TON618 Capital Equity Research · Technology · 12 June 2026 ↓ Download PDF

Priced for Mars: the $1.77 trillion listing embeds the bull case, and then some.

We initiate on SPCX at AVOID / do-not-chase. Our base-case fair value is $66 against the $135 IPO and today's $160.95 first-day close (+19%, ~$2.1T market cap) — a 51% premium at the offer, ~59% above where we see value at the close. SpaceX is a genuinely exceptional franchise: a launch near-monopoly, a hyper-scaling Starlink, and a newly-merged frontier-AI arm. But at ~90× trailing revenue at the offer (~109× at the $160.95 close) and roughly $14 billion of annual cash burn, the listing prices in flawless execution across three moonshots simultaneously. The risk/reward is asymmetric to the downside. "Data shows" denotes fact; "we" denotes the Fund's view.

01Executive Summary

Recommendation: AVOID at IPO. Fair value $66 (base); bull case $128 still sits below the $135 print. We would revisit constructively below ~$55, or on evidence the AI segment can clear its cost of capital. The franchise is not the question; the price is.

Net cash of ~$79.5B (post-IPO cash ~$99.5B less ~$20B est. debt) added to segment EV in each scenario. Shares: 13.08B Class A+B. We reach $66 bottom-up; it is independent of any third-party target.

Investment thesis

• The numbers do not support the price. ~$18.7B FY25 revenue, a ~$2.6B operating loss, and ~$19.6B of investing outflow against ~$6.8B operating cash flow — roughly $14B of FCF burn. At ~90× revenue, SPCX is the most expensive mega-cap on any conventional metric.
• Starlink is the engine — and its price is falling. Connectivity revenue +49.8% to $11.4B with $7.2B segment adjusted EBITDA, but ARPU has slid from $99 to $66 as the mix turns consumer and operating income went near-flat quarter-on-quarter despite subscribers doubling. Volume is winning; price is not.
• The xAI merger is a financing event as much as a strategy. Bolted on all-stock in February 2026 (~$250B mark, ~3× its value seven months earlier); xAI lost $6.4B on ~$3.2B revenue and is the named first use of the $74.4B of proceeds. Optionality, not earnings.
• Control is absolute. Musk holds ~85% of votes on ~42% of equity via 10-vote Class B; SPCX is a controlled company, exempt from key Nasdaq governance rules, with mandatory-arbitration and class-action waivers. Key-person and governance risk are structural.
• Consensus is split, not bullish. Published targets span $63–$227 and average ~$164 — near the price, and a barbell of Buy and Sell initiations rather than a confident mean. We do not anchor to any of them; our $66 is built bottom-up.

Source: SpaceX Form S-1/A, filed 3 June 2026 (SEC accession 0001628280-26-040364). As-of dates in §17.

The Current State of the SpaceX Story

Strip away the mystique and SpaceX is now three businesses under one roof. One launches rockets and dominates that market. The second, Starlink, sells satellite internet to more than 10 million customers and actually makes money. The third, xAI, builds artificial intelligence — and it was bolted on only in February 2026, four months before this listing. Two of the three lose money; one, Starlink, is a genuine profit machine.

Here is the whole story in a sentence: a fast-growing, highly profitable internet business is being used to bankroll two enormous bets on the future — radically cheaper rockets (Starship) and frontier AI (xAI). That is a coherent, even thrilling, strategy. It is also why the company as a whole burns roughly $14 billion of cash a year despite Starlink's profits: those profits are plowed straight back into the moonshots.

We want to be precise about what we are and are not saying. We are not betting against SpaceX the company. The launch business is the best on Earth, and Starlink is a real, durable franchise. Our concern is the price. At the $135 offer the company was valued near $1.77 trillion; on day one it closed at $160.95 — about $2.1 trillion. That is roughly 90 to 110 times the last year's revenue, a level no company this size has ever sustained. You are not paying for what SpaceX is; you are paying for a near-flawless version of what it might become.

