Financials

Financials in One Page

SpaceX is unusual: a $18.7B-revenue, three-segment industrial-plus-internet company filing to list at a target valuation of roughly $1.75 trillion. The income statement is a tale of three businesses. Connectivity (Starlink) is the profit engine — $11.4B of FY2025 revenue at a 39% operating margin and growing nearly 50% a year. Space (Falcon plus Starship development) is a high-cadence launch monopoly that earns positive segment EBITDA but is reinvested through the Starship R&D line. AI (the xAI / Grok business consolidated after the early-2026 acquisition) is a deeply loss-making, capital-hungry start-up bolted onto the rest. Reported FY2025 GAAP results — a $2.6B operating loss and a $4.9B net loss — therefore mean almost nothing on a clean look-through basis: strip out the AI segment and the legacy SpaceX business produced roughly $3.8B of operating income on $15.5B of revenue.

Earnings quality is the central debate. Operating cash flow has compounded from $4.5B to $6.8B in two years, but capex (a record $20.8B in FY2025) and the xAI build-out have driven free cash flow to negative $14.0B. The business is funded by an equity raise (>$19B issued in FY2025) and term debt, not by the income statement. The balance sheet is large but flexible: $24.7B cash against $22.9B total debt, financed by a $41.3B equity base that the IPO will materially expand. Valuation is unprecedented: at the reported IPO target, the equity prints ~94x trailing sales and ~266x trailing Adjusted EBITDA — only justifiable if you underwrite Starlink as a global telecom platform compounding at 30-50%, Starship cutting cost-to-orbit by another order of magnitude, and AI compute scaling into a real profit pool. The single financial metric that matters most right now is Connectivity Segment Adjusted EBITDA — it is the only line that pays for everything else.

Revenue, Margins, and Earnings Power

The headline question is whether SpaceX is a profitable business. The honest answer is yes for two of three segments, no for the third, and the consolidation hides that.

Revenue compounded 34% from FY2023 to FY2025 — $10.4B to $18.7B — with growth accelerating in absolute dollars each year. Gross profit grew faster than revenue (53% per year), with gross margin expanding from 41.2% in FY2023 to 49.4% in FY2025 as Starlink scaled and the launch fleet's reusability economics improved.

The margin story is the clearest signal in the financials: gross margin has improved every year and is now near 50%, which is unusual for a hardware-plus-services business. Operating margin oscillated from a deep loss in FY2023, to a slim profit in FY2024, back to a deep loss in FY2025. The driver is R&D plus AI consolidation, not core economics deteriorating.

R&D went from $2.1B (20% of revenue) in FY2023 to $8.6B (46% of revenue) in FY2025. The S-1 splits that across three buckets: $3.0B for Starship development inside Space, ongoing Starlink network improvements inside Connectivity, and the AI / Grok model and compute investment that arrived with the xAI deal. This is the single biggest reason the company looks unprofitable — it is choosing to be.

Segment view — where the money actually comes from

Connectivity carries the economics. On $11.4B of FY2025 revenue, the segment posted $4.4B of operating income (39% margin) and $7.2B of segment Adjusted EBITDA, growing roughly 50% top-line and 120% in operating profit year over year. Subscribers more than doubled to 10.3M and continue to scale. Space is breakeven on segment Adjusted EBITDA only because Starship R&D — a strategic, optional investment — drags the line. AI lost $1.2B on $3.2B of revenue, primarily because the business is being built. The investor framing is this: Connectivity is a real, growing utility; Space is a launch monopoly intentionally reinvested; AI is venture capital wearing public-company clothing.

Recent quarter — what changed in Q1 FY2026

Q1 FY2026 revenue of $4.7B is up 15% year over year, but operating income swung from positive $27M to negative $1.94B. The reason is mechanical: Q1 FY2026 is the first quarter that consolidates xAI inside SpaceX. AI segment alone lost $2.47B operating in Q1, while Connectivity posted $1.19B of operating income and Space lost $662M (with $930M Starship R&D inside that). The right way to read the quarter is "ex-AI, the legacy business earned over $500M of operating profit on $3.9B of revenue." The 15% reported revenue growth understates legacy momentum; Connectivity alone grew much faster.

