Working bottom-up from real TAMs, the author shows that hitting $4.3T requires SpaceX to capture multiple entire markets simultaneously — including the entire commercial launch market, a telecom-ARPU-sized slice of Starlink subscribers, and a meaningful fraction of long-haul aviation revenue. Even Morgan Stanley and Citi's most aggressive 2040 projections cap the entire space economy near $1T, making a single-company $4.3T figure mathematically incoherent.
The author singles out the point-to-point Earth transport line as the most speculative component, noting it is modeled as if it captures meaningful share of an $850B long-haul aviation industry that runs on notoriously thin margins. Treating a market that doesn't exist at any meaningful scale today as a multi-hundred-billion-dollar revenue line by 2040 is where the model most obviously breaks.
The piece explicitly accepts Falcon 9 reuse, ~7M Starlink subscribers, and successful Starship catches as legitimate engineering wins. The critique is narrower: late-stage private valuation math has adopted a pattern of picking distant horizons and assuming total market capture, and developers and investors should learn to read TAM-anchored forecasts skeptically regardless of how strong the underlying technology is.
An analysis published on matteast.io tears into the increasingly common claim that SpaceX will hit $4.3 trillion in annual revenue by 2040 — a figure that has been floated in investor decks, secondary-market pitches, and bullish Twitter threads ahead of the next tender offer. The author works the math the boring way: stack the company's three real businesses (Falcon/Starship launch, Starlink connectivity, and a hypothetical point-to-point Earth transport line) against the actual addressable markets they sell into, then ask what penetration rate is required to hit $4.3T.
The answer is uncomfortable. The global commercial launch market is roughly $15-20B/year today; even the most aggressive 2040 projections from Morgan Stanley and Citi top out near $1T for the entire space economy — not one company's slice of it. Starlink, the most plausible cash engine, would need to absorb something on the order of the entire global telecom ARPU pool to contribute its share. Earth-to-Earth Starship transport — moving passengers New York to Shanghai in 39 minutes — is priced in the model as if it captures a meaningful fraction of long-haul aviation, an industry whose total 2023 revenue was about $850B and whose margins are notoriously thin.
The piece is not a hit job. It accepts SpaceX's engineering wins as given: Falcon 9 reuse is real, Starlink has ~7 million subscribers as of early 2026, Starship has flown enough successful catches that the production cadence question is now serious. The argument is narrower and harder to dodge: you cannot reach $4.3T without quietly assuming SpaceX wins multiple markets in their entirety, including markets that don't exist yet at any meaningful scale.
This kind of forecast math has become the dominant valuation language for late-stage private companies, and developers building inside the resulting ecosystems should know how to read it. The pattern is familiar: pick a 2040 horizon (far enough that nobody can falsify it, close enough to feel concrete), invoke a TAM that includes adjacent industries, apply a penetration rate that sounds modest in percentage terms but is enormous in absolute dollars, multiply. The output is a number large enough to justify the current secondary valuation — SpaceX last marked at roughly $400B — and the loop closes.
The matteast.io analysis is useful because it does what most coverage refuses to do: it states the implied market share. If you back out the model, Starlink's required 2040 revenue implies more connectivity revenue than every mobile carrier on Earth collectively generates today. That is not a forecast. That is a category error dressed as a spreadsheet. The same critique applies to the Earth-to-Earth line, which would need to displace not just first-class aviation but a substantial chunk of business-class long-haul to pencil — at ticket prices that have never been tested with paying passengers and against safety/regulatory regimes that took commercial aviation 60 years to negotiate.
The community response on Hacker News (146 points, top comments from people who've actually built financial models for satellite operators) splits along a predictable line. The bulls argue that SpaceX has repeatedly violated linear extrapolation — Falcon 9's cadence in 2024 alone exceeded every other launch provider combined, and Starlink went from zero to profitable cash flow faster than any telecom buildout in history. The bears argue that physics, spectrum allocation, and the FAA don't care about momentum. Both can be right: SpaceX can be the most important industrial company of the decade and still not hit $4.3T, and the gap between those two statements is where retail investors get hurt.
There's a second-order point worth naming. The Mars program — the actual stated mission — is not in the $4.3T model in any meaningful revenue line, because Mars doesn't have customers. The trillion-dollar number is what's required to justify the valuation that funds the Mars program; the Mars program is what justifies the trillion-dollar number. It's a closed loop, and closed loops are how financial gravity gets ignored until it isn't.
For developers, the practical surface area here is narrower than the headline number suggests but real. If you're building anything that depends on Starlink as infrastructure — IoT fleets in remote geographies, maritime APIs, edge inference in places without fiber — the unit economics that matter are not the 2040 TAM but the next 24 months of pricing. Starlink Business has already raised prices twice; the consumer tier subsidy is what produced the subscriber curve everyone quotes. Build your cost model on the assumption that connectivity prices normalize toward telecom-comparable ARPU, not toward zero, because the forecast pressure runs in exactly that direction.
If you're shipping anything on Starship as a launch vehicle — small-sat constellations, hosted payloads, lunar logistics contracts — the relevant variable is cadence, not the $4.3T narrative. Falcon Heavy was supposed to be cheap rideshare and turned out to be expensive and rare; Starship's actual per-flight economics will depend on reuse cycles that haven't been demonstrated at production volume yet. Plan your launch manifest with a fallback to Falcon 9 and at least one non-SpaceX option, because the failure mode of a single-provider stack is the same here as it is in any cloud lock-in scenario.
And if you're being pitched on SpaceX secondaries by a wealth manager who quotes the $4.3T number with a straight face: the matteast.io spreadsheet is the rebuttal. Ask which line items survive being cross-checked against the actual 2040 addressable market, and watch what gets quiet.
The most useful framing here isn't "SpaceX is overvalued" or "SpaceX is undervalued" — it's that forecast math at this scale has stopped being a financial model and started being a narrative device. That's not unique to SpaceX; it's how OpenAI, Anthropic, and the entire frontier-model cohort are now priced too. The discipline for engineers is the same discipline you apply to your own systems: when a number requires capturing 100%+ of an adjacent market, treat it as a structural bug, not a forecast. The model is telling you something true, just not what it claims.
You never know what inflation gonna be in futute. In some countries that turned into autocracies with strong and long standing leaders who love traditional values and religion inflation can easily be 30-70% a year.Then not only 4.3T reachable, but even 43T.
It's funny it concludes "A line is not a law" while the entire page is just lines and charts. If you didn't know anything going in, SpaceX might as well be a company selling toilet paper for all you learned.I've got no idea what a share of SpaceX is worth today or 14 years f
It's insane to me that regulators are just letting the forced bid hack unfold with their heads in the sand as if they are unaware of what's happening or the press it's receiving.Sometimes there are clever ways to hack the system and the regulators can react and blame poor hindsight, t
It's amazing we're in a timeline were people let stuff like this happen, against any sane logic.
Top 10 dev stories every morning at 8am UTC. AI-curated. Retro terminal HTML email.
[flagged]