100 billion ARR goes hand in hand with 18.4 billion dollars: SpaceX (SPCX.US)'s first quarterly report clashes “growth and cost”

The Zhitong Finance App learned that after the US stock market on August 4, SpaceX (SPCX.US) handed over its first quarterly report since listing. Despite a sharp doubling of revenue in the second quarter and strong guidance from management to hit 100 billion US dollars of ARR at the end of the year, the company's stock price dived more than 9% after the market. Wall Street is re-evaluating the risks and rewards of this new “connection+computing power” infrastructure giant amid the combination of explosive performance and surging capital expenditure.

Starlink hematopoiesis, AI runs wild, and accumulates energy in space

SpaceX's business layout consists of three segments: Connectivity (Starlink), AI, and Aerospace.

The connectivity business is a well-deserved “basic profit market.” The connectivity business revenue for the second quarter was US$4.29 billion, up 66% year on year, and operating profit of US$1.66 billion. It was the only profitable business in the three major sectors.

The number of Starlink global users reached 12 million, doubling year on year, with a net increase of 1.7 million households in a single quarter; ARPU stabilized at 66 US dollars. Although consumer ARPU declined from 85 US dollars, corporate and government revenue surged 108% year over year to 1.8 billion US dollars, effectively hedging the decline in consumer unit prices.

In addition, the company has obtained a multi-year US government contract of more than 6 billion US dollars for the Starshield project, and the aviation scene is also being implemented at an accelerated pace — United Airlines, British Airways, Southwest Airlines, etc. have all launched Starlink Wi-Fi services.

Furthermore, in terms of Starlink Mobile (Starlink Mobile), President and Chief Operating Officer Shortwell pointed out that the recently approved EchoStar band by the FCC has built a fundamental competitive advantage for the mobile business; the next generation of mobile satellites will be launched next year, and commercial services will be officially provided to terminals by the end of next year.

The AI business is the fastest growing engine. The AI segment's revenue for the second quarter was US$2,561 billion, up 247% year over year. More importantly, AI operating losses narrowed sharply from US$2.47 billion in the previous quarter to US$1.26 billion, far better than market expectations of US$2.39 billion.

The loss reduction itself is a strong sign — indicating that AI computing power infrastructure is moving from a “pure investment” stage to large-scale commercial monetization. Musk further revealed that with the explosion of demand for commercial cloud services, the share of computing power used for Grok model training will drop to about 10% in the future, and the vast majority of computing power will be exported as infrastructure to achieve efficient monetization.

The space sector is relatively lackluster. Revenue for the second quarter was US$962 million, up 29% year-on-year, and operating losses of US$542 million. However, the progress of the Starship V3 is the real focus of the market: the 13th test flight in July successfully deployed 20 mass-produced V3 satellites. Musk revealed that the 14th test flight will be carried out as soon as the end of August, and will try to use towers to recover the upper stage — this will be another landmark leap forward for reusable rockets.

Musk said that the breakthrough upgrade of the Starlink service relies on critical mass formed by at least 1,000 V3 satellites, and this node is expected to arrive in the second quarter of next year. And with the maturity of key technologies such as heat shields, it is expected that after a year, Starship will be able to launch ultra-high frequency at least once a day (1 flight a day).

100 billion ARR: transition from goal to “bottom line”

What shocked the market most was not the results that had occurred, but the forward-looking guidance given by management.

CFO Johnson said in a conference call that based on the $6.7 billion new cloud service contract signed in the first few weeks of the third quarter (gradually realized from October) and the contribution of Cursor's acquisition, the company is expected to achieve 100 billion US dollars in annualized recurring revenue (ARR) by the end of 2026.

Musk's statement was more aggressive: “Reaching $100 billion ARR in December is not a question; we can achieve it even if we don't do anything.” He also moved the internal forecast for the trillion-dollar revenue target from before 2031 to 2030.

Behind this guidance, computing power is the core support. Musk made it clear that SpaceX exclusively uses Nvidia GPUs in chip supply. The computing power is expected to exceed 2 GW by the end of the year, and it is expected to reach 5-10 GW by the end of next year. The CFO further revealed that the payback cycle for the new computing power deployment was less than a year — which explains why management is daring to continue increasing the current high level of capital expenditure.

The “double-edged sword” of capital expenditure

The other side of high growth is the high cost of capital. Of the capital expenditure of 18.4 billion US dollars in the second quarter, about 15.8 billion US dollars was invested in AI computing power infrastructure. The CFO made it clear that capital expenditure will remain at a similar level for the next two quarters.

The market's concerns are not unfounded: SpaceX's stock price dropped from a high of $225 to around $114 after listing, and investors are weighing a proposition. When a company has both the “sexiest growth story” and “the most amazing speed of burning money,” which side of the valuation scale should it trend?

What optimists see is that the connectivity business has stabilized hematopoiesis, the AI business has reached an inflection point of loss reduction, and the 100 billion ARR guideline means that the revenue scale will increase several times more within half a year. What pessimists see is the erosion of free cash flow by continued high capital expenditure, supply pressure brought about by the lifting of the ban on shares by insiders, and the subtle temperature gap between Musk's aggressive predictions and management's cautious stance.

SpaceX's first quarterly report tells the story of “three in parallel”. Starlink proved the maturity of the business model with 12 million users and stable profits, AI verified the rigidity of demand for computing power leasing with 247% growth rate and drastic loss reduction, and Starship showed the viability of the technology route with successful V3 test flights and the goal of flying once a day.

The 100 billion dollar ARR guide means SpaceX is transforming from a “rocket company” to an infrastructure giant with “connection+computing power.” But for investors, the question that really needs to be answered may be: When a company's narrative switches from “burning money for growth” to “cashing out profits,” how many times the market is willing to pay a valuation premium?