Three of the most valuable companies in American history are asking investors for their money in the same three-month window. SpaceX goes public on June 12. Anthropic filed its S-1 on June 1 at a valuation approaching $1 trillion. OpenAI is targeting September. The combined implied value of these three companies is around $3.6 trillion – roughly the size of the French economy.
You have $10,000. Where does it go?
This is not financial advice. No reasonable financial advisor would let you treat this article as financial advice. But this is an attempt, drawing on business school research and expert analysis, to think through what a sensible person might actually do.
The IPO race to go public
SpaceX is first. The rocket and Starlink business generated $18.7 billion in revenue last year, up 33% year-on-year. It is offering 555.6 million shares at $135, raising roughly $75 billion, at a valuation of around $1.8 trillion, which would make it the largest IPO in history.
Anthropic filed last week. Its revenue run rate is $47 billion, up from $10 billion a year ago – the kind of growth that makes finance professors reach for their calculators twice. Its last funding round valued it at $965 billion.
OpenAI files in September. It generates roughly $2 billion a month in revenue and is projected to lose $14 billion in 2026. More than half of what it brings in. It has never turned a profit.
Three companies. Three compelling arguments. Three reasons to be worried.
SpaceX: real business, contested valuation, and a contract worth reading carefully
SpaceX’s S-1 revealed something that hadn’t been widely known: the company has become one of the largest AI compute landlords in the world. Google has signed a deal worth $920 million a month for access to roughly 110,000 Nvidia GPUs and associated infrastructure, running through June 2029. Anthropic is paying $1.25 billion a month for access to the Colossus 1 facility through May 2029. Together, these two contracts alone represent a run rate approaching $26 billion a year.
That is a significant number. It is also worth reading the footnotes. Both contracts include 90-day termination clauses that activate after December 31, 2026. What looks like locked-in revenue through 2029 can, in practice, be unwound on three months’ notice from the start of next year.
The underlying rocket and satellite business is more straightforward. Starlink delivered $11.3 billion in revenue in 2025, up 50%, with an operating margin of 39%. The launch business secured $6.45 billion in Space Force contracts last month. SpaceX’s competitive position in rocket launch remains close to unassailable.
The governance picture is less reassuring. Elon Musk retains 85.1% of combined voting power through a dual-class share structure. Buying SpaceX shares means accepting exposure to a $1.8 trillion company with no meaningful influence over how it is run.
On the valuation, there is a notable gap between the IPO price and what independent analysts think the business is worth. Aswath Damodaran, Professor of Finance at NYU’s Stern School of Business and one of the most respected valuation practitioners in the world, has publicly concluded that SpaceX’s equity is worth around $1.3 trillion – roughly $500 billion below the IPO target.
He flagged the S-1’s claim of a $26 trillion total addressable market for its AI division as “reaching the end of what’s plausible and pushing beyond,” and said he was passing on the IPO. Making money off it, he noted, depends on “how good you are at gauging mood and momentum”, and he described himself as “awful at both.” Morningstar goes further: its analysts have assessed SpaceX as overvalued by approximately half, and suggest that patient investors will find better entry points after the listing.
Karthik Krishnan, Associate Professor of Finance at Northeastern University’s D’Amore-McKim School of Business, adds a useful nuance on the retail opportunity: SpaceX is allocating up to 30% of shares to retail investors, which is unusual for an IPO of this scale. But as Krishnan observes, “while the price pop helps build excitement, retail investors will come in and buy it off in the secondary market, so there may actually not be money left on the table.”
Anthropic: the fastest-growing company you can’t quite value
Anthropic’s revenue grew from $10 billion to $47 billion in a single year. That kind of trajectory has not been seen in the technology industry since the early days of cloud computing.
Here is the detail worth noting: $1.25 billion of Anthropic’s monthly outgoings is going directly to SpaceX for compute. Investors considering positions in both companies are, in part, buying into the same ecosystem. If Anthropic invokes its termination clause after December 2026, SpaceX loses its second-largest disclosed contract. If SpaceX’s infrastructure underperforms, Anthropic has the mechanism to exit.
The $965 billion valuation is a bet on Anthropic capturing a significant share of what may become the largest software market in history. That bet has real merit. It also carries substantial downside if the AI market consolidates around the players with the deepest infrastructure investment, and Anthropic is not the biggest spender in the room. Google, Meta, and Microsoft are all investing at a scale that makes Anthropic’s capital requirements look modest by comparison.
Another consideration: is thatAnthropic built Claude, the AI tool that a significant number of people reading this are likely using. Familiarity with a product is a psychologically compelling reason to invest in the company behind it. It is also a notoriously poor one.