What is genuinely special — and what isn't. SpaceX is the only company that reliably lands and reuses its rockets, which is why it now flies the large majority of the world's launches; that is a real moat, not marketing. Starlink is the only satellite network of its size, with thousands of satellites relaying data to one another by laser — also hard to copy. The AI arm is the unproven piece: it is burning billions against little revenue, and whether it becomes a leader is genuinely unknown. We give it option value, not credit as a sure thing.

If you follow only three numbers, follow these: how many Starlink customers SpaceX adds, how much each pays per month, and the profit margin it keeps. Those three drive most of the value we can actually calculate. The catch is already visible in the data: the price per customer has fallen from $99 a month in 2023 to $66 today as Starlink chases the mass market. Growth is excellent; pricing is not — and that tension sits at the center of the debate.

What you must believe to pay today's price. To justify the valuation, SpaceX has to grow into a company generating about $100 billion of cash a year — which implies roughly $340 billion of revenue (eighteen times today's) and ~27% annual growth sustained for more than a decade. It is possible. It is not the most likely outcome. The market is treating a best case as the base case.

Both sides, fairly stated. The bull case: Starlink becomes the world's default internet, Starship makes orbit cheap enough to open new industries, and xAI turns into a top-tier AI lab — and $2 trillion proves cheap. The bear case: Starlink's pricing keeps sliding as rivals arrive, the rockets and the AI keep consuming cash for years, control sits entirely with one person, and a single ordinary disappointment halves the stock. Weighing both, we value SpaceX at about $66 a share. We think the company is extraordinary and the stock, at today's price, is not worth chasing.

0CMedia & Market Research

A structured sweep of reporting around the listing (NPR, CNBC, Morningstar, TechCrunch, Defense News, SpaceNews and others; ~50 dated findings, fact/opinion-tagged, in the Phase 0 snapshot). The signal: a record-breaking debut wrapped around an unusually divided sell side, with the debate hinging on whether Starship/AI monetization arrives before the cash-burn forces a re-rating.

Fact vs. opinion separated and dated in normalized/media_research.json. Two first-party sources (CNBC live blog, TheStreet) 403'd on direct fetch and are corroborated via secondary aggregators; the revenue-multiple (90× vs 94×) and average-PT ($164 mean vs $190 headline) conflicts are carried forward explicitly.

02Investment Context

The listing arrives into a liquidity-rich, risk-on tape — the only conditions under which a $75B raise clears at a fixed price and pops 19%. That same liquidity sensitivity is the macro risk: a ~90×-revenue, long-duration equity is acutely exposed to the rate and liquidity regime, and rising real yields compress terminal-value-heavy names first.

Relevance to the Fund. TON618 Capital is a Bitcoin fund; SPCX is not a BTC-correlated instrument (unlike MSTR or COIN). Its relevance is twofold: (i) a liquidity-regime barometer whose drawdowns would likely coincide with BTC risk-off; and (ii) a potential short / hedge candidate against long-duration tech beta, given the asymmetric valuation. We hold no position.

03Company Overview & Business Model

Space Exploration Technologies Corp. (founded 2002, Hawthorne CA; combined with xAI in February 2026) operates three segments:

• Space (Launch): Falcon 9/Heavy (reusable; ~165 launches in 2025, ~85% of US orbital launches), Dragon, and Starship (in ramp). Commercial & government launch; segment adjusted EBITDA positive since 2018; swung to an operating loss in 2025 on Starship investment.
• Connectivity (Starlink): consumer, enterprise, mobility and government broadband via a LEO laser-mesh constellation; 10.3M subscribers across 164 markets. The profit engine.
• AI (xAI): Grok models, the X platform, and a GPU compute build-out (Colossus). Acquired Feb 2026; consolidated; deeply loss-making (xAI lost $6.4B on ~$3.2B revenue in 2025). The named first use of IPO proceeds.

Capital structure & control. Dual-class (Class A = 1 vote; Class B = 10 votes) plus a non-voting Class C created at IPO. Musk holds ~42% of equity but ~85% of the vote. ~$99.5B pro-forma cash; debt includes legacy X notes ($0.7B 3.875% '27s, $1.0B 5.0% '30s) plus term loans and GPU financing; xAI's ~$17.5B debt is ring-fenced in a subsidiary. Founder ownership is alignment and entrenchment in equal measure.

