Financials

Financials — What the Numbers Say

SpaceX's S-1 (filed 2026-05-20, audited by PwC) is the first public look inside the company. Three audited years (FY2023–FY2025) plus Q1 FY2026 are the entire window, and the income statement is rebuilt retrospectively to fold in the xAI and X acquisitions that closed February 2026. Consolidated revenue scales from $10.4B → $14.0B → $18.7B (+34% then +33%), Connectivity (Starlink) flipped from cash drag to a $7.2B segment-EBITDA cash engine, and FY2025 net loss of $4.9B coexists with $6.8B of operating cash flow — the gap is almost entirely capex, and capex more than tripled to $20.7B. The balance sheet was repaired by a $13.1B equity raise in FY2024 and another $18.8B in FY2025; net debt is small but capex outruns operating cash flow by ~$14B and the AI segment is consuming most of the difference. With $1.75T–$2T IPO valuations being floated at ~100x sales, the one financial number that matters right now is AI-segment cash burn, because it sets the equity raise the company still needs.

Quick-reference glossary. Adjusted EBITDA strips depreciation, interest, taxes, stock-based compensation, and impairments out of earnings — management's preferred profitability lens, but it ignores the capex that actually consumes the cash. Free cash flow (FCF) is operating cash flow minus capex; it is the cash left for owners after the business reinvests. Segment Adjusted EBITDA is the company's split of that measure across the three reportable segments — Space (Falcon/Dragon/Starship launch for paying customers), Connectivity (Starlink consumer/enterprise/government broadband and direct-to-cell), and AI (the newly merged xAI/X businesses). Where this page uses "Quality Score" or "Fair Value," those headline ranking metrics were not produced for SpaceX because the underlying data provider had no probe for an unlisted issuer — there is no public price to scale against.

Revenue, Margins, and Earnings Power

This is a company with three businesses moving in three directions. Top-line growth is real (+34% then +33%), gross margin has expanded almost 8 percentage points in two years (41.2% → 49.4%), and yet operating margin swung from +3.3% in FY2024 to −13.9% in FY2025 because research-and-development spending doubled from $3.5B to $8.6B — most of it dropped in by the xAI consolidation. The clean read: the underlying launch + Starlink franchise is approaching a high-software-margin profile; the AI segment is what creates the GAAP loss.

How to read this. Gross margin rising 800 bps in two years tells you Starlink scaling is dropping cash into the contribution line — reusable Falcon-9 launches deploy satellites at marginal cost, and incremental subscribers are pure subscription. Operating margin's collapse is not deterioration of the existing franchise; it is the deliberate fold-in of xAI's compute build-out. R&D as a percent of sales went from 20% to 46% in one year — that is the AI segment, and management has explicitly said this cost line will stay elevated for a "multi-year investment horizon."

Quarterly trajectory

Revenue only +15% YoY in Q1 — the Connectivity segment grew 50% in the year but Starlink's ARPU is now $66/month vs $86 a year earlier (international mix, lower-tier plans), so subscriber growth (~105% YoY to 10.3 million) is outrunning revenue growth. The most important quarterly trend in the file: Starlink is choosing volume over price, betting that scale economics absorb the ARPU erosion.

Cash Flow and Earnings Quality

Operating cash flow looks impressive — $4.5B → $5.8B → $6.8B — but every dollar of it has been swallowed by capex. Free cash flow is deeply negative and the gap is widening: −$0.1B → −$5.4B → −$14.0B. The company prints positive EBITDA, but it is not yet a self-funding business after the AI build.

Earnings quality bridge. Net income is a poor read of cash because the income statement carries $6.7B of depreciation and amortization, $1.9B of stock-based compensation, and (in FY2023) a $3.8B intangible impairment. Strip those out and operating cash flow lands well above net income — the right read of operating quality. But the chart that matters is the capex bar: in FY2025 the company spent $20.7B on property/plant/equipment ($3.8B Space, $4.2B Connectivity, $12.7B AI), and AI capex jumped to $7.7B in Q1 2026 alone (annualized ~$31B). This is the cash-flow distortion to watch.

