SpaceX IPO: an institutional IC memo, produced by an agent for $1.87

An AI agent read the 226 MB SpaceX S-1 filed Monday, paid for live market data with USDC on Base, and produced this investment committee memo in 12 minutes. Total cost: 6 paid API calls, $1.87 USDC, no API keys.

Decision card (verdict = HOLD)

The bull case (steel-manned)

SpaceX owns three businesses no competitor can replicate. A near-monopoly on commercial space access — 80% of global mass-to-orbit since 2023, a 99% Falcon mission success rate, and a 10-year reusability lead. The only globally-deployed low-Earth-orbit broadband network — 10.3 million Starlink subscribers across 164 countries, growing 49.8% year-over-year with $7.2B of segment-adjusted EBITDA. And, since the February 2026 xAI acquisition, the only AI lab vertically integrated all the way down to the launch vehicle that will deploy orbital compute. At any reasonable valuation, this is a generational asset.

The bear case (also steel-manned)

The Connectivity segment is real and profitable. Everything else is either burning cash at extraordinary rates — the AI segment lost $6.4B on $3.2B of revenue in 2025 — or contingent on Starship, which has completed 11 flight tests but is not yet delivering payload to orbit. The IPO is partly a refinancing event. SpaceX took on a $20B bridge loan due September 2027 to fund the xAI acquisition, and the bridge lenders are exactly the same banks pricing the IPO. At any valuation above $500B, you are paying for execution that has not yet happened, governance you have no say in, and a refinancing the underwriters need to succeed.

Investment thesis

• Starlink is a great standalone business. 2025 revenue $11.4B (+49.8%), operating income $4.4B (+120%), segment-adjusted EBITDA $7.2B (+86%). Premium-priced subscription service with 10.3M paid users.
• Launch is a category of one. 80%+ of global mass-to-orbit since 2023, a 99%+ Falcon success rate, and Falcon 9 first stages flown up to 34 times each.
• Vertical integration is real and compounding. Rockets → satellites → spectrum (the EchoStar AWS-4 / H-block deal is now FCC-approved) → AI compute (~1 GW across the two COLOSSUS clusters).
• Government dependency is a moat, not just a risk. Primary launcher for U.S. national security: 11 of 12 National Security Space Launch missions in 2025, all five NASA crew and cargo flights.
• Optionality on orbital AI compute, planned for 2028. If Starship hits even 50% of stated economics — a 99% reduction in launch cost — the addressable market expands an order of magnitude.

Anti-thesis

The AI segment is a $6B+ annual cash incinerator. 2025: $3.2B revenue against a $6.4B operating loss, a negative $1.2B segment-adjusted EBITDA, and $12.7B of capex. Q1 2026 alone: $2.5B operating loss on $818M of revenue, $7.7B of capex. Annualized AI capex now exceeds $30B against $3.2B of AI revenue.

The true debt stack is roughly $42B, not the $29B headline. Composition: ~$20B SpaceX bridge loan due Sept 2027, ~$6.7B X-Corp B-1 Term Loan and ~$6.0B X-Corp B-3 Term Loan (both maturing Oct 2029, with 10-12% effective interest rates), and ~$9.1B of "Other Financings" that include failed-sale-leaseback obligations on AI infrastructure. The X-specific tranches alone produce roughly $1.2-1.3B of annual interest expense flowing through the AI segment.

A $19.6B EchoStar spectrum commitment is closing in November 2027. Equity-plus-cash consideration for 65 MHz of U.S. spectrum and global Mobile Satellite Service licenses. This is a binding capital commitment outside both the bridge loan and the FY26 capex run-rate.

A Cursor option agreement could trigger up to $10B in termination fees. SpaceX signed a compute-and-option agreement with Anysphere (Cursor) in April 2026 — one month before this S-1 — at a $60B implied equity valuation for Cursor. If either side terminates, SpaceX owes Cursor a $1.5B termination fee plus an $8.5B deferred services fee, payable in cash or Class A stock.

A $45B Anthropic contract is the AI segment's largest single external revenue stream. Cloud Services Agreements signed May 2026 commit Anthropic to $1.25B per month through May 2029. SpaceX is selling its COLOSSUS compute capacity to a direct frontier-model competitor, creating extreme counterparty concentration.

