SpaceX 135 IPO Cap|7T Sideline Cash or Overvalued Bet?

The Pricing Gap No One Can Resolve

Three Wall Street banks publicly underwriting SpaceX's IPO are simultaneously telling their clients not to chase it at $135 a share. That is the fact the market opened Monday trying to price.

SpaceX priced its initial public offering at $135 per share last week, implying a $1.77 trillion valuation — seventh-largest company in the US by market cap on day one, ahead of Tesla. Morningstar's analysts called it "significantly overvalued" before the first trade, estimating fair value at roughly half the IPO price. Jay Woods, chief market strategist at Freedom Capital Markets, told CNBC the same morning: "I would not be chasing it at these levels." The Nasdaq had just posted its worst single session in more than a year on Friday — a 4% drop — in what several desks attributed at least partly to investors rotating out of high-flying chip stocks to free up cash for the SpaceX allocation.

That is the setup: a record-breaking offering debuting into a market that sold off to fund it, underwritten by banks whose own traders are flagging caution, and entering a week when OpenAI announced a confidential S-1 filing that same Monday morning. OpenAI, valued at $852 billion and having just raised $122 billion in March, told investors it "may be a while" before it actually goes public — but the filing signals that it needs the option. Anthropic had filed its own S-1 a week earlier at a $965 billion valuation. Three companies collectively asking the public market for somewhere between $2 and $4 trillion in combined implied market cap, within a single calendar month.

The question Monday was not whether SpaceX's technology is real. The question was whether the capital pool can bear the weight of three simultaneous trillion-dollar asks — and what that demand compression does to the pricing of the ones arriving second and third.

The Lockup Clock and the Rate Wall

Tom Lee's counterargument is specific and worth examining before dismissing it. He told CNBC there is $7 trillion in cash sitting on the sidelines — high net worth capital that has not yet rotated into equities — and that SpaceX's IPO will not only be absorbed but will pull that cash into the market, lifting the broader index in the weeks that follow. His premise: these AI companies, unlike the dotcom class of 1999, are "not fully funded." They need to maintain a constructive relationship with the public capital markets to access future financing for 5-year infrastructure build-outs, which creates structural incentive to support the stock post-debut.

That premise is load-bearing. If true, SpaceX's management team, and by extension OpenAI's and Anthropic's, will manage investor relations differently than a company that raised once and is done. The lockup expiration — when insiders can begin selling — becomes less of a 1999-style flood and more of a managed cadence. Lee's "sideline cash" frame requires believing that $7 trillion in undeployed capital is accessible, patient, and genuinely waiting for this entry point rather than waiting for a price correction.

Morningstar's frame requires believing the opposite: that the current valuation already embeds a scenario where SpaceX captures the full upside of Starlink global dominance, xAI frontier model leadership, and the rocket reusability cost curve simultaneously — and that any slippage in any of those three legs creates a repricing event from a very tall start.

The 10-year Treasury yield staying above 4.5% is the condition both frames are quietly ignoring. At 4.5%, the discount rate embedded in a $1.77 trillion terminal value calculation is not trivially accommodating. The JPMorgan trading desk noted Monday that CPI and PPI data due Wednesday and Thursday could push yields higher still if inflation figures surprise. That is a named timing anchor for the week.

The unstated premise underneath Lee's thesis is that the $7 trillion in sideline cash is yield-insensitive — that those holders will buy a cash-burning AI company at 170x trailing revenue despite a risk-free rate above 4.5%. Morningstar's team does not share that premise. Neither do the traders who sold Nvidia and Micron on Friday to fund the SpaceX allocation.

The Ordering Problem and What Comes Next

Here is what the SpaceX pricing question leaves unresolved going into Friday's debut: the ordering of the three IPOs determines valuation comp precedent.

SpaceX trades first. Its reception — specifically whether institutional investors hold or distribute in the first 48 hours — sets the reference frame for Anthropic's pricing when it files a public S-1, which then constrains what OpenAI can claim in its own filing. TechCrunch noted this explicitly: "if Anthropic prices conservatively, OpenAI's path to its target valuation gets harder." OpenAI CFO Sarah Friar had already raised concerns internally about whether the company could sustain data center spending; the S-1 process will force public disclosure of exactly how large the burn is. The company's own projections show $85 billion in cash consumption in 2028 alone, even after doubling revenue.

Lockup mechanics add a second timing pressure. Reuters Breakingviews described the situation this way: SpaceX's lockup expiration does not create a single tsunami of insider selling — it creates "constant pounding" of structured tranches over time. The Nasdaq sold off 4% on Friday in part because of allocation-driven chip stock liquidations. That pattern — momentum rotation into the IPO name, temporary pressure on adjacent highflyers — will repeat at each lockup tranche.

Micron Technology recovered 9% Monday after that Friday selloff, which tells you something about how quickly allocation-driven dislocations unwind when the proximate cause passes. But the structural question does not unwind with the price recovery: the capital pool compression persists as long as all three companies are simultaneously active in primary and secondary markets.

The continuation case requires SpaceX's first week of trading to demonstrate institutional holding rather than distribution, CPI data on Wednesday to come in below consensus and ease the rate pressure on terminal value math, and OpenAI to take its time — not rushing a public S-1 that would flood the market with valuation comps before SpaceX has established a stable floor. If those three conditions hold, Lee's sideline cash thesis has a mechanism to work.

The breakdown case requires only one: a hotter-than-expected inflation print Wednesday that pushes the 10-year above 4.7%, forcing discount rate revisions across all three valuations simultaneously — at the moment when all three companies are most visible to public investors for the first time.

The verification benchmark is Wednesday's CPI print. Not as a standalone macro event, but as the specific input that either validates or invalidates the rate assumption embedded in every $1.77 trillion calculation currently on the tape.