SpaceX and Tesla's "Ecological Chemical Reaction": Merger Discussions and AI Valuation Logic
At SpaceX's 2026 Q3 earnings call, Musk confirmed that SpaceX and Tesla have each established independent board special committees to explore a merger, and he favors Tesla as the surviving entity absorbing SpaceX through a merger. After acquiring xAI, SpaceX completed a $75 billion IPO, with its market value once exceeding $2 trillion; in the first three quarters of 2026, its AI business turned profitable, with operating profit of $705 million, but AI-related capital expenditures in the first three quarters of 2026 had already reached $40 billion. Tesla delivered 1.325 million vehicles in the first three quarters of 2026, only returning to 2023 levels, with a Q3 net profit margin of 2.9%, making it difficult to support robot and autonomous vehicle R&D.
Body · AI translation
[[**SpaceX completed a $75 billion IPO through its acquisition of xAI, with market cap briefly exceeding $2 trillion, but AI investment is heavy and profitability uncertain; Tesla's auto profits are thin, the two sides are exploring a merger, and Musk favors Tesla leading the "rescue."**
**Key Points**
**1. SpaceX completes a third closed loop via xAI and pushes up its valuation**
Although the combined launch + connectivity simulated operating profit has turned positive, the valuation ceiling is about $200 billion; after acquiring xAI, the IPO raised $75 billion, and market cap briefly exceeded $2 trillion.
**2. xAI brings high growth but also cash burn and closed-loop risks**
In 2025, AI operating loss was $6.36 billion, and in the first three quarters of 2026 it turned around to a profit of $705 million; however, capital expenditure surged, space computing costs are 3–4 times those on the ground, and contracts can be canceled.
**3. Tesla's auto business cannot hold up alone**
Deliveries stagnated after peaking, and in the first three quarters of 2026 they only returned to 2023 levels; net profit margin fell to 2%–4%, making it hard to support robot and robotaxi R&D, and institutions are bearish.
**4. Musk favors Tesla as the main entity to merge with SpaceX**
On October 20, 2026, he confirmed that the two sides had set up independent board special committees to explore a merger; using Tesla as the main entity can maximize the benefits of the "trillion-dollar compensation plan."
English text is being prepared.
English text is being prepared.
Author|Eastland
Header image|Visual China
Previously, Tesla held the second-generation Roadster launch event at SpaceX's test base in Texas. Observant people can see that the so-called "flying car" is a product of aerospace technology being handed down to the Tesla platform.
It is said that the flying Roadster carries a 600-liter high-pressure gas cylinder (containing 250 kilograms of nitrogen) and can only spray for a few seconds. Musk released this white elephant to pave the way for a merger between SpaceX and Tesla.
SpaceX's triple closed loop
SpaceX was born to colonize Mars, and its main business is the R&D and construction of rockets and spacecraft.
Rome was not built in a day, and it is even less likely that humans can directly land on Mars. SpaceX obtains funds through commercial launch services to feed the Starship program. This is SpaceX's first closed-loop plan.
Although the gross profit margin of launch services is high, R&D investment is large, and net cash flow from operating activities is limited, unable to support the long-term, high-intensity R&D of the Starship program.
After mastering reusable rocket technology, SpaceX uses low-cost, high-frequency launches of low-orbit satellites to form Starlink, occupying scarce orbital and spectrum resources, and providing broadband access services to users worldwide to obtain cash flow to support Starship R&D. This is SpaceX's second closed-loop plan.
Years of practice have proven that launch services fluctuate between profit and loss while Starlink benefits are considerable. Combined, the two stably generate several billion dollars in cash flow each year. Surviving is no problem, but the dream of achieving Starship is still far away.
The logic of SpaceX's acquisition of xAI is completely different from the Starlink business: Starlink is a cash flow amplifier, while xAI is a valuation amplifier. Using the capital market's frenzied pursuit of AI concepts, SpaceX is being built into the "first stock of sky computing" to raise money in the capital market. This is SpaceX's third closed-loop plan.
Thus, SpaceX has three major main businesses—launch, connectivity, and AI. The prospectus discloses the revenue structure since 2023:
In 2023, revenue from launch, connectivity, and AI was $3.55 billion, $3.87 billion, and $1.96 billion, respectively, totaling $10.4 billion; in 2024, revenue from launch, connectivity, and AI was $3.8 billion, $7.5 billion, and $2.62 billion, respectively, totaling $14 billion; in 2025, revenue from launch, connectivity, and AI was $4.09 billion, $11.4 billion, and $3.2 billion, respectively, totaling $18.7 billion; in the first three quarters of 2026, launch revenue was $3.35 billion, up 2.2% year on year; connectivity revenue was $12.8 billion, up 56.8% year on year; AI revenue was $10.94 billion, up 380% year on year, due to the acquisition of xAI rather than natural growth of the original business.
