SpaceX (SPCX) — Filings Digest

Prepared: 2026-07-15 · Desk: Coverage · Tier: 1 (SEC-reporting, re-tiered from Tier 2 on the 2026-06-12 Nasdaq listing date — see §0)

---

0. Reporting status

Space Exploration Technologies Corp. (Nasdaq: SPCX, CIK 1181412) is a newly public SEC reporting company. It priced its IPO at $135/share on 2026-06-11, raising ~$75B at a $1.75T valuation, and began trading on Nasdaq as SPCX on 2026-06-12 — the largest IPO in history (CNBC, TechCrunch, 2026-06-12). Prior to this listing SpaceX filed no SEC disclosure of any kind; the roster's prior Tier 2 classification is stale as of the listing date and this document supersedes it per PROCESS.md §2/§4.

On EDGAR (CIK 1181412) as of 2026-07-15:

Source: SEC EDGAR company filings page, CIK 1181412, live-pulled 2026-07-15 (cross-referenced disclosure-sweep note, financials.md, same date). Because no periodic report (10-K/10-Q) has yet been filed, the S-1/424B4 is the only audited-adjacent, line-item financial disclosure that exists — all FY2023–FY2025 and Q1 2026 figures in this digest and in the model trace to that document (via secondary compilations pending a direct primary-text pull; flagged in §6).

---

1. 8-K digest (newest first)

Item numbers and subject matter per EDGAR filing index; full text has not yet been pulled line-by-line in this pass — dates/items are primary-sourced from the EDGAR index, "what it says" is a directional read pending verbatim confirmation.

8-K — 2026-06-26

• Items: 8.01 (Other events), 9.01 (Financial statements/exhibits)
• What it says: Routine post-IPO disclosure item, no items indicating a material corporate action beyond ordinary-course disclosure.
• Takeaway: No model-moving content identified; retained for completeness of the filing record.

8-K — 2026-06-23

• Items: 8.01, 9.01
• What it says: Pricing-related disclosure, consistent with post-IPO stabilization-period reporting (over-allotment/greenshoe activity typically lands here for a listing this size).
• Takeaway: Watch for a follow-on 8-K or the first 10-Q to confirm whether the underwriters' option was exercised — affects final share count and net primary proceeds.

8-K — 2026-06-22

• Items: 7.01 (Regulation FD disclosure), 8.01, 9.01
• What it says: Regulation FD item suggests a public disclosure event (investor presentation, press release, or conference remarks) made contemporaneously to avoid selective disclosure.
• Takeaway: Likely investor-relations housekeeping in the first weeks of coverage; no balance-sheet or guidance change indicated by the item codes alone.

8-K — 2026-06-17

• Items: 5.02 (Departure/election of directors or officers)
• What it says: An officer or director change — item code alone doesn't specify who or in which direction.
• Takeaway: Standard post-IPO board/management formalization; worth a name-level follow-up if the model's governance section needs the exact officer roster, but not numerically material.

8-K — 2026-06-16

• Items: 1.01 (Entry into a material definitive agreement), 3.02 (Unregistered sales of equity securities), 9.01
• What it says: A material agreement plus an unregistered equity sale — consistent with closing mechanics around the IPO (e.g., a concurrent private placement or lock-up-adjacent agreement).
• Takeaway: Flag for the model-build stage: Item 1.01 filings can carry capital-structure or customer/supplier commitments (e.g., compute or launch-services agreements) that belong in the assumption stack once the underlying agreement is identified.

8-K — 2026-06-15

• Items: 3.02 (Unregistered sales of equity securities), 3.03 (Material modification of rights of security holders), 5.02 (Officer/director changes), 5.03 (Amendments to articles/bylaws; change in fiscal year), 7.01, 8.01, 9.01
• What it says: The IPO-closing omnibus 8-K — new charter/bylaws adopted, security-holder rights modified (conversion of pre-IPO share classes into the post-IPO capital structure), officer/director slate finalized, and the IPO closing itself announced.
• Takeaway: This is the highest-information 8-K in the record — it marks the formal transition from private captable to public capital structure. The model should treat 2026-06-15 as the effective date of the post-IPO share count and governance structure (see §2 below for share/voting detail as disclosed in the 424B4).

