SpaceX runs a space program on layaway. Heavy advances from Washington, profits booked later, tax bill delayed indefinitely.
The federal dependency is stark. In 2020, about 84% of revenue came straight from federal contracts—roughly $1.4 billion that year and $1.7 billion in 2021 from NASA, the Pentagon, and other agencies.
The real kicker is the NOL wall. By the end of 2021, SpaceX had stacked roughly $5.4 billion in accumulated tax losses. Thanks to the 2017 tax law change, nearly $3 billion of those losses can be carried forward with no expiration. There's an 80% annual cap post-2020, but with credits layered on, the effective federal bill can stay paper-thin for years.
Legal? Yes.
Designed for survival-stage companies? Also yes.
SpaceX is well past survival.
The operating numbers tell the story. EBITDA roughly doubled to around $5 billion last year from $2.6 billion in 2023. Total revenue hit about $7.4 billion in 2023, with internal guidance pointing to ~$15.5 billion in 2025.
Starlink does the heavy lifting: roughly 6 million subscribers generating about $8 billion in 2023 revenue.
A critical detail often buried in footnotes: SpaceX reported paying $483,000 to foreign governments and $78,000 to U.S. states in 2021. In 2020–2021, it logged just $6,000 of income taxes without specifying where that went. Federal? Doesn't look like it. Financial reviewers called this posture "substantial and notable" given the heavy reliance on U.S. contracts.
This isn't just "a startup burning cash." The contract pipeline runs decades deep—NASA and DoD obligations measured in the tens of billions across programs, from ISS cargo and crew to intelligence payloads. SpaceX earned that business by making launches cheaper and more frequent.
Still, the public underwrites the climb while the private side pockets the compounding returns.
The Numbers in One Clean Take
Loss carryforwards: ~$5.4 billion by 2021; ~$3 billion indefinite post-2017
Federal revenue share: ~83.8% in 2020
Revenue trajectory: ~$7.4 billion (2023) → guidance ~$15.5 billion (2025)
EBITDA growth: ~$2.6 billion (2023) → ~$5 billion (2024)
Starlink metrics: ~6 million subscribers, ~$8 billion revenue (2023)
Taxes paid (2021): ~$483k foreign, ~$78k state; plus $6k income tax (2020–2021)
No need for moral posturing. Just change the rules where the incentives bite.
1. Contractor tax floor: If more than 50% of revenue in a given year comes from federal contracts, set a minimum federal income tax floor—say, 5–8% of pre-NOL taxable income. NOLs carry forward, but can't zero-out payments funded by taxpayers.
2. NOL gating for subsidy-backed firms: Losses generated in years with material federal support can offset only 50% of future income tied to those same product lines. Keep the 80% general rule for everyone else; tighten it for government-dependent giants.
3. Claw-forward, not clawback: Create a public return mechanism: once EBITDA clears a threshold (say, 20% margin or $3 billion, whichever is higher) and federal revenue remains above 30%, skim 1–2% of revenue as a surtax until NOLs from subsidized build years are exhausted. Predictable and hard to game.
4. Contract pricing hygiene: Build tax posture into bid scoring. If a bidder's effective federal rate sits below 5% over a rolling 3-year window due to NOLs, discount their non-price factors or require price give-backs. No scolding—just math in the RFP.
The point isn't punishment. It's symmetry.
The public pays to light the fuse; the public should see a return when the rocket clears the tower.