The 2026 U.S. IPO market may not just be reopening. It may be asked to do something much harder: price three of the most important private-market stories in the world.
SpaceX is the infrastructure bet. OpenAI is the gateway bet. Anthropic is the enterprise workflow bet. They are not ordinary tech companies, nor are they just another wave of short-term excitement in the IPO market. Together, they may mark the first time public markets are being asked to price, all at once, the defining themes of the next decade: the Space Age, the AGI Age, and the Enterprise Intelligence Age.
But from an investment perspective, the bigger the company, the more dangerous it is to ask only one question: “Is it great?” A great company and a great investment are always separated by one thing: price. The real question is: within these extraordinary valuations, how much is backed by actual performance, and how much is being paid for dreams that may take the next decade to realize?
A simple formula helps frame these mega-IPOs:
Investment Value = Current Performance + Future Dreams × Probability of Execution − Valuation Overhang
This is the core framework. The dream determines the ceiling; the current business determines the floor; the price determines the final return.
Start with the Big Picture
The contrast is pretty clear. SpaceX is the most tangible infrastructure story. OpenAI is the strongest distribution story. Anthropic is the cleanest enterprise adoption story. The problem is that all three already come with valuations that assume a lot has to go right.
SpaceX: The Biggest Dream, But Not Just a Dream
SpaceX is the easiest of the three to describe as a company that “sells dreams.” Rocket launches, reusable rockets, Starlink, satellite internet, the Mars plan, space infrastructure—all these keywords naturally carry that Elon Musk-style sense of the future. This isn't an ordinary business story; it's a narrative about the outward expansion of human civilization.
But saying SpaceX only sells dreams is inaccurate. It actually has the most tangible assets of the three. It has real rocket-launch capability, government and commercial customers, and a highly viable operating business in Starlink. Reuters previously reported that SpaceX was targeting an IPO valuation of around $1.75 trillion, which, if achieved, would make it one of the largest IPOs in history. More recent reports note that SpaceX may reserve 5% of its IPO shares for selected buyers under an unusual lock-up arrangement.
So SpaceX shouldn't be valued simply as a rocket company. Its ultimate valuation is a mix of Starlink monetization, its launch-business moat, and the option value of its space infrastructure—plus a Musk premium, minus a governance discount.
What makes SpaceX especially attractive is that it has both hard performance and a very distant dream. Rocket launches and Starlink are already real businesses that capital markets can easily wrap their heads around. The Mars plan, space infrastructure, and potential future links to AI computing are the longer-dated options.
This is also the core investment tension. Its moat is extremely hard, but its valuation is incredibly difficult to calculate. Rocket reuse, launch costs, satellite networks, and engineering systems cannot be replicated quickly just by throwing money at them. But if SpaceX goes public at a trillion-dollar-plus valuation, investors are buying far more than Starlink and the launch business—they are buying a Musk-led space infrastructure empire.
OpenAI: AGI is the Dream, But the Front Door is the Real Asset
On the surface, OpenAI is selling the AGI dream. But from an investment perspective, its most valuable asset may not be a stronger model—it's the distribution layer.
ChatGPT has already become the first stop for many people when they use AI. Writing, search, coding, office work, learning, investment research—more and more tasks are moving from traditional software into the AI chat interface. Most model companies sell capability. OpenAI is trying to own the front door to AI.
OpenAI officially announced in late March that it had completed a $122 billion funding round, bringing its post-money valuation to $852 billion. This is no longer the valuation of an ordinary AI startup; it's the valuation of a super-platform company. What the market is really buying is a platform poised to become the default access point of the AI era.
To break down OpenAI's valuation, you have to look at ChatGPT's distribution power, its API ecosystem, and its enterprise revenue, all weighed against massive compute costs. The company has already proven user demand and validated its commercialization path. ChatGPT subscriptions, API usage, enterprise customers, and the developer ecosystem show that OpenAI has real revenue paths, not just a technology story.
