bn raised / $35tn S&P (~1.5%) in the 2020-21 boom vs ~$600-650bn run-rate / $65tn S&P (~1%) now; the Mag Seven sit on ~$20tn of stable capital across top-50 active, top-50 passive and ~30% retail — deep enough to absorb new supply in a healthy market.
- Three day-1 risks (Paul): (1) hitting an unprecedented ~100% CAGR while transitioning from satellite- to AI-majority revenue (de-risked by billion-dollar-a-month Google/Alphabet contracts); (2) deep-cap-table divergence — Facebook ~20% total-return CAGR (beat S&P by 500-600bps) vs Uber ~6% (lagged by 600-700bps); (3) opening $30-50bn+ of single-name liquidity in a compressed window without the exchange breaking (the BATS-never-opened anecdote).
- IPO pops mean-revert: in '25 the average US day-1 pop was ~22-23% and roughly halved to ~11-12% by day 30 — driven by the marginal buyer fading once FOMO and scaled-back allocations clear; Cerebras cited as the pattern (priced $185, traded $225-230, Citi $300 PT).
- The private market now dwarfs the public one: OpenAI's ~$120bn private raise was 'the biggest equity capital markets transaction in history' — '9 times' the $14bn Paul's team raised for Ant in 2018; the panel still calls the public markets 'the most efficient place to raise capital' at scale, which should pull more private companies out.
- Paul's selection screen (private US tech only): minimum 30% multi-year revenue growth (portfolio median ~90%) married to a path to profitability — a third already cash-flow positive (Ramp, Deel); Anduril underwritten at a $28bn Series G with 5-year modelled burn but stress-tested to still clear 30% growth if forced to pivot to cash flow.
- Sector concentration: the market is paying a premium for 'four corners' — quantum, semiconductors, electrification/powering AI, and critical minerals + defense — with relative growth ordering SpaceX ~30% LTM, OpenAI ~200% YoY, Anthropic ~1,400% YoY; Polymarket (a portfolio company) implies ~$2tn Anthropic and ~$1.5tn OpenAI, i.e. headroom over last private rounds.
- Macro backdrop is tolerant for now: ~$90 oil is absorbable (the consumer bites closer to $150), one rate hike priced, inflation watched; the 2019 Lyft/Uber same-sector double-IPO is the cautionary tale, and the closing advice to issuers is blunt — only go public 'if you're ready' to beat numbers in the first two quarters.
Notable quotes
The reality is SpaceX is gonna be the 5th largest public company in the US on Friday.
We look at a business with close to a $30 trillion TAM, 3 subsidiaries, which each have over $1 billion of revenue, which has never been seen before in the IPO market.
they're forecasting a 100% revenue CAGR over the next 5 years, which if that plays out, you go from $20 billion today the biggest revenue scale ever for an IPO to $1 trillion of revenue in 5 years.
the biggest equity capital markets transaction in history was in the private market.
We raised $14 billion for Ant in 2018, and that was a huge achievement at the time. OpenAI did 9 times that.
in '25, for example, day 1 pop on average for the US IPO market was about 22%, 23%. You looked at the same statistics by day 30, it was about anywhere between 11% and 12%.
You look at the average run rate revenue for OpenAI, it's implying 200% year over year growth.
At the end of the day, the public equity markets are the most efficient place to raise capital. It's much easier, much bigger quantums of capital.
Themes
- the IPO window reopening
- AI infrastructure as the dominant asset
- liquidity-absorption capacity
- private-to-public valuation handoff
- growth-plus-path-to-profitability discipline
Mentioned
Companies
Ideas
- rule of 60
- liquidity absorption (equity raised / market cap)
- deep-cap-table divergence (Facebook vs Uber)
- IPO day-1 pop normalisation
- the marginal buyer
- private mega-raises eclipsing public ECM
- 30% growth + path to profitability screen
- index fast-tracking
- four-corners sector premium (quantum / semis / electrification / critical minerals + defense)