An AI reads the SpaceX IPO prospectus and takes 12 minutes to write this investment memo.

Author: Nick Prince

Compiled by: Deep Tide TechFlow

Deep Tide Briefing: An AI agent autonomously completed work that would take an investment analyst team days: reading through the 226MB SpaceX S-1 file, using USDC on the Base chain to purchase real-time market data, and generating an investment committee memo that includes multi-party arguments, valuation models, and a risk matrix—at a total cost of only $1.87. This is not a demo; it is a real record of paid API calls. When AI agents can pay for data themselves and decide their own analysis paths, the way Wall Street works is being reshaped.

An AI agent read the 226MB SpaceX S-1 file submitted on Monday, bought real-time market data on the Base chain with USDC, and generated this investment committee memo within 12 minutes. Total cost: 6 paid API calls, $1.87 USDC, no API key required.

Decision Card (Conclusion = Hold and Wait)

Multi-party Argumentation

SpaceX has three lines of business that competitors cannot replicate. First is its near-monopoly position in commercial space access—since 2023 it has accounted for 80% of global orbital launch quality, a Falcon mission success rate of 99%, and reusable technology that leads by 10 years. Second is the only globally deployed low-earth-orbit broadband network—Starlink, which has 10.3 million subscribers in 164 countries, a 49.8% year-over-year increase, and $7.2 billion of adjusted EBITDA after segment adjustments. Third is that, since acquiring xAI in February 2026, it has become the only AI lab vertically integrated into the rocket-launch layer, and it will deploy orbital computing capabilities in the future. No matter what reasonable valuation method you use, this is a generational asset.

Bear Argumentation

Connectivity is real and profitable. But everything else either burns money at an astonishing pace—AI division had $3.2 billion in revenue in 2025 and an operating loss of $6.4 billion—or bets on Starship, which has completed 11 flight tests but has not yet delivered payloads into orbit. This IPO is partly a refinancing event. SpaceX borrowed a $20 billion bridge loan to acquire xAI, due in September 2027, and the bridge lenders are the underwriters of this IPO. If the valuation exceeds $500 billion, you are paying for execution capability that has not yet been realized, corporate governance that you have no right to weigh in on, and a refinancing transaction that the underwriters must successfully complete.

Investment Thesis

Starlink is an excellent standalone business. 2025 revenue was $11.4 billion (+49.8%), operating income was $4.4 billion (+120%), and adjusted EBITDA was $7.2 billion (+86%). High-priced subscription services, 10.3 million paying users.

The launch business is unique. Since 2023 it has accounted for more than 80% of global orbital launch quality, Falcon success rate exceeds 99%, and Falcon 9’s first-stage rocket can fly up to 34 times.

Vertical integration is real and creates compounding effects. Rocket → satellite → spectrum (the EchoStar AWS-4/H-band deal has been approved by the FCC) → AI computing power (two COLOSSUS clusters of about 1GW).

Dependence on the government is a moat, not a risk. The #1 U.S. national-security launch provider: in 2025 it executed 11 out of 12 national-security space launch missions, and all 5 NASA crewed and cargo flights.

The option value of orbital AI computing power, planned for deployment in 2028. If Starship achieves even 50% of its stated economics—cutting launch costs by 99%—the addressable market will expand by an order of magnitude.

Counterarguments

The AI division is a bottomless pit burning over $6 billion annually. In 2025: $3.2 billion revenue versus $6.4 billion operating loss, adjusted EBITDA of negative $1.2 billion, and $12.7 billion of capital expenditures. Only in 2026 Q1 alone: $818 million revenue versus $2.5 billion operating loss, and $7.7 billion of capital expenditures. Annualized AI capital expenditure has already exceeded $30 billion, while AI revenue is only $3.2 billion.

The actual debt size is about $42 billion, not the headline $29 billion figure. Composition: about $20 billion in the SpaceX bridge loan due in September 2027, about $6.7 billion in X Company B-1 term loans and about $6.0 billion in X Company B-3 term loans (both due in October 2029, with effective interest rates of 10–12%), and about $9.1 billion of “other financing,” including obligations arising from failed sale-leaseback of AI infrastructure. Interest expense of about $1.2–$1.3 billion per year is generated by the X-related loans alone, and is included in the AI division.

The $1.86 billion EchoStar spectrum commitment is scheduled to be completed in November 2027. It is an equity-plus-cash consideration exchange for 65MHz of U.S. spectrum and global mobile satellite service licenses. This is a binding capital commitment beyond the bridge loan and FY2026 capital expenditures.

The option agreement with Cursor could trigger termination fees as high as $10 billion. In April 2026—one month before this S-1 filing—SpaceX signed a compute-and-option agreement with Anysphere (Cursor), implying a Cursor valuation of $60 billion. If either party terminates, SpaceX must pay Cursor a $1.5 billion termination fee plus an $8.5 billion deferred service fee, payable in cash or Class A stock.

