Deep Research

SpaceX (private)

5 min read

Snapshot

  • Status: Private — but barely. SpaceX has priced the largest IPO in history at 1.77T valuation, ticker SPCX), with a Nasdaq debut scheduled for June 12, 2026 — two days after this page’s date.
  • Sector: Launch, satellite broadband (Starlink), and — per Gavin Baker’s thesis — future orbital compute.
  • Latest known valuation: ~135 fixed IPO price; SpaceX plans to sell 555.6M shares for a ~11.2B greenshoe).
  • Atreides involvement: Confirmed long-standing holder; the single largest position in Atreides’ venture book (Bloomberg, Jul 2022). Position reportedly worth ~$2.5B as of April 2026 (GuruFocus — single source, treat as reported, not confirmed).
  • Likely vehicle(s): Primarily the Foundation Fund’s VC sleeve, plus at least one confirmed Special Circumstances deal-by-deal SPV.

Business Overview

SpaceX operates the world’s dominant launch business — responsible for over 90% of global mass to orbit — and Starlink, the largest satellite constellation ever built. The economics have inverted from the company’s launch-services origins: in Q1 2026, connectivity revenue (mostly Starlink) reached 4.69B total revenue (CNBC, from IPO filings). Starlink is the recurring-revenue engine that justifies a trillion-dollar-plus multiple; launch is the moat that makes the constellation’s unit economics possible. Starship — still in test/early-operational phase — is the variable that determines whether per-kilogram launch costs fall by another order of magnitude.

The IPO, at a 1.6T). Morgan Stanley and others have published forecasts in the same range. Whether the valuation holds depends almost entirely on Starlink subscriber/ARPU trajectory, Starship cadence, and optionality narratives (direct-to-cell, government constellations, and orbital data centers).

Atreides’ Involvement

  • Pre-Atreides lineage (confirmed): Baker covered and invested in SpaceX during his 18 years at Fidelity — GuruFocus dates his first investment to 2015 at Fidelity (the user-facing record varies; he was running the Fidelity OTC fund and Fidelity first invested in SpaceX in 2015). This is one of the longest-running relationships in the Atreides portfolio.
  • Largest VC-book position (confirmed): Bloomberg reported in July 2022 that SpaceX was the largest position in Atreides’ venture portfolio, alongside names like Epic Games.
  • Special Circumstances SPV (confirmed via leaked documents): TechCrunch’s review of internal SpaceX investor documents (published 2024, updated Jan 2025) shows an **“Atreides Special Circumstances Fund” allocated nearly 429,000 shares at a cost of nearly 70/share implied). At the 58M — meaning the SPV is a small slice of the total position.
  • Total position size (reported, single source): GuruFocus reported the overall Atreides SpaceX stake at **~135/share that implies a holding in the high-teens of millions of shares, the bulk of which must sit in the Foundation Fund’s VC sleeve rather than the disclosed SPV. The share-count math is our inference, unconfirmed.

Why Atreides Owns It

SpaceX is the purest expression of Baker’s “watts and wafers” framework extended to its logical endpoint: if AI compute is ultimately constrained by power and cooling, the cheapest watts and the cheapest cooling are in orbit. On Invest Like the Best (EP.451, Dec 2025), Baker argued that data centers in space are “superior in every way” to terrestrial ones — 24-hour sunlight, ~30% more intense solar flux without atmospheric loss, free radiative cooling — and that orbital compute could become “the most important technological breakthrough” within three to four years, mainstream within five to six. He has publicly mocked physics-based skepticism on X, pointing out that Musk operates two of the world’s largest coherent GPU clusters (xAI) and the only launch system (Starship) that can economically lift the mass required.

That makes SpaceX a triple thesis for Atreides: (1) Starlink as a compounding broadband monopoly-in-the-making, (2) Starship as the cost curve that breaks every competitor’s business model, and (3) the only credible platform for the orbital-compute era — the same “own the bottleneck” logic behind the firm’s optics, memory, and power positions in the public book. Alphabet’s Project Suncatcher and reports of OpenAI exploring SpaceX-rival acquisitions suggest the thesis has moved from contrarian to consensus-forming, which is exactly the window where a crossover holder of pre-IPO stock harvests the re-rating. The June 12 IPO converts a seven-plus-year illiquid conviction into a markable, tradeable position — and will finally make the stake 13F-visible from Q2 2026.

Risks

  • Valuation: $1.77T prices in flawless Starlink growth and substantial Starship/orbital-compute optionality; any cadence slip (Starship failures, Starlink churn, ARPU compression from competition like Kuiper) compresses the multiple sharply.
  • Key-man and governance: Musk concentration risk — attention split across Tesla, xAI, and politics; the IPO structure preserves his control.
  • Orbital data centers are unproven: Baker’s differentiated thesis rests on radiation hardening, thermal management at scale, serviceability, and Starship economics that do not yet exist operationally. A 5-6 year timeline can easily become 10.
  • Regulatory/geopolitical: spectrum disputes, orbital-debris regulation, and SpaceX’s entanglement with US national-security programs cut both ways.
  • Single-source position sizing: the ~$2.5B Atreides figure is one outlet’s report; actual size, cost basis, and any pre-IPO trimming are unknown.
  • Post-IPO flowback: a $75B+ raise plus early lockup expiries could pressure the stock just as Atreides’ position becomes markable.

Sources

Backlinks

Graph View