Analysis
SpaceX's median $225 analyst price target, set against 2025 revenue of under $19 billion and a $4.9 billion net loss, implies pricing the company at roughly 105 times trailing revenue -- a multiple with no direct comparable anywhere in launch services, satellite communications or even hyperscale cloud infrastructure. That number matters well beyond SpaceX itself: it's the highest-profile, most-watched AI-adjacent public print available right now, and every company still queued behind it in the IPO pipeline will be measured against whatever trading range SpaceX ultimately establishes.
What makes the multiple particularly striking is that it holds even after SpaceX has already fallen roughly 42% from its post-IPO peak near $211 back to roughly $123-$125. The stock's round trip hasn't meaningfully compressed the embedded growth assumptions in Wall Street's bull case -- it's simply repriced off a starting point that was, by the banks' own admission through Fortune's reporting, built on herd-like analyst clustering rather than fully independent modeling.
The comparison set is instructive. Databricks just marked itself at $188 billion privately with real, if undisclosed, enterprise revenue -- a fundamentally different risk profile than SpaceX's revenue-versus-loss picture, but still untested against any public listing. OpenAI's confidential S-1 and Anthropic's own reported IPO preparations, surfaced through Menlo Ventures' recent 40%-plus IRR disclosure tied to its Anthropic stake, are both still ahead of SpaceX in one sense: neither has yet had to defend a specific multiple in front of daily public trading.
For bankers structuring the next wave of AI-adjacent listings, SpaceX's actual trading range -- not its IPO price, not its post-listing peak, but wherever it settles once the current volatility clears -- becomes the working comp whether they want it to or not. If that settling point compresses further from today's roughly $123-$125, every AI-infrastructure company still private will face a lower ceiling than the one SpaceX itself was priced against in June.
The bear case: SpaceX's business mix -- launch services, Starlink, Starship development -- is different enough from a foundation-model lab or a data-platform company that treating it as a universal AI-infrastructure comp overstates how directly its multiple transfers to OpenAI, Anthropic or Databricks. What to watch next: where SpaceX's stock actually stabilizes over the next quarter, and whether the next AI-adjacent company to file explicitly addresses the comp-set question in its own roadshow materials.