04Competitive Positioning & Moat

05Proprietary Technology Assessment

• Booster reusability — the defining cost lever; a decade-plus operational lead and a reliability flywheel.
• Laser-mesh constellation — optical inter-satellite links and vertically-integrated terminals; a 10M-subscriber advantage that compounds with density.
• Starship — fully-reusable super-heavy lift. As of Flight 12 (V3 debut, 22 May 2026) it remains suborbital: no orbit, no ship catch, no orbital-refuel demo, and NASA's safety panel flags Artemis HLS as potentially "years late." The ~$10/kg target is aspirational; independents model ~$100–600/kg versus Falcon's ~$2,700. High technical risk, enormous optionality.
• AI flywheel (xAI) — the bull claim is a data (X) × compute (Colossus) × distribution (Starlink) loop, with orbital compute "as early as 2028." Grok is at 117M MAUs but 2025 AI revenue was only ~$0.5B against a multi-billion burn. We assign it option value, not franchise value.

06Financial Statement Analysis

Starlink overtakes launch in 2024 and drives all incremental growth. Source: S-1/A.

• Growth, decelerating in rate: +35% (FY24), +33% (FY25), +15% YoY in Q1'26. Connectivity (+49.8%) masks slower launch.
• Margins: consolidated operating margin negative ((13.9)% FY25) as Starship + AI absorb Starlink's profit. The 2024 net profit ($791M) was the exception.
• Cash quality: OCF is real and growing ($6.8B FY25) but dwarfed by ~$19.6B investing — structurally FCF-negative (~$14B burn), funded by financing (+$26.4B FY25).
• SBC & dilution (critical): weighted shares rose 2,759M→3,884M; xAI and Swarm equity plans assumed; ~$460M cash paid for RSU net-settlement post-Q1. We model SBC as a real, dilutive cost — fully-diluted, not adjusted away.
• Balance sheet: total assets $92.1B (Q1'26); ~$99.5B pro-forma cash; modest debt. Liquidity is not the risk — returns on deployed capital are.

07KPIs & Unit Economics

Subscribers 4.5× in two years while ARPU falls $99→$66. Volume-led growth; pricing under pressure. Source: S-1/A.

The hinge: subscriber growth and margin are exceptional, but ARPU is compressing faster than most models assume and operating income went near-flat QoQ despite subscribers doubling. If terminal ARPU settles near $45–50 in a mass-consumer mix, Connectivity's terminal value is materially lower than the price implies. The ARPU-accretive offset is verticals — aviation (+68%), maritime, and Starshield — where SpaceX still dictates price.

08Business-Model Gearing & Leverage Map

3-input model. Defensible value ≈ Starlink net adds × terminal ARPU × terminal segment margin. Those three drive ~80% of the underwritable value; Starship and xAI are call options layered on top.

Terminal ARPU is the single greatest sensitivity — the crux of the Connectivity debate. Illustrative, anchored on the base SOTP.

Operating leverage sits in Connectivity (fixed constellation, near-zero marginal cost per subscriber — the bull mechanism). Financial leverage is low (net cash). Growth leverage — the launch→constellation→data flywheel — is the qualitative upside the price is paying for.

09Forecast / Operating Model

Modeled fully-diluted with SBC treated as a real cost. Base-case shape (driver-based, illustrative):

Self-funding (FCF breakeven) is not reached until ~FY27 in the base case; the bull pulls it forward, the bear pushes it past FY30. Terminal value is back-loaded — the source of the valuation gap.

10Intrinsic Valuation — Method Selection & Reverse-DCF

Which model. For a high-growth equity, the methodologically-preferred intrinsic lenses are a multi-stage free-cash-flow-to-equity (FCFE) model or, for a dividend payer, a supernormal / multi-stage dividend discount model (two-stage, H-model, or three-stage). We choose to the company's facts — and SpaceX's facts rule the standard single models out today:

• DDM — not applicable. SpaceX pays no dividend and none is planned; cash is fully reinvested.
• FCFE — not yet meaningful. FCFE is negative (the firm is FCF-negative and actively raising capital), so a single-stage FCFE is undefined and a multi-stage FCFE rests entirely on the terminal phase. We would migrate to a two-stage FCFE / H-model once FCFE inflects positive (~FY27 in our base case).
• FCFF — noisy. Capital structure is shifting (the $74B raise, GPU financing), making a single consolidated FCFF DCF low-confidence.

