IA dispara em 12 minutos para preparar o prospecto da SpaceX, Wall Street vai mudar

AI agent autonomously completes SpaceX S-1 analysis, producing an investment memo in just 12 minutes for only $1.87, signaling a redefinition of Wall Street's work model.
(Background summary: Earth’s power grid can't keep up with AI? Analyzing SpaceX’s acquisition of xAI to build a "Space Computing Empire")
(Additional context: Elon Musk says space AI data centers "will happen even on knee-jerk thoughts," while SpaceX’s IPO warns it may not proceed as planned)

Table of Contents

Toggle

• Decision Card (Conclusion = Hold and Wait)
• Multiple Arguments
• Opposing Arguments
• Investment Thesis
• SpaceX IPO: AI 12-minute read of the prospectus
• Counterarguments
• Starlink Surge: 10.3 million subscribers as a moat
• Company Overview
• X (social platform) is a business unit, not a footnote
• Market Position — Real-time Comparable Data
• Debt Crisis: $42 billion bridge loan due
• Pending Major Transactions and Contingent Liabilities
• Valuation
• Position Size Ladder
• Major Risks (Severity × Likelihood)
• Underwriter Conflict of Interest
• Related Party Density
• Litigation Collapse: Grok class-action reserve of $530 million
• Decision Triggers
• First 180 days plus multi-year watchlist
• X Platform Integration: AI division merged into Twitter assets
• Sources

When an AI agent has the capability to autonomously pay for data and follow decision analysis pathways, the traditional Wall Street investment analysis model faces a fundamental transformation. An AI agent, after SpaceX files its S-1, reads 226MB of the prospectus in 12 minutes, purchases real-time market data via on-chain USDC on Base, and produces an investment committee memo covering multiple arguments, valuation models, and risk matrices—all at a cost of just $1.87.
This is not a demo; it’s a real paid API call record, marking that automation tools are reshaping the work of financial analysis industries.

An AI agent autonomously completed a task that would take an investment analyst team days: reading 226MB of SpaceX S-1, buying real-time market data on Base with USDC, and producing an investment memo with multiple arguments, valuation models, and risk matrices—all for only $1.87.
This is not a demo; it’s a real paid API call record. When AI agents can pay for data themselves and determine their own analysis pathways, Wall Street’s work methods are being reconstructed.

An AI agent finished reading the 226MB SpaceX S-1 filed on Monday, purchased real-time market data on Base via USDC, and generated this investment committee memo within 12 minutes.
Total cost: 6 paid API calls, $1.87 USDC, no API key needed.

Decision Card (Conclusion = Hold and Wait)

Multiple Arguments

SpaceX possesses three business aspects that competitors cannot replicate. First, its near-monopoly status in commercial space access—accounting for 80% of global orbital payloads since 2023, Falcon’s success rate over 99%, and its reusable technology leading by a decade. Second, the world’s only deployed low-earth orbit broadband network—Starlink with 10.3 million subscribers in 164 countries, growing 49.8% YoY, with segment-adjusted EBITDA reaching $7.2 billion. Third, since acquiring xAI in February 2026, it has become the only vertically integrated AI laboratory at the rocket launch level, deploying orbital computing capacity in the future.
Regardless of valuation method, this is a generational asset.

Opposing Arguments

Connectivity business is real and profitable. But everything else either burns money at an astonishing rate—AI division with $3.2 billion revenue in 2025 lost $6.4 billion—or bets on Starship, which has achieved 11 test flights but has yet to send payloads into orbit. This IPO is partly a refinancing event. SpaceX borrowed $20 billion bridge loan for the xAI acquisition, due September 2027, with the underwriters of this IPO as the lenders.
If valuation exceeds $500 billion, you’re paying for unrealized capabilities, with no say in governance, and underwriters needing to succeed in refinancing.

Investment Thesis

Starlink is an excellent standalone business. 2025 revenue of $11.4 billion (+49.8%), operating income of $4.4 billion (+120%), segment-adjusted EBITDA of $7.2 billion (+86%). High-value subscriptions, 10.3 million paying users.

SpaceX IPO: AI 12-minute read of the prospectus

Launch business is unique. Since 2023, accounting for over 80% of global orbital payloads, Falcon’s success rate over 99%, with Falcon 9 first stages flying up to 34 times.

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

Government reliance is a moat, not a risk. The top US launch provider for national security: 11 out of 12 national security space launches in 2025, all 5 NASA crewed and cargo flights.

Orbital AI computing options planned for 2028 deployment. If Starship reaches even 50% of its economic target—reducing launch costs by 99%—the market size will expand by an order of magnitude.

