SpaceX's $18.4 Billion AI Gamble Is Testing Investor Patience

SpaceX's immediate test is share supply, not just the AI story

The near-term risk is not the AI narrative on its own. It is a classic sell-pressure mismatch: a large pool of eligible shares is running into a market built on scarcity. SpaceX's first public earnings did show $7.81 billion in quarterly revenue, but the more immediate test arrives just days later, when 911.5 million shares held by employees and other pre-IPO stakeholders become eligible to sell.

Why the float matters more now

SpaceX started with an unusually tight public float. At IPO, less than 5% of its total outstanding share count was available to the public. That setup can exaggerate both upside and downside moves. After an early surge, the stock has already fallen 33% from its post-IPO record close and has traded below the $135 IPO price after six weeks as a public company.

AI spending has amplified the pressure, but the cleaner near-term issue is simpler: liquidity is changing. Lockup expiry does not create new shares, yet it can still shift a stock from a scarcity trade toward a liquidity test if insiders choose to sell.

The market now wants proof that AI spending can pay for itself

The first earnings report changed the debate. Investors are no longer asking whether SpaceX can build an AI narrative; they are asking whether the AI buildout can fund itself quickly enough to justify the valuation. The company did post a strong top-line result, with $7.81 billion in quarterly revenue and a 92% year-over-year jump. But the market focused even more on spending: capital expenditures reached $18.4 billion, well above analyst expectations, and well over 80% of SpaceX's capex went towards artificial intelligence.

The timing problem is the real pressure point

SpaceX is front-loading AI investment before the revenue base is fully proven. On the earnings call, CFO Bret Johnsen said the company believes it is getting less than a one-year payback on AI compute capital deployment. That framing helps the bull case, but it still leaves investors underwriting the gap between today's heavy outlay and tomorrow's cash-flow proof.

There is reason for optimism. Reuters reported that compute contracts could exceed $25 billion annualized once fully ramped. But contracts and future ramp figures are not the same as current proof that AI can start funding the buildout without leaving Starlink as the main cash engine.

Bulls have a future revenue case; bears have a late-mover problem

The bull case rests on the idea that today's spending is infrastructure, not waste. If contracted compute demand converts into durable utilization, the spending profile can start to look rational rather than reckless.

The bear case is more immediate. CNBC reported SpaceX is way behind OpenAI, Anthropic and Google in models and services as it tries to compete in compute against Microsoft, Amazon and Google. In a market where customer habits and ecosystem strength matter, being late can keep investors skeptical longer.

What matters from here is float expansion, insider behavior, and the $135 line

The sell-pressure mismatch is no longer theoretical. It is now a trading event.

August 6 turns the lockup into a liquidity test

Starting August 6, as many as 911.5 million shares held by employees and early investors become eligible to sell, just days after SpaceX's first public earnings release. That could significantly increases the potential supply of stock relative to the company's initially tiny public float. Because SpaceX structured a series of release valves instead of a conventional 180-day lockup, the market should expect rolling liquidity changes, not just a single cliff.

Selling pace matters more than the AI pitch

For bulls, the key signal is not a better story. It is whether pre-IPO holders retain enough skin in the game to suggest a long-term buildout rather than an orderly exit. According to CEO Elon Musk, who controls approximately 60% of outstanding shares, is subject to a separate longer lock-up and is not expected to begin selling until June 2027, so early pressure would come mainly from employees, early investors, and other insiders.

If that group sells lightly while demand improves, the market may be willing to look past near-term noise. If selling outpaces demand, investors are likely to treat public shares less as a long-duration asset and more as ready-made exit liquidity.

$135 is still the cleanest technical checkpoint

The clearest line to watch is the IPO price. Combined with volume and tape strength, whether SpaceX can hold or reclaim $135 after the unlock will say a lot about whether the market sees the company as a scaling AI and compute builder or as a business being weighed down by both spending and supply.