Skip to content

Podcast

The Most Simplified Breakdown of the SpaceX IPO on the Internet

My First Million

Source ↗ ← All highlights
  • SpaceX Is Multiple Businesses Stapled Together
    • SpaceX is four businesses stapled together: rocket launches, Starlink internet, X/XAI, and TerraFab chip and data center ambitions.
    • Launch dominance funds Starlink growth while XAI and TerraFab act as experimental, capital-hungry bets tied by Elon’s vision. (Time 0:01:20)
  • Starlink Is The High Margin Cash Cow
    • Starlink is SpaceX’s cash cow with ~10 million subscribers, ~$11B revenue, and ~40% EBITDA margins.
    • It solves rural, maritime, and crisis connectivity and accounts for ~40% of launches, creating recurring high-margin revenue. (Time 0:08:53)
  • Elon Nuked And Restarted The Starlink Team
    • Sam recounts how Elon ‘nuked’ the initial Seattle Starlink team and rebuilt it after poor progress.
    • The restart produced the fast-growing Starlink unit that later became a core revenue engine. (Time 0:10:38)
  • Starship And Space Data Centers Are The Big Bets
    • SpaceX’s next unlocks are Starship and space data centers; Starship multiplies payload and lowers per-launch cost dramatically.
    • Elon argues space lets you harvest solar energy and radiative cooling to offer far cheaper AI compute than Earth data centers. (Time 0:14:16)
  • Elon Turns Factories Into Rentable Infrastructure
    • Elon leverages vertical factory expertise: he builds massive GPU clusters (Colossus) and then rents spare capacity to top AI firms like Google and Anthropic.
    • Short-term contracts (90-day cancel) give revenue but create customer concentration and uncertainty. (Time 0:26:44)
  • Separate Cashflows From Optionality When Valuing SpaceX
    • Don’t confuse valuation math with business quality; Charlie Munger-style investors may call SpaceX a wonderful business at a silly price.
    • Consider whether you’re buying current cashflows (Starlink) or optionality on future tech (Starship, space compute). (Time 0:29:03)
  • Launch Dominance Created A Huge Cost Moat
    • SpaceX dominates launches (~80–85% of payload to orbit), which created a massive cost advantage by reducing cost per kilogram ~50–100x over legacy.
    • That launch moat enables Starlink scale and future low-cost payload strategies like Starship. (Time 0:29:18)
  • Elon’s Large Stake Amplifies The Elon Premium
    • Elon retained unusually large ownership: ~42% of SpaceX and ~85% voting control after 20 years of fundraising.
    • That concentrated ownership and control amplify the ‘Elon premium’ investors price into the IPO. (Time 0:30:40)
  • Early Operators Like Antonio Made Huge Returns
    • Antonio Gracias, an early operational partner, holds ~7% and will make ~ $90B from the IPO.
    • He provided manufacturing and production help early on and even loaned Elon money during Tesla’s brink moments. (Time 0:49:20)
  • Backing The Obvious Winners Beats Seeking Sophistication
    • Gigafund’s strategy was simply to back every Elon company rather than try to be clever or diversified.
    • That approach looked unsophisticated at the time but paid off massively because picking the right horse and holding was enough.
    • Over the last 15 years many investors succeeded by buying dominant platforms (Google, Facebook, Amazon) and doing nothing else.
    • Sam and Shaan argue people chase sophistication and activity instead of the power of simple conviction and patience.
    • Examples: Bill Gates selling Microsoft and Chamath leaving his early Facebook stake show opportunity cost of over-trading. (Time 0:52:43)
  • Pessimists Get To Be Right But Optimists Get Rich
    • When facing audacious opportunities, don’t default to pessimism; optimists capture outsized returns while pessimists get to be right.
    • Use successful audacity (Elon’s long-term bets) to expand your own risk tolerance for big ideas. (Time 0:57:17)
  • Elon’s Pay Ties To Science Fiction Milestones
    • Elon’s compensation is outcome-based with sci‑fi milestones: massive share grants require both extreme market caps and real-world feats like a million-person Martian colony.
    • The AI CEO award demands 100 terawatts of non-Earth compute—an almost inconceivable bar. (Time 1:01:55)