A recognized $530M litigation accrual sits on the balance sheet for Grok image-generation class actions — Jane Doe v. X.AI Corp (Jan 2026), Jane Doe 1 (Mar 2026), and the Baltimore Case (Mar 2026). Plaintiffs seek compensatory, statutory, and punitive damages. The S-1 explicitly states the range of additional loss is not estimable.

Q1 2026 revenue growth decelerated to 15.4% ($4.69B versus $4.07B year-over-year), down from 33.2% for full-year 2025.

SpaceX will be a controlled company with four share classes. Musk holds majority voting power post-IPO. The company will rely on Nasdaq's controlled-company exemptions, which waive requirements for an independent compensation committee and an independent nominating committee.

Adjusted EBITDA flatters the picture by approximately $9B. Management's headline number for 2025 is $6.6B of "Adjusted EBITDA" against a GAAP operating loss of $(2.6)B. The reconciliation strips out depreciation, stock-based compensation, and segment-specific exclusions.

Company snapshot

SpaceX (Space Exploration Technologies Corp; SEC CIK 0001181412) designs and operates reusable rockets, the world's largest LEO satellite constellation (~9,600 broadband satellites plus ~650 direct-to-mobile satellites), and — following its February 2026 acquisition of xAI — gigawatt-scale AI training infrastructure. Three reportable segments: Space, Connectivity (10.3M Starlink subscribers), and AI (the Grok model, the X social platform with 550M monthly active users, and the COLOSSUS / COLOSSUS II compute clusters). 2025 revenue was $18.7B; GAAP operating loss was $(2.6)B; cash on hand was $15.85B against $29.1B of long-term debt as reported on the cover-page capitalization table.

X (the social platform) is a business unit, not a footnote

The corporate chain is worth re-tracing. SpaceX acquired xAI in February 2026. xAI had acquired X Holdings in March 2025. X Holdings had acquired Twitter in October 2022. The result: Twitter/X is now consolidated inside SpaceX's AI segment, with its own balance-sheet items, its own litigation, and its own debt stack.

Scale. 1.3 billion supported accounts in the last twelve months, 550 million monthly active users (up from 520M in December 2025), and 350 million daily posts. 117M of those MAUs use Grok features — X is the primary distribution channel for the model. The Money product (payments, banking, financial services) launched in beta in November 2025 and is rolling out to general availability. X Ads Manager began a phased rollout in April 2026.

Financial contribution. Substantially all of the AI segment's 2023-2024 revenue was X-derived — advertising, X Premium subscriptions, and data licensing. In 2024 alone, advertising revenue dropped by $595M year-over-year due to "loss of advertising partners for X," partially offset by +$157M in X Premium subscription revenue and +$90M in data licensing.

Combined with the $20B SpaceX Bridge Loan (Sept 2027) and the $9.1B "Other Financings" line, gross long-term debt is roughly $42B — not the $29B headline on the capitalization cover.

X-specific risks not present elsewhere in SpaceX. EU Digital Services Act enforcement on a Very Large Online Platform. Advertiser brand-safety reversibility on short-term cancellable ad contracts — the 2024 exodus could repeat on a single news cycle. Money Product triggers payments / money transmitter / banking regulation across all 50 states and every foreign jurisdiction. Content-moderation policy reversals can trigger advertiser pause and user migration simultaneously.

Market position — live comparable data

This comp table was assembled live during the analysis by paying $0.10 to Jintel's GraphQL endpoint for batched fundamentals across all five comparables. No Bloomberg seat, no FactSet contract.

ASTS operating margin reflects pre-revenue mass investment. Source: Jintel entitiesByTickers via x402 on Base, retrieved 2026-05-22.

Reading the comp set. Rocket Lab at 104x price-to-sales is the closest narrative comp — investors pay nosebleed multiples for at-scale reusable launch plus low-Earth-orbit optionality, even with negative margins. SpaceX arguably deserves a higher multiple than RKLB, but blindly applying 104x to SpaceX's $11.4B Connectivity revenue alone implies $1.2T of equity value, which is not anchorable to anything. AST SpaceMobile at 345x is pure narrative pre-revenue valuation, useful only as a ceiling for the direct-to-cell optionality. Iridium at 7.4x sales and 14.8x EBITDA is what mature profitable LEO communications looks like — apply 7.4x to Starlink's $11.4B and you get $84B for the Starlink business standalone (a bear case anchor). NVIDIA at 31.7x EV/EBITDA on 85% revenue growth is what the AI segment would need to grow into to deserve a fundamentals-based valuation. It's not there yet.

