Elon Musk’s space company SpaceX will list on Nasdaq on the 12th, with a valuation of $1.8 trillion, or about 2,740 trillion won. It is the largest initial public offering in history, shaking up Rocket Lab’s long-held position as the crown jewel of space investing.
The offering price is $135 per share, and the company will issue 555.6 million shares. The listing will raise about $75 billion, or roughly 114 trillion won. It will break the IPO record set by Saudi Aramco in 2019. A rare feature of this deal is that 30% of the shares are being allocated to retail investors. In South Korea, money has also been flowing into the space theme, with an ETF named after NASA drawing $2.6 billion in inflows over the past two months.
◆ A $1.8 trillion valuation, a shock at 96 times revenue
SpaceX posted revenue of $18.7 billion last year, or about 28 trillion won. Dividing its $1.8 trillion valuation by that revenue gives it a price-to-sales ratio of 96 times. In other words, an investor would be paying 96 times the amount the company earns in a year.
Consider Rocket Lab, the comparison company. As of the previous day’s close on the 9th, its share price was $119 and its market capitalization was $65.8 billion, or about 100 trillion won. Although its stock has surged nearly 300% over the past year and is the hottest among space stocks, it looks cheap by comparison with SpaceX’s 96 times.
On the surface, SpaceX seems absurdly expensive. But if the revenue calculation is shifted forward by one year, that impression begins to change.
◆ AI leasing contracts change the denominator
The key is artificial intelligence. SpaceX absorbed Musk’s AI company xAI in February this year, transforming itself into a company that spans rockets, satellites, and massive data centers.
Two major contracts expanded the scale of the business. Anthropic agreed to lease an entire data center in Memphis, Tennessee, paying $125 million per month. Google will pay $92 million per month starting in October for access to 110,000 Nvidia GPUs. Together, the two contracts amount to $26 billion a year, more than SpaceX earned from all businesses last year.
If that leasing income is reflected in this year’s revenue, the calculation changes. Assuming last year’s revenue of $18.7 billion remains unchanged and adding five months of Anthropic revenue this year, or $6.25 billion, plus three months of Google revenue, or $2.76 billion, estimated revenue rises to $27.7 billion, or about 42 trillion won. Dividing $1.8 trillion by that revenue gives a P/S ratio of 65 times. If revenue reaches $30 billion, the multiple falls further to 60 times.
Rocket Lab’s estimated P/S ratio for this year is 52 times, according to FactSet consensus. The wide gap between 96 times and 52 times narrows to the slimmer difference between 65 times and 52 times. That means SpaceX’s seemingly expensive valuation is actually not very different from its peers.
At a deeper level, this raises a more fundamental question. SpaceX’s filing says it is targeting a market worth $28.5 trillion, and most of that is expected to come not from space but from enterprise AI. Investors are effectively deciding whether they are buying a rocket company or an AI infrastructure company.
◆ Similar multiples do not mean similar risk
Stopping at the multiple comparison would mean reading only half the story. Even with the same P/S ratio, the quality of the underlying revenue is completely different.
Rocket Lab is a pure space company that earns money from launch services and satellite manufacturing. Its first-quarter revenue was $200 million, and its backlog has exceeded $2.2 billion. Its next-generation rocket, Neutron, has already secured five launch contracts before its first flight. The company’s fate rests on the success of that rocket.
SpaceX’s lower multiple is built on a different foundation. The AI leasing contracts support the denominator, but those contracts include escape clauses. Google can terminate the deal with 90 days’ notice after year-end, and it also has the right to walk away if GPU deliveries are delayed.
The AI business is still operating at a loss. It posted more than $2.4 billion in operating losses in the first quarter alone. Independent assessments are also flashing warning signs. Morningstar estimates its fair value on a cash-flow basis at $780 billion, less than half of the IPO valuation.
Of course, there is a counterargument. Goldman Sachs expects SpaceX’s AI revenue to exceed $300 billion by 2030. If the leasing business performs as expected, then a 65 times multiple today could actually be cheap. Bullish and cautious views are drawing completely opposite conclusions from the same number.
This is the point investors should focus on. More important than a P/S ratio of 65 times is how long the revenue supporting that denominator can last. Rather than chasing headlines because the multiple looks lower, it is safer to ask what exactly is being bought. Since 30% of the shares are being offered to individual investors, the impact of early volatility will fall even more heavily on retail traders.
The valuations of SpaceX and Rocket Lab are now overlapping at one point. After the listing, the market will decide whether to assign a premium to SpaceX’s overwhelming scale or a discount to the uncertainty around its AI contracts. The gap between the two is likely to widen again depending on that verdict.
If Rocket Lab’s fate depends on Neutron, SpaceX’s fate depends on the durability of its AI revenue. The real battle between the two space stocks begins after the opening bell rings.