The foundation of SpaceX's trillion-dollar valuation: Who is dividing Musk's annual capital expenditure of tens of billions?

Author: nini

If you missed the Apple supply chain in 2010, missed the Tesla supply chain in 2020, or even regretted missing the Nvidia supply chain in the past two years,

the SpaceX supply chain has just begun.

Of course, I think simply chasing SpaceX itself doesn't seem worthwhile. It rose 19% on its IPO day, priced at 135 and surged to 160, with a price-to-sales ratio close to 100 times, while the company is still losing money. Retail investors rushing in on the first day of listing face considerable pressure.

So what I want to talk about is the companies supplying it.

History has repeatedly verified the same logic: super terminals crazily feed back the supply chain behind them. In 2010, when Apple launched the iPhone 4, Luxshare Precision had a revenue of 1 billion that year, and ten years later it reached 92.5 billion, with its stock price increasing 30 times. In 2019, when Tesla's Shanghai factory started production, CATL had a market value just over 100 billion, and five years later it broke a trillion. Nvidia has exploded in popularity in the past two years, and Zhongji Xuchuang has grown from tens of billions to over a hundred billion in market value.

Apple, Tesla, Nvidia------every time the super terminal stands in the spotlight, but the ones who can truly make a lot of money are the supply chain companies behind them.

SpaceX spends hundreds of billions of dollars each year on chips, materials, parts, and industrial gases. These procurement orders gradually turn into real revenue on the books of certain companies. After the prospectus was made public, this supply chain finally had data that could be traced.

Let's first look at where SpaceX's money comes from and where it goes.

Its business mainly consists of three parts. The first part is Starlink. Last year, it generated 11.3 billion dollars in revenue, accounting for 60% of the group, with over 10 million global subscribers------this is the only part of SpaceX that makes stable profits, and we can even say that all the money-burning areas rely on it for funding.

The second part is rockets. The Falcon and Starship have an annual R&D investment of 3 billion dollars, resulting in the lowest commercial launch costs globally, with plans for 100 launches in 2026 and a demand for 1,500 Raptor engines. The third part is AI. Last year it lost over 6 billion, and it is building the Colossus supercomputer on the ground, stacking 220,000 GPUs, with plans for an orbital data center in space.

So, the flow of money is simple: money earned from Starlink → invested in rockets to lower launch costs → low-cost launches send AI hardware into space → AI computing power is rented out to earn money. This is roughly the cycle.

This cycle releases hundreds of billions of dollars in procurement orders each year, so who pockets this money?

Divided by whether they can be replaced, suppliers fall into three categories.

The first category: irreplaceable in the short term

1. Nvidia, with all 220,000 GPUs for the Colossus supercomputer being theirs. But Nvidia's real moat is not hardware, but CUDA; almost all AI training worldwide uses this software ecosystem to write code. Hardware can be replaced, but the migration cost of ten years of code cannot be compensated in just a year or two. We can understand that as long as SpaceX is still building supercomputers, Nvidia will keep making money.
2. Eutelsat, ticker SATS. It holds the radio frequency spectrum for satellite communications. What is the spectrum? You can think of it as lanes in the sky; physical laws determine that there are only a few, and whoever occupies them first owns them. No matter how strong your technology is, you cannot create a segment out of thin air. Musk's phone-to-satellite feature must route through them; if you don't pay the toll, the signal will collide with other satellites. Moreover, SATS holds about 3% of SpaceX's shares. It rose 11% the day before the IPO, with options trading volume 11 times the usual.
3. Filtronic, ticker FTC, listed in London, note that it cannot be found on US stock markets. It makes millimeter-wave signal amplifiers for Starlink satellites, allowing signals to be transmitted farther and clearer. In 2024, it signed a contract worth 47.3 million pounds, with SpaceX contributing 83% of its revenue, and it also secured a maximum 10% subscription right. This may seem small, but aerospace-grade certification requires years of repeated testing under vacuum, radiation, and extreme temperature differences. Once certified, SpaceX won't easily switch suppliers because the re-certification cycle cannot keep up with production ramp-up. Moreover, Filtronic's stock price has nearly doubled in a year.
4. Materion, ticker MTRN. The only integrated producer of beryllium metal from ore to finished product globally, controlling about 56% of the supply. Beryllium is one-third lighter than aluminum, six times stronger than steel, with a melting point close to 1300 degrees Celsius; it is a metal that meets all three conditions of being light, hard, and heat-resistant, which is rare on Earth. The F-35 fighter jet, the James Webb Space Telescope's lens, and the structural components of Starship all use it. The U.S. Department of Defense lists beryllium as a strategic material, and Materion has been the exclusive certified supplier for the F-35 for over ten years. This shows its scarcity.
5. STMicroelectronics, ticker STM. It helps SpaceX make phased array antenna chips, with cumulative deliveries exceeding 5 billion, covering over 10,000 satellites. STM predicts that the low Earth orbit satellite business will reach 2 billion dollars by 2028 and 2.9 billion dollars by 2030.

