SpaceX spent $2.35 of capex for every dollar of revenue — and its own share price just pushed a bigger unlock into December

In its first quarter as a public company, SpaceX earned $7.8 billion and spent $18.4 billion building for the future — $2.35 of capital for every dollar of revenue, almost all of it on AI computing. That is the number that knocked the stock after hours. Underneath it are two things the release does not lead with: roughly 60% of the revenue that computing serves comes from a single customer who did not exist on the books three months ago, and a third of the debt funding the build is owed to a firm run by one of SpaceX's own directors.

Valuation — no earnings to divide by

At the $115.98 after-hours print the market values SpaceX near $1.65T on ~13.17B shares. There is no P/E: the company lost $541M in the quarter and $9.36B over the trailing year. So every multiple is revenue-based, and they are extreme.

The forward multiple is the telling one: at 32.4× the market already underwrites roughly $51B of forward revenue, a 63% increase on the run rate. That is not a sceptical price.

The drawdown cuts both ways. The stock is 48.6% below its $225.64 June high and 14.1% below the $135 IPO price, so a great deal of enthusiasm has already gone. It also rose 9.4% on the day of the print and gave back 7.5% after hours — the market liked the revenue and disliked the capex, in that order.

The customer

Note 3 discloses two customers above 10% of revenue. Customer A at 18.3% spans all three segments and is long-standing — 20.9%, 24.2% and 25.2% of revenue in 2025, 2024 and 2023, so it predates the cloud business. Customer B at 19.5% is new: it cleared 10% in no prior period, relates only to the AI segment, and at ~$1.52B is roughly 95% of the $1.6B of incremental cloud revenue.

Plainly: essentially all of the AI segment's growth this quarter came from one counterparty that was not material three months ago.

It is not a secret — the prospectus describes an Anthropic arrangement of ~325,000 GPUs at ~$1.25B/month running to May 2029, and the CFO named Google and Anthropic on the call, with Google ramping from October. Note 17 lists only Tesla and Valor as related parties, so the cloud customers are arm's-length — worth ruling out explicitly, and it is ruled out.

The tension the filings don't resolve. The risk factors say cloud agreements "may be terminated by either party upon 90 days' notice." The prospectus describes a term running to 2029. Both are SpaceX's own words: the contracted term is long and the cancellation right is short, and only one of them gets tested if the AI capex cycle turns.

The financing — a director's firm holds a third of the debt

Of $39,364M of total debt, $13,329M — 33.9% — is owed to Valor Equity Partners, whose founder, CEO and Chief Investment Officer Antonio Gracias sits on SpaceX's board. It comes from three equipment lease agreements for AI hardware (November 2025, January 2026, April 2026, the last through CTC Property LLC), all of which failed sale-leaseback accounting — so the assets stay on the balance sheet and the proceeds book as debt. SpaceX guarantees the lessees' payments. The balance nearly tripled in six months from $4,507M, and carries $327M of the quarter's $629M interest expense — 52%.

None of it is hidden; it is in Notes 9 and 17. The terms are not obviously off-market either: the 10-Q states these arrangements carry an average fixed rate of 5.9%, against 5.855% on the public notes.

There is a gap the filing does not reconcile. $327M of quarterly interest against an average related-party balance near $8.9B annualises to the mid-teens, not 5.9%. Failed sale-leaseback accounting can produce interest above a stated coupon, so there are innocent explanations — but the filing gives balances, interest and a rate that do not sit together, and does not say why. It is the thing I would most want asked on the next call.

Risk — each isolated, do not blur

1. The funding gap (dominant). −$25.0B of H1 free cash flow against ~$105B of liquidity — roughly 18 months at the Q2 exit rate before the capital markets are needed again. Investment-grade access today is a condition, not a guarantee.
2. Customer concentration. 37.8% of revenue in two customers; ~60% of the AI segment in one that appeared this quarter.
3. The depreciation assumption. 90% of Adjusted EBITDA is D&A and stock comp, $12.6B of asset is not yet in service, and the 10-Q publishes no useful-life table. If GPU lives are shorter than assumed, the adjustment does the damage on the way back.
4. Starlink unit economics. ARPU −22% and revenue per subscriber −28% while subscribers doubled. Management cites international mix and cheaper plans; no like-for-like is given, so a price cut in mature markets cannot be ruled out.
5. The 2027 wall. $22,244M of the $27,955M of non-cancelable commitments falls in 2027.
6. Operational and legal. The NAACP is seeking to enjoin the gas turbines powering Colossus II under the Clean Air Act; separately, class actions over Grok's image generation, with a $354M accrual.

The lockup — the weakness made the overhang bigger

On August 6, up to 911.5M shares (excluding affiliates) come free — about 1.43× the entire IPO float, roughly $106B. That much was widely reported and is right.

Less noticed: a second tranche of 455.8M "Additional Release Shares" was scheduled for the same date, but only if the stock closed 30% above the $135 IPO price ($175.50) on five of the ten trading days ending on the earnings date. Over that window the closes ran $108.37 to $118.24. The best close was 32.6% below the trigger. Zero of ten days qualified.

So they do not release now. They roll to December 8 — where up to 797.6M shares come free instead of 328.4M. The weak share price did not avoid the dilution; it deferred it and made the December block 469M shares larger. Every figure is an "up to", and the lockup is waivable with Goldman Sachs' written consent.

Horizon and sizing (kept separate)

Horizon. The next 90 days are mechanics rather than fundamentals — the August 6 unlock, the October ramp of the Google agreement, the Cursor close, and a Q3 print showing whether Customer B's revenue is a step or a spike. The thesis resolves over 3–5 years: does 1.4 GW of compute and a doubling constellation earn returns above a depreciation base compounding faster than revenue?

Sizing considerations (not a recommendation). This is the highest-beta expression of the AI-capex cycle — it concentrates an AI-infrastructure basket rather than diversifying it, and adds two risks the others do not carry: a single-customer AI revenue base and a funding requirement returning in ~18 months. Against that, Connectivity is a real, growing, profitable business worth something substantial alone. A position here is not one bet; it is three businesses on one balance sheet, two of which lose money at the operating line.