We therefore anchor intrinsic value on the sum-of-the-parts (SOTP), valuing each segment at a life-cycle-appropriate basis (§12), cross-checked by the reverse-DCF below.

Triangulating FCFF DCF. Used only as a cross-check, not the reported value: a 10% WACC (risk-free ~4.3%, ERP ~5.0%, beta ~1.4 de-levered, no size premium), a base-case ramp to ~12% FCF margin by FY30 and ~4% terminal growth yields equity value consistent with the ~$55–66 SOTP base — well below the print.

11Relative Valuation — Peers, Landscape & Indexes

Relative valuation is the most important lens here, and the most demanding: SpaceX has no single public twin, but it has deep, listed competition in all three segments. We benchmark each segment against its real peers, then against the broad market. Multiples as-of the 12 June 2026 close; N/M = not meaningful (negative earnings/EBITDA) — where EV/Sales is the only usable comparable. Sources: market data and stockanalysis.com / S&P Global, retrieved 12–13 Jun 2026.

Landscape. SpaceX's only credible listed pure-play is Rocket Lab, trading at an extraordinary ~85× revenue (no earnings) on the promise of its Neutron rocket — the market pays up for reusable-launch growth. The profitable primes (Lockheed, Northrop) anchor the floor at ~2× revenue / ~13–18× EBITDA, but they are slow-growth conglomerates, not launch pure-plays. SpaceX's reusability and ~80%+ share of global commercial launches justify a premium to the primes; Rocket Lab is the only high-multiple anchor.

Landscape. The satcom set splits in two. Mature incumbents (Iridium, EchoStar, Viasat, SES, Eutelsat) trade as capital-intensive, low-growth businesses at ~3–8× revenue / ~8–18× EBITDA (Iridium the quality outlier). The next-gen direct-to-device names price purely on optionality — AST SpaceMobile at ~376× revenue with no earnings, Globalstar ~38×. Starlink — multi-billion revenue, ~50% growth, near profitability — has no clean analog: the scale and growth of the high-multiple names with real economics, so a fair read-across is a growth-satcom multiple well above the 3–8× incumbent band but far below ASTS's pre-revenue extreme.

Landscape. Profitable hyperscalers trade at ~7–10× revenue; the AI-native growth names (Palantir, CoreWeave, Nebius) fetch ~14–68× revenue, often without profits. xAI's own last private mark (~$230B on ~$3.2–3.8B revenue) implies ~60–78× revenue — rich, pre-profit, scarcity-priced. The market will anchor SpaceX's AI segment to that private mark and the AI-native cohort, not to the cash-generative hyperscalers.

Index multiples are elevated vs. long-run medians (AI-driven). SPCX is loss-making (P/E N/M) and trades at roughly 30× the NASDAQ 100's price-to-sales and ~30× the S&P 500's.

Read-across. On a consolidated basis SPCX's ~90× (offer) to ~109× (close) EV/Sales sits above nearly every peer in every segment — exactly why a single blended multiple misleads and SOTP is required. Yet even applying generous peer multiples segment-by-segment — Starlink at a premium growth-satcom multiple, Launch at a Rocket-Lab-style growth multiple, the AI arm at its ~$230B private mark — sums to roughly $500–650B of enterprise value (our §12 base), a fraction of the ~$2T the market is paying. The relative evidence corroborates the intrinsic work: you cannot reach $135–$161 from what comparable companies are worth without assuming SpaceX becomes something no peer has yet proven possible.

Published targets span $63–$227 (mean $164, median $175); the rating split is 4 buy / 1 sell — a barbell, not a consensus. We do not anchor to any of them; our $66 is built bottom-up. Source: yfinance / Yahoo Finance consensus, retrieved 13 Jun 2026 (CFRA Sell $115 dated 12 Jun).