---

2. DEEP digest — S-1 / 424B4 (filed 2026-05-20, amended 2026-06-01/03, priced 2026-06-12)

Note on sourcing quality: the figures below trace to secondary compilations of the S-1 (Via Satellite, HL.co.uk, KraneShares, Vested Finance, Morningstar, an X/Twitter thread citing specific line items) rather than a verbatim primary-text pull, because full EDGAR document retrieval was not completed in this pass. This is flagged explicitly in §6 and should be closed out before the model locks its actuals column.

2a. Business segments (as presented in the S-1)

The S-1 discloses a three-segment structure, reflecting the 2026-02-02/03 all-stock merger with xAI Corp. (implied combined valuation $1.25T: SpaceX $1T + xAI $250B — CNBC, 2026-02-03):

Source: HL.co.uk S-1 breakdown, 2026; KraneShares S-1 takeaways; Vested Finance S-1 deep dive (all 2026-05-20 vintage, digesting the filed S-1).

Q1 2026 (first quarter reported post-filing): total revenue $4.694B; AI segment alone $818M revenue / -$2.469B operating loss / -$609M adjusted EBITDA — KraneShares/Vested Finance S-1 takeaways, 2026-05-20. AI-segment losses are the dominant swing factor in consolidated profitability (FY2025 AI operating loss of $6.355B exceeds the entire company's net loss of $4.9B — i.e., Connectivity + Space combined were net profitable before AI-segment losses are layered in).

2b. Summary financial statements as disclosed (S-1 actuals)

$B unless noted

Source: S-1 compiled figures via X/Mo Islam thread citing S-1 line items, 2026-05-20; corroborated by diversifiedchaos.substack.com "SpaceX by the Numbers"; Morningstar "6 Charts on SpaceX's Pre-IPO Financials"; satellitetoday.com (Via Satellite), 2026-05-20.

Read-through: capex/revenue moved from 42% (FY2023) to 80% (FY2024) to 111% (FY2025) — SpaceX is now spending more on capex than it books in revenue, driven overwhelmingly by AI compute buildout rather than Starship or Starlink production capacity. This is the single largest disclosed swing factor for the model build.

2c. Share count and class structure

The 2026-06-15 8-K (Items 3.02/3.03/5.03) marks the conversion from the pre-IPO private capital structure to the post-IPO public structure. Specifics not yet pulled verbatim from the 424B4 in this pass — flagged as an open item: the desk needs the exact post-IPO share count (basic and fully diluted), the dual/multi-class structure (SpaceX has historically maintained founder super-voting shares), and the resulting implied market cap reconciliation against the $135 IPO price and $136.12 last close (2026-07-14). Directionally:

• IPO priced $135/share, ~$75B raised in primary + secondary proceeds combined, implying a ~$1.75T pricing-basis valuation (CNBC/TechCrunch, 2026-06-12).
• First-day close $160.95 (+19%), implying a first-day market cap of roughly $2.0–2.1T (CNBC/Forbes/CNN, 2026-06-12).
• Stock has since round-tripped: peaked near $2.7T market cap within days of listing, then pulled back to $136.12 close on 2026-07-14 (Nasdaq quote, checked 2026-07-15) — roughly flat to the $135 IPO price and well below the post-listing peak.
• Implied shares outstanding at $1.75T / $135 ≈ 12.96B shares (arithmetic back-solve, not a disclosed figure — treat as approximate pending the actual share count from the 424B4 cover page).

2d. Use of proceeds

Not independently pulled from primary text in this pass. Secondary sourcing (Vested Finance S-1 deep dive) indicates the ~$75B of IPO proceeds is earmarked across Starship development/production, Starlink (Gen2/Gen3) constellation buildout, and AI-datacenter/compute infrastructure — consistent with the capex trajectory in §2b. Flagged as an open item for verbatim confirmation against the 424B4 "Use of Proceeds" section before citing specific dollar allocations in coverage.md.

