The Round Trip: What SpaceX’s Falling Stock Is Actually Asking
Since its June 12 IPO, SpaceX has completed a full round trip, pricing at $135, surging above $225 within days, and now trading below its offer price. The debate is no longer about what SpaceX was worth as a private company. The market is trying to determine what kind of public company it is becoming. One increasingly useful way to think about SpaceX is as an emerging hyperscaler. It may not follow the path of Amazon, Microsoft, or Google, but its combination of launch infrastructure, Starlink, and AI compute is creating a business model that looks increasingly familiar, while pointing toward a very different destination.
Since listing on June 12, SpaceX has done a full round trip.
It priced at $135, pushed above $225 intraday within days, and now trades below its offer price. The market is no longer arguing over a private valuation in the abstract. It is pricing a live public company, and the question under that repricing is becoming clearer.
What exactly is SpaceX turning into?
One useful way to read it is this: SpaceX is starting to take on the shape of a hyperscaler. Not in the way Google, Microsoft, or Amazon did, and not necessarily toward the same end state, but the structure is recognizable enough that the comparison matters.
Why the Comparison Exists at All
The broad outline is familiar.
SpaceX now has a first-party AI stack through Grok and Cursor, while also renting compute capacity to third parties. That combination, running internal AI while monetizing infrastructure externally, looks a lot like the dual model the major cloud platforms operate.
The numbers behind that narrative are large enough to matter. Reported contracts with tenants like Anthropic, Google, and Reflection imply a potential run-rate that is meaningful relative to SpaceX’s broader revenue base. On paper, compute and AI infrastructure began to look like a third pillar next to launch and Starlink almost overnight.
That is what made the market willing to entertain the hyperscaler analogy in the first place.
But the more important question is whether the resemblance is structural enough to last, or whether it is only temporary because the market is still trying to understand what kind of company it is looking at.
Why This Is Not a Normal Hyperscaler Story
The first difference is the direction of travel.
The major cloud incumbents built broad infrastructure businesses first and then layered AI demand on top. SpaceX appears to be moving in the opposite direction. It built a large amount of compute capacity for its own AI ambitions and is now renting some of that capacity externally while its own usage catches up. Over time, those paths could converge. In the present, they are still meaningfully different. That matters because renting spare capacity is not the same thing as having built a cloud platform around external demand from the beginning.
The second difference is the type of revenue.
Traditional cloud is diversified, elastic, and made up of large numbers of customers with usage that expands and contracts across a broad platform. SpaceX’s current compute business looks closer to a handful of very large reservations. That is a different kind of business. It can still be valuable, but it behaves more like wholesale leasing than like a mature hyperscale platform.
The contract structure matters too. If large portions of the revenue can leave on relatively short notice, the market will not treat it the same way it would treat a deeply embedded, recurring cloud business. That makes upcoming earnings especially important, because the next print is where the market starts to test whether the run-rate story is durable or simply directional.
Why the Real Bet Sits Somewhere Else
The deepest reason this is not a normal hyperscaler story is that the terrestrial compute business may not be the final destination at all.
The more ambitious version of the thesis is orbital.
The xAI merger was framed in part around the idea of building data-center capacity in space, and the filings point toward orbital AI compute infrastructure later in the decade. On that reading, ground-based compute rental is not the endgame. It is the bridge. It generates revenue, helps absorb current capacity, proves outside demand exists, and buys time while the more differentiated piece of the strategy is being built.
That is where SpaceX starts to look unlike any incumbent cloud company. The established hyperscalers own land, data centers, networks, and chips. They do not own launch systems. SpaceX does. If orbital compute ever becomes economically and technically viable, the company would be building from an asset base that the existing cloud leaders simply do not have.
That is the core of the bullish case. It is also the part of the story that remains least proven.
Why the Market Is Pulling Back
The round trip in the stock is the market beginning to separate these layers.
At the highs, the stock was trading more like the orbital version of the story was already halfway realized. At lower levels, the market appears to be asking a more disciplined question: how much of the current valuation is justified by the business that exists now, and how much depends on a future that is still highly contingent?
That is where the recent selloff becomes more understandable.
The company’s AI arm is still deeply loss-making. The underlying buildout is capital intensive. Some of the current revenue is concentrated. And the most differentiated part of the thesis — compute in orbit, tied to Starship’s cost curve and space-based infrastructure — remains conceptual enough that it still belongs to the future more than the present.
That does not make the thesis wrong. It does mean the stock is now being asked to support a more precise conversation than the early post-IPO momentum phase required.
What the Next Earnings Print Has to Answer
The first major checkpoint is approaching quickly.
The next earnings report is the first real public test of how much of the compute narrative is translating into something durable enough for the market to underwrite. It is also where the company has to begin showing whether the new pillar is becoming operationally legible, not just strategically compelling.
That does not mean one quarter will resolve the debate, but it will start to answer the question the stock is now asking: is SpaceX becoming an AI infrastructure company with a moat that could eventually extend beyond the terrestrial cloud model, or is it still better understood as a strong but more conventional infrastructure business carrying a much larger future narrative on top of it?
That is the real meaning of the round trip. The market is no longer just pricing hype, scarcity, or post-IPO momentum. It is beginning to test which version of the story deserves the valuation. And that is a much more interesting phase than the pop ever was.
T-Tokens are tokenised loan participation rights, not equity. High risk. DYOR. Not financial advice. Not available in the US or other restricted territories. tessera.pe/terms