12Valuation Synthesis — Football Field

Life-cycle stages differ by segment, so SOTP dominates (weighted 70%, the triangulating DCF 30%). The consolidated fair-value range is $27–$128 with a $66 point estimate — a range, not false precision.

13Risk Analysis

Market risk. A newly-listed, ~90×-revenue, low-float, long-duration equity. We expect 30-day realized volatility well above 50% annualized; a 5th-percentile 12-month outcome below $40 is plausible on a risk-off + ARPU-disappointment combination. No listed options as of 13 Jun 2026 (typical for a day-old IPO), so an option-implied VaR/skew read is not yet available — we anchor on peer and early realized vol until the options market opens. First-day turnover (522M shares, ~$85B) confirms ample liquidity.

14Catalysts & Monitoring

Invalidation trip-wires (what would change our call): (i) Starlink ARPU stabilizing above $70 with continued net adds; (ii) the AI segment showing a credible path to cash-flow positive; (iii) a de-rate below ~$55. Any two would move us from Avoid toward Hold/Accumulate.

Ownership — Who Owns SpaceX, and Who Benefits Most

The single most relevant ownership fact is control. The S-1 states it plainly: on completion of the offering, Elon Musk "will beneficially own a majority of the outstanding shares of our Class B common stock and a majority of the voting power… and therefore will be able to elect all the members of our board." EDGAR confirms he is the only 10%+ beneficial owner. SpaceX is, functionally, Musk's company — and the gap between how much he owns and how much he controls is the story.

The wedge comes from the share structure: Class B carries 10 votes to Class A's one. So Musk holds a large minority of the economics (~42% on the S-1's beneficial-ownership basis) but a majority of the votes (~85%). New public shareholders are the mirror image — about 4% of the economics and 1% of the votes.

Reconciled against EDGAR. Nine Form 3s were filed 11 Jun 2026 by SpaceX directors/officers; Musk is the only 10%+ filer, and the largest outside holder is Valor (~503M Class A). Institutional 5%+ holders (Alphabet/Google, Fidelity) had not filed a Schedule 13G as of the report date, so their exact stakes are not yet on EDGAR. Musk's economic/voting figures use the S-1's as-adjusted beneficial-ownership basis; his Form 3 direct holdings do not foot to it share-for-share — a 4 May 2026 five-for-one split, restricted/performance Class B, family trusts/GRATs, and convertible preferred all sit between the two — but every source agrees on the conclusion: Musk controls the company. Sources: SpaceX S-1/A (3 Jun 2026); EDGAR Forms 3 (11 Jun 2026).

Who benefits most. Overwhelmingly Elon Musk. He holds the largest economic stake and near-total voting control, so the upside of success accrues first and foremost to him — and he alone decides how it is pursued (capital allocation, the xAI merger, whether and when to ever pay a dividend). After Musk, the winners are the early venture backers who bought in years ago at a fraction of today's price, and employees holding equity. New public shareholders are last in line on both counts: they are funding the company's ~$14B-a-year cash burn while owning ~4% of the economics and ~1% of the votes. You are buying a minority economic interest in a founder-controlled company, with essentially no ability to influence how it is run. That is not automatically bad — aligned founders can compound value for decades — but it is the central governance fact, and it is why we treat key-person and control concentration (§13, §15) as the dominant non-financial risk.

15ESG & Governance

• Governance (material). Musk ~85% of votes on ~42% of equity via 10-vote Class B; controlled-company exemptions from Nasdaq independence rules; a non-voting Class C; mandatory arbitration plus jury-trial and class-action waivers; removable only by Class B vote. The xAI merger was a related-party deal priced by the controller through Nevada entities (the SolarCity precedent). Minority Class A holders own cash-flow exposure with effectively no control. This is the dominant non-financial risk and the basis of our governance discount.
• Insider ownership. High alignment, high entrenchment; the $1T Tesla pay-package fight is the precedent for how retention can be wielded. Watch Form 4 activity into the lock-up.
• Environmental/social. Net-positive narrative (global connectivity, launch reuse) against orbital-debris, launch-site environmental, and X content-moderation scrutiny now inside the perimeter.

