SpaceX’s stock closed at $135.27 on Wednesday, barely clinging to the $135 price Elon Musk set for its June 12 IPO, according to TechCrunch AI. At one point during the day it dipped below $133. That’s a full month of steady decline from the $200-plus peak it hit days after going public, when its valuation briefly rivaled Amazon and Microsoft.
Why does a rocket company’s stock chart belong in an AI newsletter? Because Anthropic and OpenAI have both filed confidentially for IPOs, and SpaceX is the test case everyone’s reading.
What’s actually happening
- Thin float. Only 4% of SpaceX’s total shares trade on the Nasdaq. Tiny supply plus enormous attention equals wild swings in both directions.
- Bonds are sagging too. The debt SpaceX sold after the IPO is also suffering, which matters more than the equity dip. Stock can swing on sentiment. Bond investors are pricing actual repayment risk.
- The vision discount is arriving. Markets are “sobering up on CEO Elon Musk’s grand vision,” as TechCrunch AI puts it, part of a broader tech deflation over the last month.
What stands out is that third point. SpaceX didn’t miss a number. It didn’t lose a contract. Investors simply started discounting the story, and the story was most of the valuation.
The read-through for AI IPOs
SpaceX and the frontier labs are selling the same instrument: a narrative about a future that doesn’t exist yet, priced as if it already does.
Neither Anthropic nor OpenAI has set a date. They’re watching. And what they’re watching is a company with real revenue, real government contracts, and a real monopoly on orbital launch, struggling to hold its own IPO price for thirty days.
That’s the uncomfortable part. SpaceX has a business. The labs have a business plus a much bigger promise and much bigger losses. If the market won’t extend patience to Musk, the pitch that AGI justifies the multiple gets harder to land.
Expect a few things over the next 12 to 24 months:
- Later IPOs, or smaller ones. If public markets stay skittish, both labs can stay private longer. Private capital is still flowing. The pressure to list isn’t financial, it’s about employee liquidity.
- A pivot in the pitch. Less “we’re building AGI,” more “here’s our enterprise ARR and gross margin.” The companies that IPO successfully will be the ones that sound boring.
- Float games. SpaceX’s 4% float created the volatility. Any lab going public will study that number carefully, and probably float more.
The Starship test
SpaceX launches Starship Thursday, its first flight since a booster failure in May and its first since the IPO. Under the company’s “fly, fail, fix” approach, both the booster and upper stage will end in an explosion regardless of whether anything goes wrong. They’re simulating landings in the Gulf of Mexico, not recovering hardware.
So the market gets to watch a planned explosion and decide what it means. That’s the whole problem with narrative-priced assets in one image.
What to do with this
If you’re a founder raising on AI narrative: the vision premium is compressing. Start building the boring case now. Revenue quality, retention, unit economics. You’ll need it sooner than you planned.
If you’re a buyer of AI tools: vendor durability just became a real diligence question. A lab that needs public markets to fund its compute bill is a different counterparty than one that doesn’t. Ask about runway.
If you’re watching for the IPO window: don’t read SpaceX’s daily chart. Watch the bonds and watch the float. Those tell you what institutions actually believe.
The frontier labs spent three years being told they were priced fairly because the future is enormous. SpaceX just showed what happens when the market decides to price the present instead. Both labs are taking notes.
More details at the original source.