SpaceX (SPCX)

Executive Summary & Action Plan

Verdict

SELL (Reduce into the lockup wall) — The S-1 finally shows what the tape is pricing: a superb connectivity monopoly (64% segment EBITDA margin) welded to a ~$10B/yr AI burn via the February xAI merger, with ARPU down 33% in nine quarters and 2.6B insider shares becoming saleable four weeks from now. Our $124 fair value sits below the $145 tape, and the supply calendar owns the next two quarters.

Price $145.30 | Market cap $1.91T (fully diluted) | Target $124 | Upside -14.7% IWANNAVY Fair Value · price as of 2026-07-11 close · Street consensus $242.22 (reference; n=18, 304.7% dispersion, range $62–$800) — the widest disagreement in our universe, because SPCX trades on scenario capitalization, not on any comps framework

Abstract

The S-1 reframes SpaceX: the February 2026 all-stock xAI merger (xAI at $250B) folded Grok, X, and the Colossus compute build into the listed entity, so Q1'26's −41.6% operating margin and −91% net margin are conglomerate accounting — the AI segment lost $2.47B and spent $7.7B of capex in one quarter, while the core Space+Connectivity business earned a +14% operating margin and Connectivity produced $2.09B of EBITDA at a 64% margin. The engine has a governor, though: disclosed ARPU fell $99→$66/month in nine quarters as growth shifted to emerging-market tiers, launch revenue fell 28% as internal Starlink flights crowd the manifest, and FCF ran −$9.1B in Q1 against $23.7B of cash. Our $124 fair value is the sleeve-level SOTP-DCF (launch, Starlink, Starshield, plus explicitly option-priced Starship/xAI); comps-anchored lenses land at only $46–48, which we treat as the air-pocket floor, not the base. Rating: SELL/reduce — ~2.6B shares unlock around 8/10 and ~34% of all shares by 9/10, against no remaining index bid.

Action Plan

Supply owns the tape into December (Facebook-2012 analogue maps $95–125 before clearing). Reduce now and into any pre-print rally toward the $175.50 early-release trigger; re-enter at $124 only after the first tranche clears, size at $95.

• Risk/Reward 0.65:1 at market (bull +43% vs bear −67%) | Prob-weighted 12M return −18.8% (Bull 20% × +43.2% + Base 50% × −14.7% + Bear 30% × −67.0%) | Confidence: Low — two disclosed quarters, one month of trading

---

Last Four Quarters

Only two quarters exist publicly (S-1 filed 5/20). The YoY compare is the whole story: headline revenue +15.4% is mostly the xAI consolidation ($818M of AI segment revenue that didn't exist in Q1'25), while ex-AI revenue was roughly flat and Space fell 28%. The margin collapse decomposes cleanly: −53pp from the AI segment loss, −20pp from Starship R&D expensing ($930M in the quarter), plus a satellite-depreciation step-up (~$0.9B/qtr) — the core business stayed profitable.

GAP verification (−4.51% on the 7/10 session): lockup front-running, no fresh fundamental news — the stock printed an all-time low ($145.07) after fully absorbing ~$4.3B of Nasdaq-100 forced buying (added 7/7), the clearest possible supply signal.

---

Revenue & Profit Mix

Three reporting segments since the merger. Connectivity carries the company; AI consumes it; Space is the famous monopoly and the smallest line. Q1'26 per the S-1.

---

Business Lines

• Starlink Connectivity (69% of revenue, 64% segment EBITDA margin): 9,600+ satellites, 10.3M subs (+45% YoY). Must know — unit economics are bifurcating: subs nearly doubled while ARPU fell 33% since 2023 ($99→$66); enterprise/mobility/direct-to-cell (10–50x consumer ARPU) is the real dollar driver. This EBITDA services the debt and funds xAI.
• Space / Launch (13% of revenue, −28% YoY): 80%+ of global orbital mass. Must know — the monopoly monetizes indirectly: it lets SpaceX deploy Starlink at cost (Kuiper pays retail for launch). As a segment it is small and shrinking.
• Starshield / Government (~20% of total company revenue across segments): Must know — highest-quality recurring revenue ($22B cumulative awards + $6.45B of May-2026 wins + NSSL-3 ~$5.9B ceiling), but concentrates key-man/political risk on one individual with 85.1% voting power.
• AI — xAI/Grok/X/Colossus (17% of revenue, ~100%+ of losses): Must know — this is why the listed entity exists in its current form: −$2.47B quarterly op loss, $7.7B quarterly capex, Anthropic compute contract on Colossus through May 2029; the Musk award (1.0B Class B shares, ~$93B max) currently carries zero recognized expense under ASC 718 — a latent expense overhang.

---

IWANNAVY Fair Value

Sleeve-level SOTP-DCF is the primary: launch (defense-like, 8.5% WACC) + Starlink (sub growth vs ARPU deflation modeled separately) + Starshield (government counterparty) + Starship/xAI as probability-weighted options = $124, with ~25% of base EV being pure optionality. Comps-anchored lenses ($46–48) can't reproduce the price even at ceiling multiples — they are the discipline floor if scenario capitalization dies. Our 6/18 $145 mark is cut to $124 on the steeper disclosed ARPU path ($66 vs ~$81 modeled) and the consolidated xAI burn.

• Thesis breaker (for the SELL): an 8/6 print showing AI-segment losses peaking, Starlink ARPU stabilizing on the May price increases, and clean absorption of the first tranche — that combination plus a Starship V3 success re-opens the bull path ($208) and would force our floor lenses to concede durable scenario premium.

---

Catalysts & Risks

---

References

• SpaceX Form S-1/S-1A (SEC EDGAR, filed 2026-05-20) — segment P&L, ARPU disclosure, lockup schedule, Musk award · CNBC — Nasdaq-100 fast-track add (2026-07-06)
• Yahoo Finance — ~31% of tradable float now short (2026-07-08) · Investing.com — lockup countdown mechanics · Motley Fool — first earnings: what to watch (2026-06-22)
• Fortune — Musk award carries zero ASC 718 expense (2026-06-06) · CNBC — Shotwell: "I do not want to focus on quarterly earnings" (2026-06-12)