The $4.5 billion Anthropic contract is the largest single external revenue source for the AI division. A cloud services agreement signed in May 2026 requires Anthropic to pay $1.25 billion per month through May 2029. SpaceX is selling its COLOSSUS computing power to directly competitive frontier-model companies, creating extreme counterparty concentration risk.

On the balance sheet, a $53 million litigation reserve has been recorded for Grok image generation class-action lawsuits—Jane Doe v. X.AI (January 2026), Jane Doe 1 (March), and the Baltimore case (March). Plaintiffs are seeking compensatory, statutory, and punitive damages. The S-1 explicitly states that the scope of additional losses cannot be estimated.

In 2026 Q1, revenue growth slowed to 15.4% ($4.69 billion versus $4.07 billion year-over-year), below 33.2% for all of 2025.

SpaceX will be a controlled company with four classes of equity. After the IPO, Musk will hold majority voting rights. The company will rely on Nasdaq’s controlled company exemption, which waives the requirements for an independent compensation committee and an independent nominating committee.

Adjusted EBITDA inflates the figure by about $9 billion. Management’s 2025 headline figure is $6.6 billion “adjusted EBITDA,” while the GAAP operating loss is negative $2.6 billion. The adjustments exclude depreciation, equity incentive expenses, and segment-specific exclusions.

Company Profile

SpaceX (Space Exploration Technologies Corp.; SEC CIK 0001181412) designs and operates reusable rockets, the world’s largest LEO satellite constellation (about 9,600 broadband satellites plus about 650 direct-to-phone satellites), and—after acquiring xAI in February 2026—gigawatt-scale AI training infrastructure. Three reporting segments: Space, Connectivity (10.3 million Starlink subscribers), and AI (Grok models, X social platform with 550 million monthly active users, and COLOSSUS/COLOSSUS II computing clusters). 2025 revenue was $18.7 billion; GAAP operating loss was negative $2.6 billion; cash on hand was $15.85 billion versus $29.1 billion of long-term debt listed on the capitalization table.

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

The corporate chain is worth retracing. SpaceX acquired xAI in February 2026. xAI acquired X Holdings in March 2025. X Holdings acquired Twitter in October 2022. Result: Twitter/X is now consolidated into SpaceX’s AI segment, with its own balance sheet line items, its own lawsuits, and its own debt structure.

Scale. Over the past 12 months it has supported 1.3 billion accounts, 550 million monthly active users (higher than 520 million in December 2025), and 350 million posts per day. Of these monthly active users, 117 million use Grok functionality—X is the primary distribution channel for this model. Money products (payments, banking, financial services) launched a beta in November 2025 and are moving toward full availability. X Ads Manager began phased rollout in April 2026.

Financial contribution. The AI segment’s revenue in 2023–2024 came almost entirely from X—ads, X Premium subscriptions, and data licensing. In 2024 alone, ad revenue fell by $595 million year-over-year due to “X losing advertising partners,” partially offset by a $157 million increase in X Premium subscription revenue and a $90 million increase in data licensing.

Add the $20 billion SpaceX bridge loan (due September 2027) and the $9.1 billion “other financing” line item, and total long-term debt is approximately $42 billion—not the $29 billion headline figure on the capitalization cover.

X-specific risks that do not exist in SpaceX’s other businesses. Enforcement of the EU Digital Services Act on super-large online platforms. Reversibility of brand safety in short-term ad contracts that can be canceled at any time—2024’s mass exodus could be repeated within a single news cycle. Money products trigger payment/money transfer/banking regulation across all 50 U.S. states and every foreign jurisdiction. A reversal in content moderation policy could simultaneously trigger advertisers pausing and users migrating.

Market Position—Real-time Comparable Data

This comparison table was assembled in real time during the analysis process, by paying $0.10 to Jintel’s GraphQL endpoint to obtain batch fundamental data for all five comparable companies. No Bloomberg Terminal required, no FactSet contract required.

ASTS operating margin reflects large-scale investment before revenue. Source: retrieved from Jintel entitiesByTickers via Base chain’s x402, retrieval date 2026-05-22.

Interpretation of the comparison group. Rocket Lab’s 104x price-to-sales multiple is the closest narrative benchmark—investors are willing to pay very high multiples for scaled reusable launches with low-earth orbit option value, even if margins are negative. SpaceX should command a higher multiple than RKLB, but blindly applying 104x to SpaceX’s $1.14 billion revenue from only the connectivity business implies an equity value of $1.2 trillion, which cannot be anchored to anything. AST SpaceMobile’s 345x is a narrative valuation purely before revenue—only used as an upper bound reference for direct-to-phone option value. Iridium’s 7.4x sales and 14.8x EBITDA reflect what mature profitable LEO communications looks like—applying 7.4x to Starlink’s $1.14 billion yields an $84 billion standalone value for Starlink (bear anchor). NVIDIA’s 31.7x EV/EBITDA corresponds to 85% revenue growth—that is the level the AI segment needs to reach to justify a fundamental-based valuation. Not yet.