OpenAI: the brand name with a complicated path
ChatGPT is a cultural phenomenon. OpenAI has 50 million consumer subscribers and 9 million business users. On pure name recognition, it wins the room.
The financial picture runs deeper than the headline loss figures. OpenAI is projected to lose $14 billion in 2026, and analysts project cumulative losses exceeding $140 billion between now and 2029. Frédéric Fréry, Professor of Management at ESCP Business School, identifies the structural reason: every ChatGPT interaction generates revenue for Microsoft and Nvidia before OpenAI sees a penny.
“Like the Gold Rush era, when shovel sellers grew rich at the expense of prospectors, it is the infrastructure providers making a fortune, not the model designers.” At current burn rates, some analysts suggest insolvency is possible as early as 2027 without a fundamental shift in the business model.
Fréry also points to the intensifying competition. “Google, the inventor of generative AI, is making rapid progress with Gemini. China’s DeepSeek has claimed to use less expensive processors. France’s Mistral AI advocates for a frugal approach and European digital sovereignty.”
Apple recently replaced ChatGPT with Gemini for Siri – a visible sign of how quickly the landscape is reorganising around OpenAI’s position.
The most structurally unusual aspect of this IPO remains what the listing actually represents: a company originally built as a non-profit research laboratory, now converting to a conventional for-profit corporation answerable to public shareholders. That transition has no close precedent in corporate history. Developing AI for the benefit of humanity faces the obligation to grow quarterly earnings.
OpenAI’s September listing at least offers one practical advantage: time to observe how SpaceX and Anthropic perform before committing. That option to wait is worth something.
What the history of big IPOs actually shows
In 2019, Uber listed at $45 per share, raised $8.1 billion, and spent most of its first three years trading below its offering price. In 2021, Rivian raised $13.7 billion at IPO, the largest since Facebook and lost more than 90% of its value within eighteen months. Both were companies with genuine technology, real revenue, and credible long-term stories. Neither was a fraud. The timing, the valuation, and the weight of retail enthusiasm at the moment of listing did the damage.
Jay Ritter, who has studied IPO performance at the University of Florida’s Warrington College of Business for decades, puts a number to the pattern. IPOs underperform companies of comparable size by an average of 3.6% per year in the five years after listing. The first-day “IPO pop” goes almost entirely to institutional investors who receive allocations before the stock opens. The retail investor buying on day one typically pays at or above the IPO price.
None of this makes SpaceX, Anthropic, or OpenAI poor long-term investments. It is a reason to be clearheaded about when you buy.
The $200 billion problem the market hasn’t fully priced in
The entire US IPO market raised $45 billion across all of 2025. These three companies alone are targeting more than $200 billion in the same quarter. Analysts at Yahoo Finance have noted that the 2026 pipeline could raise more capital than all US listings combined since 2022.
Capital does not appear from nowhere. Institutional portfolios rebalancing into SpaceX, Anthropic, and OpenAI will be selling something else to do it. Retail investors making new allocations may pull from existing positions to participate.
The market is at historic highs. A draw of this magnitude on available capital, at this particular moment in the cycle, is a variable that most commentary on these individual IPOs treats as someone else’s problem.
The savings account deserves a fair hearing
The S&P 500 is at historic levels. Cash still earns something. The emotional case for putting $10,000 into some of the most exciting companies of the era is powerful. The historical case is more complicated.
Amazon IPO’d at $18 in 1997. Investors who bought in 1999 at the peak of dot-com enthusiasm paid around $100 per share. By 2001, those shares had fallen below $6. The people who held for twenty years did extraordinarily well. The people who needed their money back in 2003 did not.
All three companies here may prove to be Amazon. The relevant question is whether IPO day at these valuations is the right entry point, or whether allowing the first several months of volatility to play out is the more intelligent position. Damodaran, who values businesses for a living, is waiting. Morningstar’s analysts are waiting. Patience, in IPO investing, has a historically better track record than enthusiasm.
A rough allocation, for those who want one
If you had to put $10,000 to work across these three, a reasonable distribution might look like this:
- half in SpaceX, where the underlying rocket and satellite business is the most defensible, though the valuation gap identified by Damodaran and Morningstar means patience with your entry point matters;
- a quarter in Anthropic, where the growth trajectory is extraordinary and a long horizon absorbs the risk;
- and a cautious position in OpenAI, or none at all, with the remainder held in cash while you watch the first two listings perform.
The capital reallocation effect means the weeks after each listing will be volatile. The opportunity to buy below the IPO price may well present itself. That is not a reason to stay away. It is a reason to wait.
This article is for informational and academic purposes only and does not constitute financial advice. Consult a qualified financial advisor before making investment decisions.
Interested in this topic? You might also like…