What SpaceX's IPO Revealed About Modern Market Mechanics
The SpaceX public debut offers a striking lesson in how IPO mechanics can be engineered to support a share price independently of company fundamentals. After months of detailed analysis — including a comprehensive report on how the company operates — the most revealing fact about the deal is the size of its float. A study of the last roughly 100 IPOs of any size over nearly the past year shows that a typical offering releases about 21% of its shares into the public float. SpaceX released just 4.3%.
That deliberately tiny float was designed to "juice the stock." By offering so few shares, the company created scarcity. This scarcity was compounded by index dynamics: many indexes were compelled to buy the stock in the aftermarket if they couldn't secure allocations during the offering. The result was a flood of forced buyers chasing a very small supply of shares.
This matters because SpaceX is, by any conventional measure, an extraordinarily expensive deal — trading at over 100 times earnings, or, framed more charitably, over 100 times revenues. The fact that the stock has held its price at all, currently trading only a few dollars above its IPO price, is described as something close to a miracle. The key takeaway: the price stability speaks far more to the manufactured market dynamics of the deal than to the underlying performance of the business.
How This Changes the Calculus for Future IPOs
If a company can support its valuation by offering only a sliver of its shares, it raises an obvious question: why would any company choose to offer more? The logic of the small float becomes self-reinforcing, especially if the rules governing lockups are also set to change.
Two additional developments around SpaceX underscore this shift. First, insiders are positioned to sell far more easily than ever before. Goldman Sachs has effectively released the people who would normally be held back. Under a conventional lockup, insiders must wait six months before selling. In SpaceX's case, insiders are expected to sell right after the quarterly earnings report — and to sell in large volumes immediately thereafter.
Second, the scale of incoming supply is dramatic. The initial offering was 555 million shares (a figure full of fives — 555 million, 555,000, 555,000). But before the six-month lockup period even concludes, over 1.2 billion additional shares are expected to come to market. So while the float started tiny, a substantial wave of new supply is approaching quickly.
Why Anthropic Could Be the Defining Debut of the AI Era
Anthropic stands out as potentially the most important debut of the AI era. By a wide range of measures, it is arguably the number one AI company in existence.
Consider the competition. One could argue that Nvidia, a semiconductor company, is not truly an "AI company" — and there is history to support that skepticism. Nvidia was a Bitcoin mining beneficiary only a few years ago, and before that spent decades making graphics processing units for video games. By contrast, Anthropic is a pure large language model business: its product is offered, widely used, and — from what can be gathered — has achieved some profitability.
Anthropic is described as the biggest company of its kind and the standard against which everyone else is measured, including ChatGPT and OpenAI. It is growing at an incredible pace. In Silicon Valley, the constant topic of conversation is using Claude Code and using Anthropic. As a pure AI offering, it is positioned to be the best of breed.
A pointed contrast is drawn with other companies that wrap themselves in AI narratives. SpaceX, for instance, has an XAI unit framed as artificial intelligence — but it isn't really a pure AI play. Its principal revenues over the past year come from Twitter housed within that AI unit. Anthropic, by comparison, is a genuine pure-play AI company.
The Missing Exposure for U.S. Investors
What makes Anthropic especially significant is that U.S. investors have not previously had access to a pure-play AI LLM company of this kind. Chinese investors have had considerably more experience with domestic pure-play AI and LLM companies, some of which have seen explosive, "absolutely bonkers" price action. An Anthropic listing would, for the first time, give American public-market investors direct exposure to this category.
Capital Rotation: Trading Dynamics Versus Company Fundamentals
A crucial distinction runs throughout: the difference between trading dynamics and company fundamentals. In the lead-up to the SpaceX IPO, capital visibly rotated out of momentum trades, out of Bitcoin, and out of riskier corners of the market to fund participation in the new offering. Will a similar pattern repeat with Anthropic?
Conversations with traders on the floor of the New York Stock Exchange suggest it might. Stocks in sectors like photonics and optical networking — names that have become popular among retail traders and the broader trading class — appear to have been sold down so investors could free up money for the SpaceX IPO. Now, as those investors exit SpaceX positions, money may flow back into those same names.
The critical point is that these rotations have nothing to do with the fundamentals of the underlying companies and everything to do with the mechanical aspects of the market at a given moment. These are two entirely different things. Inside the trading room, prices are treated simply as names; some traders have spent years trading companies without knowing what those companies actually do. But an investor, as opposed to a trader, should focus on what a company does and what its prospects are. It is essential to understand what lies behind a moving share price rather than reacting only to the "wiggly line on the screen."
A New Playbook for Mega-Cap Private Companies
If Anthropic follows the SpaceX blueprint — a float below 5% — it could fundamentally reshape how future AI companies approach going public, with OpenAI being the obvious next candidate among the mega-names. In effect, the script is being rewritten in real time.
This represents a genuinely new playbook, and Databricks belongs in the same conversation. These companies share a defining trait: they have remained private far longer than was historically typical in Silicon Valley. They have raised enormous amounts of capital, progressing through D, E, and F funding rounds. In the past, a B round and a C round were often all a company would complete before an IPO.
The nature of these companies has also inverted. Historically, the companies pushed toward an IPO were often the desperate ones — businesses not strong enough to go public cleanly but still able to raise venture money. Venture capitalists would then force them into the public markets simply to recover their investment, "putting lipstick on those pigs" in the process. (That colorful phrase, it's noted, originated on the NYSE floor itself, where "there were a lot of pigs running around.")
Today's crop is the opposite: these are very large, successful companies with substantial revenues. Some may have "some hair on them," but they carry big revenues and command enormous market capitalizations in the private markets — valuations they are likely to maintain once public. SpaceX, Anthropic, OpenAI, and Databricks are all potentially trillion-dollar companies upon hitting the public markets, and they may sustain those valuations at IPO — something never seen before.
The Regulatory Backdrop and the Scale of Modern Listings
The talent migration reinforces the trend: a recent Bloomberg report indicated that even more executives are leaving Google's parent, Alphabet, to join Anthropic and OpenAI — a telling signal of where the center of gravity now sits.
The sheer scale of recent deals has also become normalized. SK Hynix's ADR listing, filed via an F-1 (the foreign equivalent of the S-1 used by domestic issuers) and slated for July 10th, is a $29 billion offering. For perspective, Alibaba's landmark IPO was $25 billion. These once-staggering figures now pass as routine.
Behind the move toward staying private longer lies a specific regulatory change. There used to be a rule tied to share count: if a company exceeded a certain number of shareholders — perhaps around a thousand, possibly a hundred in an earlier era — it was forced to go public. (The exact threshold would need fact-checking.) That rule effectively compelled companies to list once they had accumulated many investors. When it was removed, companies gained the freedom to remain private far longer.
That rule change is judged, on balance, not to have been a terrible one. There is a strong argument that companies should be well seasoned before the average retail investor gets involved — particularly given that indexes can force everyone to own these shares whether they want to or not. The consequence has been longer timelines before companies come public, but the markets are currently well equipped to absorb these very large IPOs.
Conclusion
Taken together, these developments mark a structural shift in how the largest private companies enter public markets. The combination of ultra-small floats, looser lockup behavior, sustained mega-cap valuations, and a regulatory environment that lets companies mature in private has produced a new kind of IPO — one where mechanics can dominate fundamentals in the short term, where trillion-dollar debuts are plausible from day one, and where a pure-play AI leader like Anthropic could become the defining listing of its generation. For anyone watching, the essential discipline remains separating the engineered movement of share prices from the real prospects of the businesses underneath them.