Major cash-flow line items

The single biggest takeaway: FY2025 is funded by external capital, not by the business. $18.8B of equity and $16.1B of debt proceeds were raised; that combined ~$35B inflow covered the ~$14B FCF deficit with room to refinance and stockpile $24.7B of cash. SBC is moving up fast ($1.9B in FY2025, $2.5B annualized in Q1 2026) — at IPO valuations this is real per-share dilution and should be tracked.

Balance Sheet and Financial Resilience

The balance sheet is the strongest part of the story but is also where the IPO mechanics land. Cash + marketable securities ended FY2025 at $24.7B; after the xAI close in Q1 2026 it fell to $23.7B (cash $15.9B + securities $7.8B), debt rose to $30.3B, and shareholders' equity exploded from $2.6B to $34.5B as $38.8B of redeemable convertible preferred converted to common.

Two things to understand. First, redeemable convertible preferred sits above common equity in the capital structure: at FY2025 it was $38.8B — larger than total liabilities ex-debt — and effectively converted at the Q1 2026 close, blowing additional paid-in capital from $35.9B to $74.1B. Existing private holders own most of that converted stack. Second, goodwill of $11.8B is ~13% of total assets and is almost entirely tied to the X (Twitter) acquisition — a known impairment risk if AI monetization disappoints. The Altman Z-Score, Piotroski F-Score, and Beneish M-Score are not available (no public probe for an unlisted issuer), so resilience must be read off the raw line items.

Returns, Reinvestment, and Capital Allocation

Standard ROIC and ROE are not meaningful when GAAP net income is negative and total invested capital just stepped up by ~$30B via merger. Asset turnover gives a cleaner read:

Asset turnover falling from 0.25 to 0.20 even as revenue grew 33% tells you the asset base is growing faster than sales — capex is being installed in advance of monetization (the AI compute build, V3 Starlink satellites, Starbase Starship expansion). Consistent with a company at the front of an investment cycle, not the middle of one.

Share count and dilution

The FY2024 spike in diluted share count to ~10.0B reflects the impact of the convertible preferred under the if-converted method when GAAP net income flipped briefly positive. The Q1 2026 jump to 3.88B is the actual xAI/X consolidation. Stock repurchases of $1.0B–$1.1B per year offset some employee award dilution but are small versus an SBC expense that just reached $1.9B annually — net dilution is real and accelerating.

Capital allocation summary

The judgment. Management is allocating capital exactly the way an early-stage compounder should — pour it into the highest-return franchise (Connectivity, where segment EBITDA grew 86% to $7.2B) and the segment with the largest unproven option (AI). They are not buying back stock at scale, not paying a dividend, not doing tuck-ins. Whether this works depends on AI cash-flow timing — there is no buffer if the optionality fails.

Segment and Unit Economics

The three-segment cut is where this report stops looking like an aerospace company and starts looking like three businesses stitched together: a reusable-rocket franchise that is barely profitable on its own, a connectivity business that is the actual cash engine, and an AI build-out funded by both.

Connectivity is the franchise. $11.4B of revenue (61% of total), $7.2B of segment EBITDA — a ~63% segment-EBITDA margin. Growth was 49.8% in FY2025 and 50%+ continuing in Q1 2026 (Q1: $3.3B revenue, $2.1B segment EBITDA). 10.3 million subscribers, ~75% of all active maneuverable satellites in orbit. ARPU is falling ($91 → $81 → $66 in Q1) but that is intentional — international expansion at lower price points trading ARPU for volume.

Space is the platform. $4.1B revenue, segment EBITDA of $653M despite funding $3.0B of Starship R&D. Reported customer launches were 43 in FY2025 versus 122 internal Starlink launches; reported Space revenue therefore understates the actual cadence and economics of the launch business because internal launches are capitalized into the Connectivity segment as satellite deployment cost. This is the textbook reason segment GAAP can mislead — the launch franchise's true economic contribution is hidden inside Connectivity's capex line.