From 2023 to 2025, the proportion of the three main businesses in revenue changed greatly:
Launch business revenue grew 11.5%, with an average annual growth of 7.2%; its share of total revenue fell from 34.2% in 2023 to 21.9% in 2025;
Connectivity business revenue grew 194%, with an average annual growth of 71.6%; its share of total revenue rose from 37.2% in 2023 to 6.1% in 2025;
AI business revenue grew 8.1%, with an average annual growth of 4%; its share of total revenue fell from 28.5% in 2023 to 17.1% in 2025;
"Launch + connectivity" was already a closed loop
Rocket launch is the business SpaceX started with, and its technology and scale are second to none globally, with gross profit repeatedly hitting new highs.
Unfortunately, rocket R&D is a bottomless pit, and the launch business cannot even earn back R&D expenses:
In 2023, gross profit was $1.89 billion and R&D expenses were $1.54 billion;
In 2024, gross profit rose to $2.26 billion, and R&D expenses also rose to $1.84 billion;
In 2025, gross profit was $2.73 billion and R&D expenses were $3 billion (surpassing gross profit);
In the first three quarters of 2026, gross profit was $2.3 billion and R&D expenses were $3.23 billion, $930 million higher than gross profit.
Launch cannot earn back Starship R&D costs, so the connectivity business fills the gap:
In 2025, gross profit was $5.47 billion, with a gross margin of 48%; R&D and selling/administrative expenses totaled $1.04 billion, for a total expense ratio of only 9.2%; in the first three quarters of 2026, gross profit was $6.72 billion, with a gross margin of 52.5%; R&D and selling/administrative expenses totaled $1.51 billion, for a total expense ratio of 11.8%;
After merging with the connectivity business, SpaceX's performance immediately returned to normal. The chart below shows simulated gross profit, expenses, and operating profit for "launch + connectivity":
In 2023, gross profit was $3 billion, total expenses were $2.5 billion, and simulated operating profit was $470 million; in 2024, gross profit was $5.1 billion, total expenses were $3 billion, and simulated operating profit was $2.09 billion; in 2025, gross profit was $8.2 billion, total expenses were $3.8 billion, and simulated operating profit was $4.4 billion; in the first three quarters of 2026, gross profit was $9 billion, total expenses were $5 billion, and simulated operating profit was $4 billion;
If there were only the launch and connectivity businesses, SpaceX would already have sustained profitability. But this is a low-level closed loop, mainly for two reasons:
First, for the grand narrative of "colonizing Mars," a surplus of a few billion dollars a year is a drop in the bucket.
Second, with $5-6 billion in operating profit from "launch + connectivity," a valuation at 30x PE would not exceed $200 billion, and the proud Musk would not be satisfied. More critically, an IPO would not raise much money either.
In early 2026, SpaceX acquired xAI (including X platform, Grok, computing power leasing, and businesses such as X platform, Grok, Colossus computing power leasing, and Cursor).
SpaceX was implanted with the hot AI concept, successfully completed an IPO, raised $75 billion, and its market value once exceeded $2 trillion. The third closed loop achieved phased success.
xAI—a bull in a china shop
1) Whether SpaceX makes money all depends on AI's fortunes
Injecting the AI business made SpaceX a new darling of Wall Street, at the cost of a completely transformed revenue and profit landscape.
In 2025, gross profit was $1.02 billion, with a gross margin of 32%; R&D expenses were $5.06 billion, for an expense ratio of 158%; selling/administrative expenses were $1.83 billion, for an expense ratio of 57.1%; the total expense ratio was 215.3%;
In Q1 2026, the gross margin increased to 44.3%, and the R&D expense ratio climbed to 291%;
In Q2 2026, computing power leasing revenue from major customers such as Anthropic and Google surged, revenue grew 213% quarter-on-quarter, while R&D expenses instead fell 8.4%, causing the R&D expense ratio to drop to 85%;
In Q3 2026, revenue rose another 195% quarter-on-quarter to $7.56 billion; R&D expenses grew 46% quarter-on-quarter to $4.89 billion; because R&D expenses grew less than revenue, the R&D expense ratio further dropped to 42%.
In 2025, the AI business had an operating loss of as much as $6.36 billion; in the first three quarters of 2026, the AI business turned profitable, with operating profit of $705 million.
SpaceX's performance depends entirely on xAI:
In 2025, gross profit was $9.22 billion. Of that, xAI contributed $1.02 billion;
Total expenses were $11.29 billion, of which xAI's total expenses were $6.89 billion.
Ultimately, SpaceX had an operating loss of $2.59 billion and a net loss of $4.94 billion.
Entering 2026, xAI's performance improved rapidly, and by the third quarter SpaceX's gross profit reached $9.1 billion (of which xAI's gross profit was $4.9 billion); total expenses were $6.3 billion (xAI's total expenses were $4.45 billion).
Because the gross profit contributed by xAI exceeded the expenses it consumed, SpaceX's quarterly operating profit was $2.8 billion.
But the AI business is a cash-devouring beast. In the first three quarters of 2026, SpaceX's AI-related capital expenditures had already reached $40 billion. The bulk is still to come; according to Goldman Sachs estimates, SpaceX's capital investment before 2030 will reach $350 billion, of which 80% will go to AI computing power.
2) Huge uncertainty
Whether SpaceX's AI business can form a closed loop in the future is highly uncertain.
First, space computing power costs are too high and lack competitiveness.