2e. Governance / control — Musk voting power

Not independently pulled from primary text in this pass. SpaceX has historically maintained a founder-controlled voting structure (consistent with Musk's public statements about wanting to retain control given Starship/Mars mission risk). Flagged as an open item: the desk needs the exact post-IPO voting-power disclosure (Item 5.03/3.03 8-K on 2026-06-15 modified security-holder rights, which is where any share-class conversion mechanics would be codified). Given the pattern at other founder-controlled tech IPOs, a supervoting structure preserving Musk's board control is the working assumption but is unconfirmed pending a direct S-1 "Principal Stockholders" / "Description of Capital Stock" section pull.

2f. Risk factors that actually matter

Directional read from secondary S-1 coverage (The Information, Vested Finance, KraneShares) — not a verbatim risk-factor list, but the risks that recur across independent write-ups of the filing:

1. AI segment cash burn and unproven unit economics. FY2025 AI operating loss ($6.355B) exceeds total company net loss; Q1 2026 AI segment burned $2.469B operating loss on $818M revenue. If AI-segment losses don't narrow, they structurally offset Connectivity's profitability.
2. Single-counterparty AI compute concentration. Secondary S-1 commentary (Vested Finance) references a Cloud Services Agreement with Anthropic (~$1.25B/month compute capacity through May 2029, signed May 2026) as a disclosed but single-sourced-in-our-sweep item — flagged as unverified; needs confirmation against the primary S-1 text or Anthropic's own disclosures before being treated as a modeling input.
3. Starlink ARPU decline. $99/mo (2023) → $66/mo (Q1 2026), -33%, explicitly framed in secondary S-1 coverage as an intentional volume-over-pricing strategy — a risk if subscriber growth ever decelerates faster than price cuts.
4. Capex outrunning revenue. Capex/revenue at 111% in FY2025; self-funding capacity is a live question even with $75B of fresh IPO proceeds, given the AI-segment burn rate.
5. Starship program/regulatory risk. FAA grounded the program after the Flight 12 mishap (2026-05-22, booster loss on Super Heavy boostback burn); recurring FAA mishap-investigation cadence (Flight 8, Flight 9, Flight 12) is a structural, recurring regulatory friction point for the Space segment, which is already loss-making.
6. Founder/key-person and control risk. Musk's central role across SpaceX, Tesla, xAI, and X — plus the concentration of voting control likely retained post-IPO (§2e) — is a standard but material risk factor at this scale of combined enterprise.
7. Federal revenue concentration. ~20% of company revenue from US federal agencies (NASA, Space Force) — a customer-concentration and political/budget-cycle risk, more acute for the already-loss-making Space segment.

---

3. Pre-IPO context

3a. Valuation / tender ladder

Grading: A = filing/regulatory/exchange-quote sourced; B = major-outlet reporting; C = analyst estimate. Valuation roughly doubled every 5–6 months through 2025 ($350B→$400B→$800B), then re-rated to $1.75T at IPO — a 13x increase over four years from the Jan 2023 $137B primary mark, with the last 18 months alone (Dec 2024→Jun 2026) accounting for a 5x move, coincident with the AI-segment (xAI) narrative entering the story.

3b. Primary funding history (pre-2023)

Cumulative primary funding through Jan 2023: ~$12B (SpaceXStock.com compilation). No primary capital was raised between Jan 2023 and the June 2026 IPO — every intervening valuation mark was a non-dilutive secondary/tender event (grade B/C).

xAI merger (structural, not a capital raise): all-stock combination closed 2026-02-02/03, implied combined valuation $1.25T (SpaceX $1T + xAI $250B) — CNBC, 2026-02-03 (grade B). This is the origin of the S-1's three-segment (Space/Connectivity/AI) structure.

3c. Regulatory record (FCC/FAA)

Recurring FAA mishap-investigation cadence is a structural feature of the Starship test program under the Starbase, TX launch license — not a one-off.

3d. Government contract record

Individual award lines (grade A/B) do not fully reconcile to the $22B company-cited aggregate (grade C) without a primary USASpending.gov pull — treat the $22B as directional.