The second category: technically replaceable, but the cost of replacement is too high

1. Honeywell, ticker HON. It controls the flight control and inertial navigation systems of rockets------the rocket knows where it is, where it is flying, and what posture to maintain, all controlled by it. From Apollo to the Space Shuttle to commercial spaceflight, decades of certification have gradually accumulated. Changing suppliers is equivalent to re-implanting the rocket's brain; all underlying code must be rewritten, and new certifications must start from scratch. SpaceX launches over a hundred times a year, and it cannot afford to stop launch schedules just to save on procurement costs.
2. Carpenter Technology, ticker CRS. It is used to refine special steel alloys for Raptor engines. Vacuum melting and repeated purification control impurities to a level of one part per million. A slight error could lead to disaster in the combustion chamber. This material process cannot be transmitted solely by blueprints; building an equivalent production line could take decades or more.
3. Hexcel, ticker HXL. It supplies aerospace carbon fiber; every kilogram less in rocket weight means one kilogram less in effective payload. Carbon fiber is half the weight of metal for the same strength. It has collaborated with SpaceX for over a decade, with material formulas and weaving processes specifically adjusted to meet SpaceX's needs. If replaced, the entire material system must be validated from scratch.
4. Broadcom AVGO, responsible for 10Gbps data exchange between space and ground. To ensure high-speed data flow without congestion, it relies on them. Linde Group invested 100 million dollars in 2025 to build an air separation plant near the Texas Starbase, exclusively for liquid oxygen and liquid nitrogen, due to the large consumption of high-purity industrial gases during rocket launches; the closer the distance, the lower the cost, making this location itself a moat.

The third category: requires stable mass production and cost minimized

You may not have seen the Starlink dish in person, but think about it; they need to deploy a total of 30 million units globally. Each unit contains thousands of components and dozens of processes, needing to be produced on an assembly line like making a phone, while also withstanding aerospace-level vibrations and temperature differences.

At this scale, technology is no longer the primary factor; the most important thing is who can deliver stably and who can minimize costs.

The logic of Foxconn manufacturing for Apple applies here exactly. Qiqi Technology, ticker 6285, is the largest global contract manufacturer for Starlink terminals and routers. The quality control standards were developed through years of collaboration with SpaceX, so it's not just any factory that can take on this work.

Moving up are several A-share companies. Xunwei Communication, 300136, is the exclusive global supplier of high-frequency connectors in Starlink terminals, with SpaceX-related orders expected to be about 1.05 billion by 2025. Parker New Materials, 605123, is the only Chinese supplier of forged components for Starship's body and engines, with orders around 680 million, accounting for 35% of the company's revenue. Western Materials, 002149, is the exclusive supplier of niobium alloys for Raptor engines, with orders around 1.02 billion. Yingliu Co., 603308, supplies core castings for Raptor turbine pumps, accounting for 42% of its own revenue------SpaceX's orders have already become the largest source of income for this company.

Going smaller, Tianyin Electromechanical can be likened to the star sensors on Starlink satellites, which help the satellites determine their posture by observing stars, and it has over 60% market share. Tongyu Communication makes ground antenna modules for Starlink, with expected orders of 300 million in 2026.

On the US stock market, there are also a few companies. Trimble, ticker TRMB, manages time; thousands of satellites are flying in the sky, and each clock must be synchronized to the same beat; a microsecond of difference can lead to communication errors. Astronics, ticker ATRO, manages rocket power distribution. CTS, ticker CTSH, manages heat dissipation. None of these are black technologies, but they are all indispensable screws in the entire system.

You may ask, these companies have always been there, why now?

Three reasons.

• First, the procurement volume has just begun to increase. There are plans for 100 launches in 2026, Starship is accelerating testing, and the AI data center will start deploying into space in 2028. The target for Starlink terminals is 30 million units, and there are currently only 10 million subscribers. The speed at which SpaceX is spending money is far from its peak.
• Second, transparency has opened up for the first time. Previously, SpaceX was a private company, and procurement data was a black box. After the prospectus was made public, quarterly and annual reports will continuously disclose, allowing the order growth rate of supply chain companies to be tracked and verified.
• Third, referencing the rhythm of history. The Apple supply chain took ten years to go from the iPhone 4 to its peak. The Tesla supply chain has taken seven years from the Model 3 mass production to now. The current position of the SpaceX supply chain is more like Tesla in 2018, with mass production just starting, suppliers just being finalized, and order growth rates just beginning to steepen. Meanwhile, Starship is still in testing, Starlink is gradually expanding, and the AI data center has not yet started construction; this is equivalent to its 2018.

Finally

Buying SpaceX on its IPO day, I believe, is paying for Musk's dream, and it is a very high-priced space dream. Of course, you could also say you believe in Musk, and that is your dream too.

But perhaps we can look at it from another angle,

Looking along the supply chain, what we are betting on is something else, because no matter how SpaceX's stock price moves, its hundreds of billions of dollars in procurement orders each year must be fulfilled by someone. These orders are unrelated to stock prices; they are revenue that arrives on time every month.

This article does not constitute investment advice. There are still some issues, such as the cyclical nature of beryllium metal, geopolitical discounts for Taiwanese manufacturers, insufficient liquidity for small companies, and potential reshuffling of certifications due to technological iterations. Each company needs to be judged individually.

But if you didn't get a share on the day SpaceX went public,

then you can change your strategy; instead of chasing high prices, let's look at those quietly supplying.

The giant has already ignited; this time, the shovel is within your reach~