SpaceX's first public earnings report lands: $7.8bn revenue, $100bn cash, and an Nvidia satellite deal

At 14:38 UTC on 4 August 2026, Polymarket reported that SpaceX would release its first-ever public earnings report after the US market close the same day. By 19:32 UTC, Polymarket reported the stock had moved roughly 10% higher into the print. At 20:31 UTC, an Unusual Whales post said SpaceX had disclosed $100bn in cash. At 20:49 UTC, Polymarket reported SpaceX had selected Nvidia for its orbital AI data-centre satellites. At 20:50 UTC, Reuters filed a piece headed "SpaceX's satellite ambitions squeeze out rivals reliant on its rockets."

The day's thread evidence is a sequence of relayed headlines, not a single integrated announcement. The headline numbers are large. The sequencing matters. The earnings print, the cash disclosure, the Nvidia selection, and the Reuters squeeze-framing on launch-dependent rivals all landed inside a six-hour window, and reading them together is tempting. The more rigorous read is to say what each item says, note what it does not specify, and flag the inferences that require evidence the thread does not contain.

What the release says, and what it does not

The figure in the Polymarket relay at 20:10 UTC is $7.8bn in revenue for the first quarter as a public company, with growth of 92% year-on-year. The comparator base is not specified in the relayed item. The Unusual Whales post at 20:31 UTC puts the cash position at $100bn. The Polymarket contract opened at 19:57 UTC implies that a sell-side consensus existed; the consensus figure itself is not in the available thread.

The Polymarket item at 14:38 UTC frames the release as a written earnings report, not an earnings call. The thread does not establish a 20:00 UTC release time; the Polymarket item specifies only "after the market closes today." Segment-level revenue (launch versus Starlink versus any nascent line), net income, and forward guidance are not specified in the available thread. For any investor-grade read, the absence is the story.

What the Nvidia headline says, and what it does not

The Polymarket item at 20:49 UTC reports that SpaceX selected Nvidia for its orbital AI data-centre satellites. The headline does not specify Nvidia's role beyond "selection," nor does it detail volumes, satellite counts, launch timelines, or the deployment architecture. The Reuters wire at 20:50 UTC, on the same hour, describes SpaceX's satellite ambitions squeezing rivals reliant on its rockets; the Reuters headline and the Polymarket item are separate items, and the Reuters headline does not name Nvidia or describe the orbital AI data-centre arrangement.

The temptation is to read the two items as one move: SpaceX acquiring the in-space compute layer, with Nvidia as silicon supplier, and squeezing rivals who depend on SpaceX for launch. The thread evidence supports a narrower statement. SpaceX has selected Nvidia for orbital AI data-centre satellites. Reuters has separately reported that rivals reliant on SpaceX rockets are being squeezed. The causal link between the two, that the squeeze is being driven specifically through the Nvidia arrangement, is not stated in the Reuters headline and is not entailed by the Polymarket item.

A counterpoint reads the Nvidia headline as a procurement event, not a structural shift. Selection of a chip supplier for a specific satellite line is consistent with a normal supplier decision. What would push the reading toward structural is evidence about deployment scale, exclusivity, or downstream customer commitments, none of which the thread provides.

What the Reuters item does, and does not, attribute to SpaceX

The Reuters wire filed at 20:50 UTC frames SpaceX's satellite ambitions as squeezing rivals reliant on its rockets. The framing is on the supply side: rivals depend on SpaceX for launch, and SpaceX's own satellite buildout squeezes them. The thread does not specify which rivals, which satellite lines, or which markets. The Reuters item is a standalone wire piece; whether it casts the squeeze as a consequence of the Nvidia arrangement, of Starlink's existing constellation, or of any other specific SpaceX action is not specified in the available thread.

What we verified, and what we could not

What we verified against the source items: SpaceX reported $7.8bn in quarterly revenue, up 92% year-on-year, on 4 August 2026, in its first release as a public company. The release was reported to disclose roughly $100bn in cash. On the same day, SpaceX was reported to have selected Nvidia for its orbital AI data-centre satellites. Reuters filed a wire piece on 4 August 2026 describing SpaceX's satellite ambitions as squeezing rivals reliant on its rockets. The Polymarket contract opened at 19:57 UTC implies a sell-side consensus existed; the consensus figure itself is not in the available sources.

What we could not verify against the available thread: whether the release was a written report or an earnings call; the precise release time; segment-level revenue split between launch services and Starlink broadband; quarter-on-quarter or year-on-year net income; any forward guidance figures; the precise terms of the Nvidia arrangement; the specific rivals Reuters refers to; whether the squeeze framing is tied to the Nvidia selection or to other SpaceX action; and any statement on Starship test cadence or commercial schedule. The thread evidence does not tie the Reuters squeeze-framing to the Nvidia arrangement directly.

Monexus analysis: the day's items are best read as four separate relays that landed in the same window, not as four parts of a single orchestrated announcement. The headline numbers are real and large. The structural read, vertical integration on a scale that did not exist in the launch business five years ago, is a plausible inference but an inference, not a sourced claim. The next quarter's segment disclosure, if SpaceX provides it, is the test of whether the day was a procurement plus a launch-squeeze story, or the first public confirmation of a deeper integration thesis.

Desk note: Monexus treated the Polymarket and Unusual Whales items as relays of the underlying release, not as primary sources, and did not attempt to reconcile the Reuters squeeze-framing with the Nvidia arrangement beyond what the Reuters headline itself states. The vertical-integration framing was kept visible as assessment, not fact.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://x.com/Polymarket/status/2084733738181628068
• https://x.com/unusual_whales/status/2084738896726810760
• https://x.com/Polymarket/status/2084743433726427489
• http://reut.rs/4heVIDo
• https://x.com/Polymarket/status/2084730340149809430
• https://x.com/Polymarket/status/2084724061200924756
• https://x.com/Polymarket/status/2084650061519331697
• https://x.com/Reuters/status/2084743700811304962