---

4. Takeaways for the model — anchor numbers

• Share price / market cap: last close $136.12 (2026-07-14, Nasdaq); IPO priced $135/share (2026-06-11); implied shares outstanding ≈13.0B (back-solved, unconfirmed against 424B4 cover page — flagged in §2c).
• FY2025 actuals (S-1): revenue $18.67B (+33% YoY); net loss -$4.9B; adjusted EBITDA +$6.58B; capex $20.7B (capex/revenue 111%).
• FY2024 actuals (S-1): revenue $14.0B (+35% YoY); net income +$791M (only profitable year on record); capex $11.2B.
• FY2023 actuals (S-1): revenue $10.4B; net loss -$4.6B; capex $4.4B.
• Q1 2026 actuals: revenue $4.694B; operating loss -$1.943B; adjusted EBITDA +$1.127B.
• Segment mix (FY2025): Connectivity 69% of revenue / +$4.4B op. profit (only profitable segment); Space ~22% / loss-making; AI ~9% / -$6.355B op. loss (dominant swing factor).
• Starlink KPIs: 12M+ subscribers (2026-06-04, 160+ countries); ARPU $66/mo (Q1 2026), down from $99/mo (2023), -33%.
• Launch cadence: 165 launches in 2025 (record year); Falcon 9 dedicated price $74M (late-Feb 2026).
• Capital structure: last primary round pre-IPO Jan 2023 at $137B; IPO raised ~$75B new primary/secondary capital combined at $1.75T pricing basis.
• Structural driver: the 2026-02-02/03 SpaceX-xAI merger ($1.25T combined) is why the company now reports an AI segment at all — model this as a distinct, currently loss-making growth segment, not folded into "Space."

---

5. What we will not know until the first 10-Q/10-K

• Audited financial statements. Everything in §2b is S-1-disclosure-vintage (unaudited or audited-for-IPO-purposes only); the first 10-Q (quarter ended 2026-06-30, expected ~August 2026) will be the first GAAP periodic filing subject to ongoing SEC review, and the first 10-K will bring a full-year audited FY2026 statement.
• Exact post-IPO share count and vote structure. §2c/§2e are back-solved/directional; the 424B4 cover page and "Description of Capital Stock" section need a direct pull to confirm fully-diluted share count and any founder super-voting mechanics.
• Capex-by-segment split. No clean public breakdown of capex between Starlink satellite production, Starbase/Starship tooling, and AI compute infrastructure exists yet in sources reviewed; a 10-Q/10-K footnote (or a direct S-1 exhibit pull) should resolve this.
• Anthropic Cloud Services Agreement details. The ~$1.25B/month, through-May-2029 compute agreement referenced in secondary S-1 commentary is unverified against primary text or Anthropic's own disclosures — needs confirmation before it's treated as a revenue-line modeling input.
• Segment reporting stability. Whether the three-segment (Space/Connectivity/AI) structure persists, or gets further subdivided (e.g., X/Grok separated from AI/compute), is unknown until at least one full quarterly cycle of post-IPO segment reporting is observed.
• Underwriters' over-allotment exercise / final share count. The 2026-06-23 8-K (Items 8.01/9.01) is plausibly related but not confirmed verbatim; final net primary proceeds and share count should be confirmed in the first 10-Q.
• Full risk-factor text. §2f is a directional synthesis of secondary S-1 coverage, not a verbatim risk-factor extraction; a direct primary-text pull of the S-1 "Risk Factors" section is still owed before coverage.md cites specific risk-factor language.
• Guidance/outlook. SpaceX has not yet held a public earnings call or issued formal guidance; the first 10-Q filing and accompanying investor materials will be the first opportunity for management commentary in a reporting-company context.

---

Sourcing note: per PROCESS.md §5, every specific figure above carries an inline source and date on first use. Figures traced to secondary compilations of the S-1 (rather than a verbatim primary-document pull) are flagged as such throughout; closing that gap — a direct EDGAR pull of the S-1/424B4 primary text — is the top-priority follow-up before the model-build stage locks its actuals column.