Cash Flow and Earnings Quality

The pattern here is a classic capital-intensive scaler — strong and growing operating cash flow, but capex larger than OCF, so headline free cash flow is negative.

Two important observations. First, operating cash flow is dramatically higher than net income in every year — the gap is depreciation (rockets and satellites being amortized through the income statement), stock-based compensation, and customer prepayments (notably for Starlink hardware and launch contracts). That is good earnings quality at the operating line: the business converts customer demand into cash. Second, free cash flow has gone the other way as capex exploded — capital expenditure rose from $4.4B in FY2023 to $20.8B in FY2025. The biggest single bucket inside that FY2025 capex line is the AI segment ($12.7B), followed by Connectivity ($4.2B) and Space ($3.8B).

The takeaway: SpaceX's legacy Space + Connectivity business spent ~$8.0B of capex in FY2025 on roughly $11.7B of operating cash flow attributable to those segments — i.e., it is already self-funding. The negative consolidated FCF is the AI investment phase, paid for with equity issuance.

Stock-based compensation deserves attention. SBC rose from $679M (FY2023) to $1.95B (FY2025) and is excluded from Adjusted EBITDA. On the IPO-target market cap of $1.75T, $1.95B of SBC is only ~11 basis points of dilution — small. On a peer basis, however, it should be tracked, because the xAI deal materially expanded the share base (period-end shares jumped from 3.08B to 5.80B in Q1 FY2026).

Balance Sheet and Financial Resilience

The balance sheet has grown faster than the income statement. Total assets nearly doubled in one year — $57B to $92B — driven by the $11.8B xAI goodwill, $21B more property, plant and equipment (Starlink constellation, Starbase build-out, AI compute), and $12.6B more cash.

Net debt definition: total debt minus cash. SpaceX ended FY2024 with $1.6B of net debt and ended FY2025 with $1.85B of net cash (cash exceeds debt). On Adjusted EBITDA of $6.58B, gross debt is 3.5x EBITDA, but net debt is essentially zero — comfortable for a company with this revenue scale and this share of contracted government and subscription revenue.

The IPO itself materially changes the resilience picture. The reported raise target is $75B of fresh equity. Even if only a fraction is primary, the company will end the IPO with a cash balance well above $40B and net cash near $20B — more than enough to fund the AI segment build-out without further dilution for several years.

Accounts payable nearly tripled to $11.8B — partly the timing of large supplier payments on Starship and AI compute, partly customer-prepayment dynamics. Inventory and receivables grew in line with revenue. Long-term debt rose by $8.5B in the year as the company drew on credit lines and term loans to bridge the heavy capex period before the IPO.

Returns, Reinvestment, and Capital Allocation

GAAP return on equity is meaningless here — equity is being raised faster than income is reported, and most of the equity is funding assets that will produce returns in future years.

The more useful question is whether the cash being invested earns a sensible return. The clean way to look at it: Connectivity (Starlink) is throwing off ~$7.2B of segment Adjusted EBITDA on roughly $20B of cumulative invested capex over the past three years. That implies a 35%+ cash return on installed capital base, before you give credit for further subscriber growth. Space, ex-Starship R&D, is also profitable. AI returns will not be known for years.

Capital allocation in FY2025 was almost entirely about growth: $20.8B of capex, $7.2B of debt paydown, and a small $1.1B buyback used to mop up secondary-market share grants. The $19.1B equity issuance funded the gap. There is no dividend, no large repurchase program, and the financial logic is straightforward — every dollar of internal cash and external financing is being directed at Starship (Space), constellation expansion (Connectivity), and AI compute build-out.

Share count was nearly flat 2023-2025, then jumped 88% in Q1 FY2026 from the xAI stock-for-stock consolidation. The dilution is dramatic but should be evaluated against what was added: a $3.2B-revenue AI franchise plus the long-dated optionality on Grok and AI compute. Whether that consideration was paid for at a fair price is a separate question — answered partly by the IPO multiple the market is willing to attach to the combined entity.

