Ron Baron, founder of the US fund manager Baron Capital, has shifted his case for how big Starlink’s revenue can get away from household broadband and towards artificial intelligence, according to a September 17, 2026 report by 24/7 Wall St. The outlet reported that Baron Capital holds roughly $25 billion in SpaceX shares and that Baron has staked a headline projection that Starlink alone can reach $1 trillion in revenue.
Key facts
- Baron Capital holds about $25 billion in SpaceX shares, its biggest illiquid conviction, now a public stock, 24/7 Wall St reported.
- Baron’s earlier model projected $1 trillion from Starlink within a decade, against reported revenues of $18 billion last year.
- His current operational claim is that 100,000 satellites are needed, with each generation deorbited every five years and replaced by hardware ten times more capable.
- SpaceX’s second-quarter connectivity revenue — the line that carries the satellite internet business — was $4.29 billion, up 66% year over year, with enterprise and government inside it up 108%.
- SpaceX spent $18.37 billion on capital expenditure in the second quarter alone, $15.83 billion of it on AI compute, and placed a $25 billion inaugural investment-grade bond issuance, according to Baron Capital as cited by the outlet.
What Baron is now arguing
Baron has pivoted, 24/7 Wall St reported, and now says what will overwhelm everything are agentic agents — machine users that run continuously rather than in bursts. The outlet described the contrast he draws: a human subscriber generates traffic in short sessions with long idle gaps, while an AI agent booking travel, monitoring a factory line or negotiating a contract does not sleep, does not close the browser, and multiplies per account rather than per person.
Describing the model he had used before that shift, Baron was quoted in the report as saying: “SpaceX was planning on going from 10,000 satellites to 20,000 satellites.” 24/7 Wall St said Baron treats the agentic-traffic shift as an observation, but that it is an assumption, because agentic workloads at commercial scale barely exist yet.
What it means for Starlink price and availability
The only consumer price signal in the report is revenue per user: 24/7 Wall St reported that consumer ARPU slipped from $85 to $66 in the same quarter that enterprise and government connectivity more than doubled. The outlet does not forecast Starlink prices in any market, does not address regional plan tiers, and says Baron’s $1 trillion figure no longer depends on residential pricing rising. Read strictly, the reporting supports only that the household plan is a shrinking share of the revenue mix.
On this site’s own reading — an inference neither Baron nor 24/7 Wall St makes — a business priced around enterprise capacity has less reason to defend a consumer price point and more reason to manage congestion, which would show up as cheaper, more heavily shaped tiers where there is spare capacity and slower availability where there is not. That remains a prediction, not a reported fact.
Availability is a licensing question before it is a satellite question. A constellation of the size Baron describes would need spectrum and landing rights country by country, a process that in Africa has put regulators rather than incumbent telcos in the frame.
What the quarter does and does not prove
The revenue mix
The outlet treated the quarterly results as partial support: growth is running fastest in exactly the segment Baron now points to. It stopped short of saying the figures establish machine traffic as the cause, since enterprise and government demand also covers maritime, aviation and state contracts that predate any agentic workload.
The capital bill
Against that, the quarter’s capital spending ran more than four times the connectivity revenue it supported, most of it on AI compute and part-funded by the bond issuance. Baron’s five-year refresh cycle, the outlet noted, only closes if launch cadence, manufacturing throughput and cost per kilogram all move together. Starship V3, which SpaceX itself says will cut cost to orbit by 99% or more, is the piece the outlet says the model has no fallback for. Our September 28 report on Starship Flight 14 framed the same dependency from the launch side, projecting that launch economics would take the cost of delivered Starlink capacity down roughly 16-fold if V3 flies as advertised.
Where the report says the thesis is weakest
24/7 Wall St challenged Baron’s ground-versus-orbit cost comparison, saying nothing in space actually runs on a one-time cost basis. Its support for that is its own characterisation of an April 2026 US Government Accountability Office review of space-based data centres, which the outlet said flagged cooling at scale as unproven — heat does not disperse easily in near-vacuum — and on-orbit servicing as underdeveloped. The piece neither links nor quotes the GAO document, and Starlink News has not been able to match those findings to a published GAO report, so they should be read as the outlet’s account of the review rather than verified GAO conclusions.
The outlet also put a question to Baron’s ratio that it said he did not answer: what wears out first, and who fixes it.
The bear case as stated
24/7 Wall St set out a bear case in which the revenue ceiling was never the binding constraint. On that reading, the constraint is simultaneous execution across constellation expansion, orbital compute and power generation, with the Starship V3 test-flight cadence as the near-term tell. The outlet added that Baron is a holder talking his book, with an unusually large incentive to sound certain about a ten-year forecast it described as a stack of compounding assumptions.
Frequently Asked Questions
Where is SpaceX stock trading relative to its IPO?
SPCX changed hands at $150.88 on September 16, 2026, up 5.15% on the session but still 6.26% below its post-IPO reference price, according to 24/7 Wall St. The outlet put the company’s market capitalisation near $1.16 trillion on that date.
What cost comparison does Ron Baron use for orbital versus ground infrastructure?
Baron said a terrestrial unit costs about $10 in capital and $2 a year in maintenance, while in space he claims it costs $2 one time, 24/7 Wall St reported. The outlet objected that the "unit" in that comparison is undefined, noting that a rack, a server, a watt of compute and a satellite bus are not interchangeable.
What does 24/7 Wall St say the bull case for Starlink rests on?
The outlet set out a bull case in which Baron is right about demand and directionally right about the constellation. It cited SpaceX’s 12.0 million Starlink subscribers, a $47.5 billion backlog and 1.4 GW of AI compute already in place as a head start it said no competitor can match this decade.
Is the $1 trillion figure a SpaceX forecast?
No. 24/7 Wall St presented it as Ron Baron’s own projection as a shareholder, not guidance from SpaceX. The numbers the outlet sources to the company and to Baron Capital — quarterly connectivity revenue, capital spending and the bond issuance — are historical; the $1 trillion revenue figure and the 100,000-satellite requirement are Baron’s characterisations of where the business goes next.
Follow Starlink News on Google. Make us a preferred source to see more of our reporting in Google search results.
