Anthropic just crossed $30 billion in annualized revenue. That’s more than OpenAI. That’s more than almost every enterprise software company that existed 15 years ago. And the kicker — the company is reportedly targeting an IPO as early as October at a $380 billion valuation.

OpenAI and SpaceX aren’t far behind. Between the three of them, we’re staring at what could be the largest concentration of IPO wealth creation in modern financial history. Trillions in new equity, funneled through Wall Street, allocated mostly to insiders and institutions before a single retail share ever hits your brokerage.

If you’re reading this thinking “I’ll just buy it on day one,” I’ve got bad news. That’s not how the biggest IPOs work. And the math on why is uncomfortable.

Why this is different from any IPO wave we’ve seen

The dot-com IPOs of 1999 were small — most companies went public at valuations under $5 billion. Today’s AI trio is different. Anthropic alone, at $380B, would IPO larger than Coca-Cola. OpenAI is widely expected to price higher. SpaceX at $400B+ would rival Tesla’s market cap.

When IPOs are this large, three things happen — and all three are bad for retail:

Allocations go to institutions first. The “friends and family” shares are usually gone before retail brokerages even get a crumb.
Pop-day dynamics are brutal. A stock that lists at $100 and opens at $145 has already extracted the easy money. Retail buys the retail price, not the allocation price.
Lockups and dilution come later. Insiders sell, employees cash out, and retail holds the bag through the first earnings cycle that “disappoints expectations.”

We’ve watched this movie before. Facebook, Uber, Airbnb — each one had a retail investor class that bought in at the top and spent 18 months underwater.

The real wealth opportunity isn’t the IPO — it’s the supply chain

Here’s the contrarian take most finance Twitter is missing: the people who made the most money off the last AI wave weren’t the people who bought OpenAI (they couldn’t). They were the people who bought the companies selling to OpenAI.

NVIDIA went from $300 to $1,400+. Companies building AI data centers, cooling infrastructure, and power generation — boring names like Vertiv, Eaton, Constellation Energy — saw 300-500% runs that barely made headlines.

Right now, Anthropic is reportedly evaluating its own AI chips. That’s a signal. The compute supply chain is still the real economy underneath the AI story. And it’s still (relatively) open to retail.

The question isn’t whether to buy the Anthropic IPO at $380B. The question is who Anthropic is going to spend their IPO capital with — and are any of those companies still reasonably valued?

What you can do right now (before the IPO window opens)

Three concrete steps:

Define your AI allocation. If you don’t know what percentage of your portfolio is currently tied to AI (directly or indirectly), you’re already behind. A reasonable number for most investors is somewhere between 10% and 25% — anything higher is speculation, anything lower means you’re betting against the most obvious macro trend of the decade.
Build a “boring infrastructure” basket. Power utilities with AI data center exposure, cooling and electrical specialists, semiconductor equipment. These are where IPO capital flows next — and they’re still buyable.
Wait for the secondary IPO window. Most of the real money on big IPOs is made 6-12 months after listing, when the lockup expires and insiders dump. That’s often when the stock bottoms and the patient buyers win. Don’t chase day one. Set a watchlist now.

The broader truth about AI wealth

The AI IPO wave is going to mint thousands of new millionaires — but almost all of them will be employees, early investors, and institutional allocators. That’s been true for every major IPO wave since the 1980s. What’s different this time is the scale: we’re talking about a combined $1 trillion+ in new public equity across three companies.

Retail investors can still win this decade, but not by trying to get IPO allocations they’ll never get. The game is to own the pipes, not the platforms. Own the compute, the power, the connectivity — the things every AI company has to pay for, whether they IPO or not.

And track your progress. Because the investors who come out of this cycle wealthy won’t be the ones who picked the best stock. They’ll be the ones who knew their number, stuck to their allocation, and didn’t panic when the headlines went crazy.

Want to track your own path through the AI wealth cycle? The Wealtharian Wealth Tracker lets you monitor your net worth, AI allocation, and FU money progress in one place. Try it free →