Counterarguments

AI division burns over $6 billion annually. In 2025: $3.2 billion revenue, $6.4 billion operating loss, segment-adjusted EBITDA negative $1.2 billion, capital expenditure $12.7 billion.
Q1 2026: revenue $818 million, operating loss $2.5 billion, capex $770 million.
Annualized AI capex exceeds $30 billion, while AI revenue is only $3.2 billion.

Real debt is about $42 billion, not the headline $29 billion. Composition: roughly $20 billion SpaceX bridge loan (due September 2027), about $6.7 billion X B-1 term loan, about $6 billion X B-3 term loan (both due October 2029, interest 10-12%), plus ~$9.1 billion “other financing,” including obligations from failed sale-leaseback of AI infrastructure.
Loans related to X alone generate about $1.2–1.3 billion annual interest.

$1.96 billion EchoStar spectrum commitment to be completed by November 2027, in exchange for 65MHz US spectrum and global mobile satellite licenses.
This is a binding capital commitment outside the bridge loan and FY2026 capex.

The Cursor option agreement could trigger up to $10 billion termination fee. In April 2026—one month before this S-1—SpaceX signed a deal with Anysphere (Cursor) on compute and options, implying Cursor’s valuation at $60 billion.
If either party terminates, SpaceX must pay $1.5 billion termination fee plus $8.5 billion deferred service fee, payable in cash or Class A stock.

$4.5 billion contract with Anthropic is the largest external revenue source for AI division. Signed in May 2026, it commits Anthropic to pay $1.25 billion monthly until May 2029.
SpaceX is selling its COLOSSUS compute power to frontier model companies, creating extreme counterparty concentration risk.

On the balance sheet, a $530 million litigation reserve is recorded for Grok image output class action—Jane Doe v. X.AI (Jan 2026), Jane Doe 1 (Mar), and Baltimore cases (Mar). Plaintiffs seek compensatory, statutory, and punitive damages. S-1 states additional losses are unquantifiable.

Q1 2026 revenue growth slowed to 15.4% (from $4.69B to $4.07B YoY), below 2025’s 33.2%.

Starlink Surge: 10.3 million subscribers as a moat

SpaceX will be a controlled company with four classes of equity. Musk holds majority voting rights post-IPO. The company relies on Nasdaq’s controlled company exemption, waiving independent compensation and nominating committees.

Adjusted EBITDA is approximately $9 billion. Management’s 2025 headline figure is $6.6 billion “adjusted EBITDA,” while GAAP operating loss is -$2.6 billion. Adjustments exclude depreciation, stock-based incentives, and segment-specific exclusions.

Company Overview

SpaceX (SEC CIK 0001181412) designs and operates reusable rockets, the world’s largest LEO satellite constellation (~9,600 broadband satellites plus ~650 direct-to-phone satellites), and—after acquiring xAI in February 2026—gigabit-scale AI training infrastructure. Three 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 clusters). 2025 revenue: $18.7 billion; GAAP operating loss: -$2.6 billion; cash on hand: $15.85 billion versus $29.1 billion long-term debt on the capitalization sheet.

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

The corporate chain warrants re-tracing. SpaceX acquired xAI in February 2026. xAI bought X Holdings in March 2025. X Holdings acquired Twitter in October 2022.
Result: Twitter/X now integrated into SpaceX’s AI division, with its own balance sheet projects, lawsuits, and debt structure.

Scale. Over the past 12 months, supports 1.3 billion accounts, 550 million monthly active users (up from 520 million in Dec 2025), with 350 million daily posts.
Among these monthly active users, 117 million use Grok features—X is the main distribution channel for that model.
Money products (payments, banking, financial services) launched beta in Nov 2025 and are progressing toward full rollout.
X Ads Manager phased launch began in April 2026.

Financial contribution. AI division in 2023-2024 derives nearly all revenue from X—ads, X Premium subscriptions, and data licensing.
In 2024 alone, ad revenue declined by $595 million YoY due to “X losing ad partners,” offset by $157 million increase in X Premium subscriptions and $90 million in data licensing.

Adding $20 billion SpaceX bridge loan (due September 2027) and $9.1 billion “other financing,” total long-term debt is about $42 billion—more than the headline $29 billion on the cap sheet.

X-specific risks not present in other SpaceX businesses.
EU Digital Services Act enforcement on super-large online platforms.
Reversible brand safety issues on short-term ad contracts that can be canceled at any time—2024’s mass exodus could recur within a single news cycle.
Money products trigger payments/money transfer/banking regulation across all 50 US states and every foreign jurisdiction.
Reversal of content moderation policies could simultaneously trigger advertiser suspensions and user migration.