A signal worth noting. Rocket Lab filed a 424B5 prospectus supplement on May 20, 2026 — the exact same day SpaceX dropped its S-1. RKLB raising secondary equity into the SpaceX news cycle suggests management views the IPO window as wide open and competitive supply pressure as imminent.

Pending material transactions and contingent obligations

These four items are each individually material and they compound. Two were signed within the 60 days before this S-1 was filed.

Why this matters for valuation. A clean "adjusted net obligations" view is: $42B gross debt plus the $19.6B EchoStar commitment plus up to $10B of Cursor contingent liability, minus $15.85B of cash on hand, equals roughly $55B of net obligations before any IPO proceeds. That is three-to-four times the figure produced by a naive read of the capitalization cover page and meaningfully shifts the bear case.

Valuation

Method 1 — Trading multiples on the Connectivity segment standalone, since it is the only segment with positive standalone economics.

Position-sizing ladder

Top risks (severity × likelihood)

Underwriter conflicts of interest

This is buried deep in the Underwriting section, rarely covered in the press, and material. The five lead representatives (Goldman Sachs, Morgan Stanley, BofA, Citigroup, J.P. Morgan) plus five additional book-runners (Barclays, Deutsche Bank, RBC, UBS, Wells Fargo) all have affiliates that are lenders under the $20B SpaceX Bridge Loan they are now pricing the IPO to refinance. Morgan Stanley separately advised SpaceX on the xAI acquisition that the bridge loan funded. The syndicate has a direct financial interest in maximizing IPO proceeds. This should make pricing discipline at the IC table.

Related-party density

No single line is alarming. The density is the concern — the Musk-controlled entity graph has at least nine distinct financial touchpoints with SpaceX. Public-company governance committees typically scrutinize one or two such relationships. This is an order of magnitude more.

Decision triggers

Upgrade to overweight if the deal prices at or below $350B implied equity, AND Starship achieves commercial payload delivery in H2 2026 as guided, AND Q2 2026 Connectivity revenue growth exceeds 40% year-over-year.

Downgrade to PASS if the deal prices above $510B, OR a Starship loss-of-vehicle event delays V3 satellite deployment beyond 2027, OR the AI segment burn accelerates above $8B annualized operating loss in Q2-Q3 2026, OR the FAA imposes an extended Starship grounding.

First 180 days plus multi-year watch list

• D+1: first-day-pop benchmark vs. comparable IPOs
• D+30: first quarterly earnings (Q2 2026) — drives the early-release lockup tier (20% release immediately, +10% if stock is +30% vs. offer)
• D+70, +90, +105, +120, +135: staged early-release lockup tiers of 7% each
• D+90: quiet period ends, sell-side analyst initiations begin
• D+180: full standard-tier lockup expires
• H2 2026: Starship guided to commercial payload delivery
• Q2-Q3 2026: Grok image-generation class-action procedural milestones (watch for any increase in the $530M accrual)
• April 2027: first anniversary of the Cursor option agreement — watch for exercise or termination signal
• September 2027: $20B SpaceX Bridge Loan matures (must be refinanced or repaid)
• November 2027: $19.6B EchoStar spectrum close — V2 Mobile global rollout is gated on this
• May 2029: $45B Anthropic compute contract ends; renewal terms will define AI segment economics for years afterward
• October 2029: combined $12.7B X-Corp B-1 and B-3 Term Loans mature

Sources

• SpaceX S-1, SEC Accession 0001628280-26-036936, filed 2026-05-20: https://www.sec.gov/Archives/edgar/data/1181412/000162828026036936/spaceexplorationtechnologi.htm
• Live comparable fundamentals via Jintel GraphQL entitiesByTickers, x402 on Base, retrieved 2026-05-22
• Live SEC composite profiles via x402helper /companies/profile for RKLB, IRDM, VSAT, retrieved 2026-05-22
• Sector IPO context via Parallel Search, x402 on Base, retrieved 2026-05-22
• Four scenarios for the SpaceX IPO — Acadian Asset Management: https://www.acadian-asset.com/investment-insights/owenomics/four-scenarios-for-the-spacex-ipo

Produced by the IPO Analysis bundle on agentic.market. 6 paid x402 calls. $1.87 USDC on Base. No API keys. No signups. Pay-per-request.

A Bloomberg terminal seat is $24,000 per year. This memo is what agents can produce now when they can pay for the data themselves.