But OpenAI still needs to prove something much harder: Can the AI interface ultimately become a high-margin platform? AI is not traditional software. Model training, inference, chips, data centers, and compute procurement require massive, ongoing investment. User growth and revenue growth matter, of course, but public markets will eventually ask a more brutal question: How much high-quality profit is actually left over after all those costs?
So the biggest debate around OpenAI is not whether it has a future. The real question is whether it becomes the next Google, Microsoft, or Apple-style platform company—or whether it becomes an AI giant with massive revenue and influence, but with margins under long-term pressure from compute costs and model competition.
Anthropic: The Most Enterprise-Like AI Business, and the Ultimate Valuation Test
Anthropic does not have SpaceX’s civilization-level narrative, nor does it have OpenAI’s mass-consumer mindshare. But among the three, it may be the one that most closely resembles a real enterprise business.
Claude, Claude Code, safety and reliability, enterprise workflows, multi-cloud partnerships—these keywords are not as sexy as Mars or AGI, but they are easier to connect to real paid use cases. Enterprises do need AI. Developers do need coding assistants. Companies are willing to pay for productivity gains. Compared with consumer AI, enterprise AI is more likely to form stable paid demand and become deeply embedded in organizational workflows.
Anthropic officially announced in late May that it had completed a $65 billion Series H round, bringing its post-money valuation to $965 billion. It later announced that it had confidentially submitted a draft S-1 registration statement to the SEC for a proposed IPO.
Anthropic's valuation essentially hinges on enterprise customer growth, Claude Code penetration, workflow stickiness, and a safety/reliability premium, balanced against model commoditization risk. What makes the company attractive to the market is that it doesn't seem to be telling a distant story. It is gradually pushing AI into everyday enterprise work. Code, documents, customer service, analysis, process automation—once these use cases become habits, they can turn into long-term stickiness.
But Anthropic’s risk is also very direct. If model capabilities gradually converge, where will Claude’s long-term differentiation come from? Will enterprise clients ultimately buy the strongest model, or the lowest price? Will they pay for the safety brand, or for cloud integration? Will they buy Claude itself, or the broader solution bundles offered by AWS, Google, and Microsoft?
So the core question for Anthropic is this: Is it a next-generation enterprise software platform, or is it an AI model company that could be repriced under the pressure of model commoditization? Its path to commercialization may be the clearest. The problem is that the valuation already assumes much of that success, leaving investors the least room for error.
Put Them Together: It's Not About Who is Greater, But Who Can Execute More Easily
When you put the three companies side by side, the most interesting question is not which one is greater. The real questions are: Whose dream is bigger? Whose performance is easier for the market to understand? And whose valuation is harder to digest?
If we rank them by narrative power, SpaceX ranks first, OpenAI second, and Anthropic third. SpaceX is about the space age and the outward expansion of human civilization; OpenAI is about AGI and the next generation of human-computer interaction; Anthropic is about enterprise AI and the restructuring of productivity. All three stories are huge, but they operate on entirely different narrative levels.
In terms of performance visibility, OpenAI and Anthropic are easier for the market to understand. They have users, subscriptions, APIs, enterprise clients, coding tools, and cloud partnerships—languages capital markets already know well. SpaceX’s business is harder, but also heavier. Rockets, satellites, launches, communication networks, and future space infrastructure all imply higher capital expenditure, more complex regulation, and longer return cycles.
If we map them on a two-dimensional chart, it would look roughly like this:
That is why they should not be treated as the same kind of IPO. SpaceX is a long-duration infrastructure option on the space age. OpenAI is a platform-layer bet on the next interface shift. Anthropic is an enterprise productivity bet on AI becoming embedded in daily workflows.
$NVIDIA(NVDA)$ $Tesla Motors(TSLA)$ $Meta Platforms, Inc.(META)$ $Alphabet(GOOGL)$ $Alphabet(GOOG)$ $Apple(AAPL)$ $Microsoft(MSFT)$ $Amazon.com(AMZN)$ $C3.ai, Inc.(AI)$
Comments