AI is the option. $3.2B revenue, $6.4B loss from operations, $12.7B capex in FY2025 and $7.7B in Q1 2026 alone. Nameplate compute draw is now 1.0 GW (COLOSSUS + COLOSSUS II online). The math here is straightforward: AI segment burned ~$8B of cash in FY2025 (operating loss plus capex net of depreciation), and if Q1 run-rate holds, AI burn is closer to $30B/year. Either AI revenue inflects sharply within 2–3 years or this segment becomes the primary funding driver of further equity raises and the primary impairment risk on the goodwill stack.

Valuation and Market Expectations

Standard public-market valuation multiples are not yet available for SpaceX — no listed price, no listed market cap. What is observable is the IPO-process anchor: secondary markets and broker reports point to $1.75T to $2T target valuation on the IPO, against $18.7B of FY2025 revenue and $6.6B of Adjusted EBITDA. That implies:

For context, an AJ Bell analyst quoted at $1.75T called it "67 times sales, three times Nvidia's rating." A separate broker analysis pegged it at ~113x sales. EV/FCF is undefined because FCF is negative.

The valuation judgment. At $1.75T:

• Versus history: SpaceX's last private round (Q3 2025 tender) implied ~$500B. IPO ask is ~3.5x that price on the same business — most of the step-up is the xAI fold-in plus the AI optionality framing.
• Versus peers: RKLB and ASTS both trade at extreme EV/sales multiples too, but they are tiny ($600M and $71M of revenue), not yet profitable, and trade on optionality. The mature LEO comparable, Iridium, trades at 7.4x EV/sales. SpaceX deserves a premium for scale, margin, and growth — how much premium is the entire question.
• Versus growth and margin: 93% pro-forma revenue growth, 49% gross margin, 35% adjusted-EBITDA margin (consolidated, FY2025), 30%+ free-cash-flow-margin if you carve out the AI segment. That justifies a substantial multiple — perhaps 20–40x sales for the launch + Starlink core — but the $1.75T+ ask requires the AI segment to be valued like a pre-revenue AI infrastructure pure-play.

Bear / base / bull on FY2027 revenue and exit multiple:

A $1.75T IPO sits at the upper edge of the base case — the market is paying for the bull and giving the seller most of the upside. EV/sales valuations of 35–50x are reserved for high-growth software at scale; SpaceX's underlying mix is more capital-intensive than that comparison flatters.

Peer Financial Comparison

Peer-gap judgment. SpaceX is uniquely placed on this table — the only company with both Iridium-like profitability and RKLB-like growth, at 20-30x the revenue scale of either pure-play. Iridium is the closest mature LEO comparable on margin, but trades at 7.4x sales because growth is ~4%. RKLB and ASTS trade at the same elevated multiples as the SpaceX IPO ask but on early-stage, unprofitable businesses. The honest read: SpaceX deserves a meaningful premium to LMT/IRDM (it grows faster, earns better margins) but the IPO valuation is asking the market to underwrite AI segment success — that is the gap with reality.

What to Watch in the Financials

Closing read. The financials confirm that SpaceX has built two distinct franchises — a launch platform with structural cost leadership and a connectivity business with rapidly expanding margins. They contradict the framing that this is a single growth business: it is two profitable segments (Space + Connectivity, combined ~$15.5B revenue and ~$7.8B segment EBITDA) plus a third segment (AI) absorbing roughly all of it in capex and operating losses. The IPO valuation is therefore not pricing the cash-generating franchise; it is pricing the AI option. The first financial metric to watch is AI segment Adjusted EBITDA and capex run-rate — because every other line item in this report (further equity raises, balance-sheet leverage, FCF trajectory, goodwill impairment risk, ultimately the multiple) flexes with it.

The first financial metric to watch is AI segment capital expenditure as a multiple of AI segment revenue — currently 4.0x in FY2025 and rising to 9.4x in Q1 2026, and either it bends down or further dilution is the funding mechanism.