Based on public information, the cost of space computing power per GW exceeds $100 billion, with annual depreciation of $20 billion (over a 5-year depreciation period), 3-4 times that of U.S. ground-based computing centers.
The claim of "zero-cost electricity in space" does not hold either. Because the sun can shine for free, the cost of photovoltaics is fixed asset depreciation. The utilization hours and light intensity of space photovoltaics are far higher than ground-based photovoltaics, but manufacturing and installation costs are also substantially higher. After calculating depreciation, it is really hard to say which has a lower cost per kWh, space photovoltaics or ground photovoltaics.
SpaceX has no cost advantage compared with U.S. ground-based computing centers, and there is no comparison with China (at the Mengxi computing center, the annual "depreciation + electricity" for GW computing power is only 10 billion RMB).
Second, contractual revenue may not be 100% secured.
xAI is highly dependent on computing power leasing contracts with several major customers. Contracts between large U.S. companies often come with cancellation clauses—who wouldn't want to leave themselves a way out?
As for computing power leasing, it usually includes a near-term non-cancellable portion (within 6–12 months) and a far-term cancellable portion (either party may terminate with 90 days' written notice).
So far, SpaceX's incorporation of xAI has been successful, but whether a true closed loop can be achieved in the future is highly uncertain.
Tesla's sales peaked at 1.81 million vehicles in 2023, then fell continuously:
In 2024, 1.79 million vehicles delivered, sales index 99 (with 2023 as the base);
In 2025, 1.636 million vehicles delivered, sales index 90;
In the first three quarters of 2026, 1.325 million vehicles delivered, sales index 100.
From January to September 2026, sales rebounded, but only back to the 2023 level, lacking room for imagination.
Tesla's per-vehicle profit and gross margin stabilized after several consecutive years of decline:
In 2022, the average selling price exceeded $50,000 and the gross margin was 26%;
In 2025, the average price fell to $40,000 and the gross margin was only 13%;
In 2026, both the selling price and gross margin recovered; in the third quarter, the average selling price was $43,000 and the gross margin returned to 20%. Gross profit per vehicle was about $8,300.
Net profit fell continuously after peaking in 2023:
In 2023, Tesla's net profit reached $15 billion, with a profit margin of 15.5%;
In 2024, net profit halved to $7.2 billion, with a profit margin of 7.3%;
In 2025, net profit halved again to $3.9 billion, with a profit margin of 4.1%.
Entering 2026, the net profit margin declined further: Q1 net profit was $490 million, with a margin of 2.2%; Q2 net profit was $1.1 billion, with a margin of 4%; Q3 net profit was $900 million, with a margin of 2.9%.
The car-making business has meager profits, with annualized profits of three to four billion dollars, which is a drop in the bucket for robot and robotaxi R&D and large capital expenditures, and Tesla may fall into losses.
SpaceX's new mission—coming to the rescue
On November 7, 2025, Tesla shareholders approved Musk's "trillion-dollar compensation plan." Musk is to lead Tesla to accomplish multiple ambitious goals. For example, for every $500 billion increase in Tesla's market capitalization, reward shares worth $10 billion are unlocked; there are also goals such as raising annual adjusted EBITDA to $400 billion, delivering 1 million humanoid robots, and deploying 1 million robotaxis.
All of Tesla's profit comes from vehicle manufacturing, where the net profit margin is less than 3%. The valuation built on the "light of technology" persona is precarious, and Wall Street is collectively bearish:
Investors have become jittery. Muddy Waters released an 87-page bearish report (directly pointing to false advertising of autonomous driving, questionable financial data, and severely inflated valuation), and Tesla's stock price plunged 17.8% that week.
As early as the end of 2025, the author raised the possibility and necessity of a merger between Tesla and SpaceX. Since then, discussions on this topic have increased. On October 20, 2026, Musk personally confirmed on SpaceX's Q3 earnings call that the two companies had each established independent special committees of the board to explore a merger.
Technically, there are multiple structures available for the merger of the two companies, such as Tesla absorbing and merging SpaceX, SpaceX injecting some assets into Tesla, SpaceX absorbing and merging Tesla, and establishing a new company to actually control Tesla and SpaceX...
The "trillion-dollar compensation plan" approved by Tesla shareholders is a rare opportunity for Musk. Therefore, only a merger with Tesla as the main entity can maximize his personal interests (such as the first and second options above).
The story of ecological synergy between SpaceX and Tesla is already taking shape. For example, relying on xAI computing power to train FSD, SpaceX purchasing robots, xAI being a major customer of Tesla energy storage, Starlink being deeply tied to robotaxis... There is also a joint venture project called Terafab (SpaceX + Tesla + Nvidia + TSMC) said to involve total investment of $119 billion.
Coming to the rescue (saving Musk, saving Tesla) has become SpaceX's new mission.
*The above analysis is for reference only and does not constitute any investment advice!
Article title: The "ecological synergy" of SpaceX and Tesla
Has published 1,120 articles on Huxiu. This content may not be reproduced without permission. For authorization, please contact hezuo@huxiu.com. Everyone who is changing and wants to change the world is on the Huxiu APP[ ]