Market Position — Real-time Comparable Data

This comparison table was assembled on the fly during analysis, using a $0.10 payment to Jintel’s GraphQL endpoint to fetch basic fundamentals for five comparables. No Bloomberg terminal, no FactSet contract needed.

Debt Crisis: $42 billion bridge loan due

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

Interpreting the comparison group. Rocket Lab’s 104x EV/Sales is the closest narrative benchmark—investors are willing to pay high multiples for scaled reusable launches and low-earth orbit options, even with negative margins.
SpaceX should command a higher multiple than RKLB, but applying 104x to SpaceX’s $11.4 billion revenue implies a $1.2 trillion equity value, which is unanchored.
AST SpaceMobile’s 345x is purely a revenue-based narrative valuation, serving as an upper bound for direct-to-phone options.
Iridium’s 7.4x sales and 14.8x EBITDA reflect mature, profitable LEO communications—applying 7.4x to Starlink’s $11.4 billion yields an $840 billion standalone value (opponent anchor).
NVIDIA’s 31.7x EV/EBITDA with 85% revenue growth indicates the level of growth needed for the AI division to justify valuation based on fundamentals.
It’s not there yet.

Notable signals. Rocket Lab filed a 424B5 supplement on May 20, 2026—the same day SpaceX released its S-1. RKLB’s secondary issuance during SpaceX news cycle suggests management sees IPO windows open and competitive supply imminent.

Pending Major Transactions and Contingent Liabilities

These four items are individually significant and overlapping. Two were signed within 60 days before this S-1.

Why are these important for valuation?
A clear “adjusted net obligation” perspective: $42 billion total debt + $19.6 billion EchoStar commitments + up to $10 billion Cursor contingent liabilities, minus $15.85 billion cash, equals about $55 billion net obligation—excluding IPO proceeds.
This is three to four times the cap sheet’s simple number, substantially altering the opponent scenario.

Valuation

Method 1—based on independent transaction multiples of the Connectivity segment, as it is the only segment with positive independent economics.

Position Size Ladder

Major Risks (Severity × Likelihood)

Underwriter Conflict of Interest

This is buried in the underwriting section, rarely discussed but critical.
Five lead underwriters (Goldman Sachs, Morgan Stanley, BofA, Citi, JPM) plus five additional book managers (Barclays, Deutsche Bank, Royal Bank of Canada, UBS, WFC) are creditors of the ~$20 billion SpaceX bridge loan, now involved in IPO pricing for refinancing.
Morgan Stanley also advised SpaceX on the xAI acquisition (funded by bridge loan).
Underwriters have a direct financial interest in maximizing IPO proceeds.
This should alert the investment committee to pricing discipline.

Related Party Density

Litigation Collapse: Grok class-action reserve of $530 million

No single item appears alarming alone.
What’s concerning is the density—multiple financial contacts between Musk-controlled entities and SpaceX, at least nine different points.
Typically, governance reviews examine one or two such relationships; here, it’s an order of magnitude more.

Decision Triggers

If the deal’s implied equity valuation is $350 billion or less, and Starship achieves commercial payload delivery as guided in H2 2026, and Q2 2026 connectivity revenue growth exceeds 40% YoY, then upgrade to overweight.

If valuation exceeds $510 billion, or if Starship suffers a vehicle loss delaying V3 satellite deployment beyond 2027, or if the AI division’s burn rate accelerates to an annualized operating loss over $8 billion in Q2-Q3 2026, or FAA imposes long-term Starship flight bans, then downgrade to abandon.

First 180 Days plus Multi-year Watchlist

D+1: First-day gain benchmark compared to comparable IPOs

D+30: First quarterly report (Q2 2026)—triggering early lock-up release (immediately release 20%, then another 10% if stock exceeds +30% of issue price)

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

D+90: silence period ends, sell-side analysts initiate coverage

D+180: all standard lock-up periods expire

Second half 2026: Starship achieves commercial payload delivery

X Platform Integration: AI division merged into Twitter assets

Q2-Q3 2026: Grok image output class action milestone (monitor if reserve increases to $530 million)

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

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

November 2027: $19.6 billion EchoStar spectrum deal closed—V2 mobile global launch constrained by this

May 2029: $45 billion Anthropic compute contract ends; renewal terms will define future AI division economics

October 2029: $12.7 billion X B-1 and B-3 term loans mature

Sources

SpaceX S-1, SEC Registration No. 0001628280-26-036936, filed 2026-05-20

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

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

Industry IPO background via Parallel Search, on Base chain, retrieved 2026-05-22

Four scenarios for SpaceX IPO—Acadian Asset Management

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

A Bloomberg terminal costs $24,000 per year.
This memo demonstrates what AI can produce when it can pay for data itself.