SpaceX Has Lost an Entire Tesla

On June 12, 2026, SpaceX listed on the Nasdaq under the ticker SPCX – $135 a share, roughly $75 billion raised, the largest IPO in history. Four trading days later the stock hit $225.64, up 67 percent.

On Monday, SPCX closed at $113.50, its 13th losing day out of 16 sessions. Tuesday’s pre-market took it down to around $110. That is minus 50 percent in six weeks and more than $1.2 trillion in market value erased – almost exactly Tesla’s entire market capitalization.

1. The price was the bet

At the offer price, SpaceX was worth about $1.75 trillion on 2025 revenue of $18.67 billion. Even after the crash, that is a price-to-sales ratio north of 80; conventional aerospace groups trade at 1.5 to 3. What carries that number is not the 33 percent growth rate but expectations – and expectations are the first thing a market reprices.

2. A four percent free float

Only around four percent of the shares were freely tradable at the IPO. With a float that thin, modest capital flows move the price hard in both directions. The rally to $225 was momentum, not an institutional re-rating. Even inclusion in the Nasdaq 100 on July 7 ended down seven percent the same day.

3. The rocket builder became an AI company

With the xAI acquisition in February, SpaceX repositioned itself for investors as a vertically integrated AI company. That explains both the valuation and the losses: in Q1 2026, Starlink generated $1.19 billion in operating profit on $3.26 billion in revenue, while the AI segment posted a $2.47 billion operating loss. At the same time, the market is punishing AI capex – the post-Fed tech selloff and a $25 billion bond issue landed at the worst possible moment.

4. The orbital data centers don’t survive scrutiny

The most speculative part of the story is unravelling in public. Analyst Peter Zeihan calculated that a single one-gigawatt orbital data center would require roughly 500 Starship launches, calling the concept dumb on multiple levels. The prospectus itself concedes the technology is unproven and may never be commercially viable. The real AI business is terrestrial: Anthropic pays $1.25 billion a month for Colossus capacity, Google $920 million a month for GPU access starting in October.

5. Skepticism meets euphoric analysts

Value investor Jeremy Grantham told Morningstar this was the craziest IPO in human history and puts the odds of a collapse at 90 percent – the valuation rests on AI progress, he argues, while SpaceX’s own AI offering is third-rate. Against him stands a nearly unbroken wall of buy ratings from Goldman to JPMorgan, with an average target of about $237. That gap is itself a warning sign. Short sellers are sitting on billions in gains.

6. August 6 casts a long shadow

SpaceX reports its first quarterly results as a listed company on August 4. Two trading days later, the first lockup expires: up to 911.5 million shares become tradable – a multiple of the current free float. Anyone selling today need not believe the company will fail, only that supply pressure is coming in ten days.

The other side

Cathie Wood’s ARK keeps buying and calls SpaceX a company with the potential to become the most important in history – backed by nearly 10,800 Starlink satellites and the thesis that AI compute on Earth is hitting physical limits. Macquarie has a $250 target. And operationally it is working: Starship Flight 13 met all mission objectives on Friday.