Notable signals. Rocket Lab filed a 424B5 prospectus supplement on May 20, 2026—exactly the same day SpaceX released its S-1. During SpaceX’s news cycle, RKLB issued secondary equity, indicating management believes the IPO window is open and that competitive supply pressure is imminent.

Pending Major Transactions and Contingent Obligations

These four items are individually significant and overlap with each other. Two were signed within 60 days before this S-1 was submitted.

Why this matters for valuation. A clear “adjusted net obligations” perspective is: $42 billion of total debt plus $19.6 billion of EchoStar commitments plus contingent liabilities of up to $10 billion for Cursor, minus $15.85 billion of cash on hand, equals approximately $55 billion of net obligations—before including any IPO proceeds. This number is three to four times the simple reading of the figure on the capitalization cover page, materially changing the bear case.

Valuation

Method 1—Based on the standalone trading multiple of the connectivity segment, because it is the only segment with positive standalone economic benefit.

Position Size Tiers

Major Risks (Severity × Likelihood)

Underwriter Conflict of Interest

This is deeply buried in the underwriting section and rarely covered by news, but it is important. The five lead underwriters (Goldman Sachs, Morgan Stanley, BofA, Citi, JPM) plus five additional bookrunners (Barclays, Deutsche Bank, Royal Bank of Canada, UBS, Wells Fargo) are all lenders on the $20 billion SpaceX bridge loan, and they are now involved in IPO pricing for refinancing that loan. Morgan Stanley also served as an advisor to SpaceX for the xAI acquisition (funded by the bridge loan). The syndicate has a direct financial incentive to maximize the IPO proceeds. This should keep the valuation committee vigilant about pricing discipline.

Related-party Density

No single item looks concerning on its own. What is concerning is the density—the network of entities controlled by Musk has at least nine different financial touchpoints with SpaceX. Corporate governance committees typically review one or two relationships like this; here there is an order of magnitude more.

Decision Triggers

If the transaction pricing implies an equity valuation of $350 billion or less, and Starship achieves commercial payload delivery in the second half of 2026 as guided, and if 2026 Q2 connectivity revenue grows by more than 40% year-over-year, then upgrade to Overweight.

If the transaction pricing exceeds $510 billion, or if Starship experiences a vehicle loss event causing V3 satellite deployment to be delayed beyond 2027, or if AI division burn accelerates in 2026 Q2–Q3 to an annualized operating loss of over $8 billion, or if the FAA imposes a long-term Starship flight ban, then downgrade to Abandon.

First 180 Days + Multi-Year Watchlist

D+1: First-day price increase benchmark versus comparable IPOs

D+30: First quarterly earnings report (2026 Q2)—trigger early release of the lock-up tier group (immediately release 20%; if the stock is +30% versus the offering price, release another 10%)

D+70, +90, +105, +120, +135: phased early releases of the lock-up tier group, 7% each

D+90: Silence period ends; sell-side analysts initiate coverage

D+180: All standard lock-up tiers expire

Second half of 2026: Starship achieves commercial payload delivery as guided

2026 Q2–Q3: Grok image generation class-action procedural milestone (watch whether the $530 million reserve increases)

April 2027: 1-year anniversary of the Cursor option agreement—watch for exercise or termination signals

September 2027: $20 billion SpaceX bridge loan matures (must refinance or repay)

November 2027: $19.6 billion EchoStar spectrum transaction completed—constraining the V2 global mobile rollout

May 2029: $45 billion Anthropic compute contract ends; renewal terms will define the AI segment’s economic benefits for the years after

October 2029: Maturity of a total of $12.7 billion in X Company B-1 and B-3 term loans

Sources

SpaceX S-1, SEC registration number 0001628280-26-036936, filed 2026-05-20

Real-time comparable fundamentals via Jintel GraphQL entitiesByTickers on Base chain’s x402, retrieval date 2026-05-22

Real-time SEC filings via x402helper /companies/profile for RKLB, IRDM, VSAT, retrieval date 2026-05-22

Industry IPO background via Parallel Search on Base chain’s x402, retrieval date 2026-05-22

Four scenarios for SpaceX IPO—Acadian Asset Management

Generated by the agentic.market IPO analysis package. 6 paid x402 calls. $1.87 USDC on Base chain. No API key required. No registration required. Pay per request.

A Bloomberg Terminal seat costs $24,000 per year. This memo demonstrates what agents can produce when they can pay for data themselves.

LAYOUT REFERENCE (source): total_lines=147, non_empty_lines=74, blank_lines=73