This article is over 1 month old.
SpaceX S-1 filing - an analyst’s breakdown
Key takeaways from the financial disclosure that every investor needs to know.
Updated 4 Nov 2026
SpaceX is not a rocket company that filed for an IPO. It's three businesses stapled together, a profitable satellite ISP (internet service provider), a money-losing launch infrastructure operation that cross-subsidises everything else, and a freshly-bolted-on AI play (xAI + X, merged in early 2026) that is burning extraordinary amounts of capital. From the S-1 I can see that apart from the extraordinary connectivity business, the others are lackluster.
SpaceX EBITDA for each Business Unit
Chart of each segment EBITDA generation (B$)
Segment 1: Connectivity (Starlink + Starshield)
This is the brightest spot in SpaceX’s overall dark business. The connectivity segment has generated $11.4B in revenue and $4.42B in operating income in FY2025, representing year-over-year growth of 49.8% and 120.4% respectively. Segment Adjusted EBITDA was $7.17B, up 86.2% YoY. That is an extraordinary combination, a nearly 50% top line grower with over 60% EBITDA margins at the segment level. These are big tech levels of growth, not infrastructure levels, however given the valuation it should be expected.
As of March 31, 2026, Starlink serves approximately 10.3 million subscribers across 164 countries, territories, and other markets, powered by 9,600 broadband and mobile satellites in LEO.
The moat here is structural and formidable. Starlink's LEO constellation delivers low-latency broadband (<70ms) to locations where terrestrial infrastructure is uneconomical. No competitor has anything close in scale, OneWeb is years behind, Amazon's Kuiper is unproven, and traditional GEO satellites have physics-imposed latency disadvantages. SpaceX's vertically integrated manufacturing gives them an advantage currently no one can match.
Segment 2: Space (Falcon, Dragon, Starship)
Since 2023, SpaceX has achieved nearly a monopoly on the global space market launching more than 80% of mass to orbit globally each year with an over 99% mission success rate with Falcon rockets. That degree of market dominance in a strategically critical industry is without precedent in modern industrial history. Falcon 9 has effectively monopolised commercial launch.
However, the Space segment is currently a financial disaster. In Q1 2026, Space generated only $619M in revenue against a $662M operating loss and $351M Adjusted EBITDA. In FY2025, it managed $4.09B revenue but only $653M Adj. EBITDA despite being the backbone of the company's competitive advantages.
The reason is Starship. The Space segment funded $930M of R&D expense in Q1 2026 alone and $3.0B in all of FY2025 for the Starship programme. These are pre-revenue development costs being expensed through the P&L. The narrative is that Starship is an investment that unlocks everything else, and that might be correct, but investors are being asked to fund a programme with no GAAP revenue yet against enormous capital consumption.
Segment 3: AI (xAI, Grok, X)
This segment is where I think SpaceX has the most weakness. xAI was officially bought by SpaceX on February 2, 2026 with the goals of developing space-based data centers powered by SpaceX's Starship megarocket and expanding the capabilities of xAI’s Grok chatbot. These goals seem very unattainable and a waste of CAPEX.
The financials show an even worse picture of this segment of the whole business. The AI segment lost $6.36B from operations in FY2025 on just $3.2B of revenue. In Q1 2026, it generated $818M revenue against a $2.47B operating loss. That's a -300% operating margin. The Adj. EBITDA was $1.24B in FY2025.
FY2025 capex for the AI segment was $12.73B. In Q1 2026 alone it was $7.72B. Annualise that Q1 number and you're looking at $30B+ of capex in the AI segment in 2026. This further reinforces my point of AI being a waste of CAPEX.
Bottom Line:
From researching SpaceX’s S-1 filing I can see that its connectivity business is the reason to own this. It's one of the best infrastructure businesses in the world by any objective quality metric, growing at 50%, with a widening moat and zero credible competition at scale.
However, with two very unprofitable segments at its side this case gets more complicated. On one side there is a profitable satellite company, on the other side is an unprofitable AI and rocket business.
If you are willing to buy a company valued at $1.8 trillion and has not made any profits, but believe in Elon’s vision for it, then SpaceX will definitely be a buy, otherwise it will be a definite no for me and I’m sure other traders around the world as well.
SpaceX Stock (NASDAQ)
SPCX