Morningstar pegs SpaceX 30% below the $400B whisper number

What happened

Morningstar published a pre-IPO note arguing that SpaceX is worth roughly $280 billion, about 30% below the ~$400 billion valuation circulating in private secondary markets ahead of the expected 2026 listing. The post climbed to 175 points on Hacker News, where the comment section turned into a referendum on whether public-market analysts can model a vertically integrated launch + connectivity + (eventually) Mars-logistics company using anything resembling a standard DCF.

The analyst's math is unromantic. Falcon 9 and Falcon Heavy revenue is modeled off a launch cadence that's already close to the 2025 run rate — Morningstar isn't betting against reusability. Starship is in the model as an option, not a hockey stick. The disagreement with the private-market mark is concentrated almost entirely in one line item: Starlink subscriber growth and ARPU through 2030. Private investors are pricing in something close to a global consumer-broadband monopoly in underserved geographies plus a meaningful enterprise/maritime/aviation book. Morningstar is pricing in roughly half that, on the argument that terminal subsidies are still suppressing unit economics and that Amazon's Kuiper plus the Chinese Guowang and Qianfan constellations will compress margins before the moat sets.

The $120B gap between Morningstar's number and the secondary-market mark is not a disagreement about rockets. It's a disagreement about how durable Starlink's terminal economics are once a second LEO constellation hits commercial scale.

Why it matters

For a developer audience, the instinct is to dismiss this as a finance story. It isn't. SpaceX is now load-bearing infrastructure for a non-trivial chunk of the internet — Starlink crossed 6 million subscribers in early 2025, sells direct-to-cell service via T-Mobile, and is the de-facto backhaul for a growing set of edge deployments where fiber economics don't pencil out. The valuation question is really a capacity-planning question dressed in a suit.

Consider the second-order effects. If Morningstar is right and Starlink's growth curve is closer to a regulated utility than a hyperscaler, capex on the v3 satellites slows, Starship cadence stays tied to NASA and DoD contracts rather than self-funded throughput, and the price-per-GB curve for LEO connectivity flattens earlier than the bulls expect. If the private market is right, Starlink is on a path to be the dominant non-terrestrial network globally, and every competing constellation is racing to be a viable #2 before the cost-per-gateway gap becomes unbridgeable.

Either way, the practical question for anyone designing a multi-region system in 2027-2030 is whether you'll have one credible LEO provider or two — and the IPO prospectus is going to tell you which world the smart money thinks we're in. The HN thread surfaced an underappreciated point: Kuiper's first operational batch is supposed to hit commercial service in 2026, and the Chinese constellations are launching faster than Western press coverage suggests. The duopoly-vs-monopoly question may be resolved before SpaceX even prices.

There's a governance angle the Morningstar note sidesteps but the comments don't. SpaceX is structurally a controlled company — Musk's voting stake and the dual-class structure expected at IPO mean public shareholders are minority partners on a platform whose CEO is, to put it gently, not always optimizing for shareholder return. Analysts can model launch cadence; they cannot model the discount rate on key-person risk for a founder who runs five companies and a social network. Morningstar's 30% haircut is arguably *too small* once you price that in, which is the contrarian read worth holding in your head.

The other thing worth noting: this is the first major pre-IPO note where a tier-1 sell-side shop has put a number meaningfully below the private mark. Secondary-market SpaceX shares have traded up and to the right for six years on the assumption that public-market validation would be a formality. Morningstar just made it not a formality.

What this means for your stack

If you're architecting anything that touches non-terrestrial networking — IoT fleets, maritime telemetry, rural CDN edges, disaster-recovery backhaul — the valuation debate maps directly onto your vendor-risk spreadsheet. A Starlink that IPOs at $400B and uses the proceeds to crush Kuiper before it scales gives you one excellent provider and no leverage on pricing. A Starlink that IPOs at $280B and spends the next three years defending margins against Kuiper and Guowang gives you a real second source by 2028.

Plan for the second scenario. Concretely: don't sign multi-year exclusive contracts with any single LEO provider in 2026. Design your radio and antenna abstraction layer so that swapping terminals is a config change, not a forklift. If you're building developer-facing APIs on top of LEO connectivity (and a surprising number of YC companies in the last two batches are), keep the satellite-network specifics behind an interface — the underlying economics are about to get genuinely competitive for the first time, and the winning provider in 2030 may not be the winning provider today.

For anyone holding SpaceX secondary shares through a Forge or EquityZen vehicle: Morningstar's number is the first credible public anchor. The bid-ask spread on private SpaceX has been wide and lightly traded; expect the secondary market to compress toward the IPO range as the S-1 approaches, and don't be surprised if that direction is down.

Looking ahead

The S-1 will settle most of this. Two numbers matter more than the rest: Starlink ARPU by segment (consumer vs. enterprise vs. mobility) and gross margin on the v3 terminal. If either is meaningfully better than Morningstar's model, the bull case holds and the IPO prices closer to $400B. If either disappoints, $280B is the floor and the gap closes from above. Either way, the era of pricing SpaceX off Musk's tweets is ending — and the engineers building on top of LEO connectivity will get the first honest look at the unit economics underneath their backhaul.