16Recommendation

AVOID at the IPO / first-trade. Fair value $66 (base), $128 (bull), $27 (bear); probability-weighted $66 — 51% below the $135 print and ~59% below the $160.95 first-day close.

• For the Fund: no core position (non-BTC-correlated). The actionable expression is a tactical short / put overlay as a hedge against long-duration tech beta and a liquidity-regime turn — sized small, given borrow cost, squeeze risk on a beloved name, Musk-headline convexity, and the tiered lock-up that delays the natural supply catalyst to ~December 2026. Entry discipline matters more than the thesis.
• Revisit long below ~$55, or on ARPU stabilization + AI cash-flow visibility.
• Expected value: 0.25($128) + 0.50($55) + 0.25($27) = $66. The asymmetry (−80% bear vs. −5% bull) is the point: you are paid little to be right and lose much to be wrong.

17Disclosures & Data Integrity

Primary source. SpaceX Form S-1/A, filed 3 June 2026 (SEC accession 0001628280-26-040364; CIK 0001181412), the operative amended prospectus; original S-1 20 May 2026. Financial figures as-of 31 March 2026 (Q1'26) or fiscal-year-end as labelled. The full Source & Evidence Log is the Phase 0 manifest.json (status: partial — the only gap is pre-IPO earnings transcripts, which do not exist).

Key assumptions & limitations. SOTP segment multiples and scenario probabilities (25/50/25) are TON618 judgments, stated explicitly. Total debt ~$20B is an estimate pending full balance-sheet extraction; net-cash sensitivity is <5% of equity value. Starlink churn, cohort retention, and magic-number are not disclosed at full granularity. Vertical-segment revenue figures (aviation/maritime/Starshield) are third-party estimates, lower-confidence than S-1 consolidated numbers. The DCF is low-confidence given terminal-value dominance and only triangulates the SOTP.

Regulatory status — publisher's exclusion. This report is published solely as general, impersonal information of regular circulation. It is not tailored to the investment objectives or circumstances of any individual and is not issued in connection with compensation from any client. TON618 Capital has no clients and distributes all research free of charge. On that basis the Fund publishes in reliance on the publisher's exclusion from the definition of "investment adviser" under the Investment Advisers Act of 1940 (§202(a)(11)(D); cf. Lowe v. SEC, 472 U.S. 181 (1985)). Nothing herein is personalized investment advice.

Conflicts, positions, compensation & registration. Conflicts of interest: TON618 Capital is a Bitcoin fund and may hold or transact in securities it discusses; material conflicts are disclosed where they exist. Ownership: the Fund holds no position in SPCX as of the report date, and SPCX is not a Fund holding or a BTC-correlated instrument. Compensation: the Fund received no compensation from any party in connection with this report and charges nothing for it. Registration: TON618 Capital is not registered as an investment adviser or broker-dealer in any capacity.

Use of AI & feedback. Artificial intelligence was used in the creation of this report; all methodology and data integrity have been reviewed and approved by a professional. Direct feedback to CIO Keyth Beck, keyth@ton618capital.com. Version 2.0 · company name & ticker now lead the cover in a display-size title; v1.9 analyst consensus verified via yfinance (free; mean $164 / median $175 / high $227 / low $63; 4 buy / 1 sell); v1.8 verified prices & volume vs Massive Market Data ($160.95 close, 522M sh / ~$85B) and noted no listed options yet (§13); v1.7 rebuilt the ownership section with a formal EDGAR reconciliation (control-first; economic-vs-voting wedge; Valor ~503M Class A largest outside holder); v1.5 added the Ownership section + Download-PDF link; v1.4 "The Current State of the SpaceX Story" and the §11 relative valuation; v1.3 compliance disclosures; v1.2 today's $160.95 close and §10 intrinsic-method selection · analyst: TON618 Equity Research.

TON618 Capital. For information purposes only. Not an offer to sell or a solicitation to buy any security or interest. Past performance is not indicative of future results. Digital assets and equities are volatile and may result in loss of capital.