Segment and Unit Economics

The S-1 discloses three reportable segments and the key business metrics inside them. The segment table is the single most important page in the financials.

Connectivity revenue grew 49.8% in FY2025 and Connectivity operating income grew 120%. Underneath that: Starlink subscribers reached 10.3M by March 2026, up 105% year over year. The subscriber number is the single most important unit-economics input — at roughly $90/month average revenue per user across the residential, roam, business, mobility, and maritime tiers, the run-rate Starlink revenue is on a trajectory through $13-15B for FY2026 before counting government Starshield growth.

Inside Space, the relevant unit metric is launch cadence: 170 launches in FY2025 from 159 flight-proven boosters, with the company stating it carried over 80% of mass to orbit globally. Falcon launch services and Dragon contracts generate the cash that funds Starship R&D — which is itself the largest single R&D bucket the company discloses.

Valuation and Market Expectations

There is no historical valuation chart because there is no public trading history. The S-1 filed 2026-05-20 targets a listing on 2026-06-12 at a reported $1.75 trillion valuation raising $75B. The benchmark question is what the implied multiples look like on that reference price.

These are extreme multiples in absolute terms but have to be evaluated against three things at once: a $11.4B Connectivity business growing 50% and already at 39% operating margin, an effectively monopoly Space business hidden inside the loss line because of Starship R&D, and an AI business still in the heavy-investment phase whose terminal economics are unknown.

A simple way to triangulate is to value Connectivity standalone. At a 12x revenue multiple — high but defensible for a 50%-growth, high-margin, infrastructure-class subscriber business — Starlink alone would be worth ~$137B. Apply 7x to Space revenue with the Falcon manifest building and Starship optionality on top — call it $29B. The IPO target therefore embeds roughly $1.58 trillion of value for the AI business plus optionality (Mars program, satellite-to-cell, defense, deep-space). That is the gap the buyer is paying for.

A Quality Score, Fair Value, and Predictability rating from any third-party model are not available for this company at this stage (no public-market history, no consensus estimates). That absence is itself information — the multiple cannot be sanity-checked against a long earnings record. Investors are underwriting management forecasts and the long-run market sizes for satellite broadband, defense launch, and AI compute, not stable cash earnings yet.

Peer Financial Comparison

The peer set is the listed connectivity and launch universe — but none of them is anywhere near SpaceX in scale, and only Iridium is profitable.

What the peer table says: SpaceX is 4-30x the revenue of any listed competitor, with a gross margin (49%) above all of them except IRDM (which has a smaller, mature LEO business) and ASTS / GSAT (which have negligible revenue). On EV/Sales it sits between RKLB (67x, pre-profit) and the asset-heavy LEO incumbents. The premium to RKLB on revenue multiple is modest given that SpaceX has a vastly larger and more diversified revenue base and is already producing meaningful Adjusted EBITDA. The valuation gap to IRDM (which trades at 4x EV/Sales as a mature, profitable but slow-growth peer) is large and reflects the bet on growth and AI — not on current cash earnings.

What to Watch in the Financials

What the financials confirm. Starlink works as a business — high gross margin, scaling revenue, scaling Segment EBITDA, and a real subscriber metric to track it. Falcon launch services generate real cash and have built the world's only economically reusable orbital launch system. Operating cash flow has compounded faster than revenue.

What the financials contradict. The "$18.7B revenue, GAAP loss" headline is misleading. The legacy Space + Connectivity business produced ~$3.8B of operating income in FY2025 on $15.5B of revenue. The reported loss is the AI consolidation. Equally, the deep free-cash-flow deficit is a choice — capex is running ~3x operating cash flow because management is choosing to build Starship, expand Starlink, and grow AI compute simultaneously, not because the core economics are weak.

The first financial metric to watch is Connectivity Segment Adjusted EBITDA. If Starlink holds its growth and margin path, the IPO multiple becomes underwritable from cash earnings within five years. If it slows — through Kuiper pricing, D2C competition, or churn — the valuation thesis rests on AI and Mars, both of which are decades from contributing comparable cash.