Industry overview

Anthropic, the company behind the Claude AI models, is expected to go public next month, and bankers have floated valuations of around $1.5 trillion — with some reaching as high as $2 trillion. That would make it one of the most valuable companies on Earth from its first day of trading. Eric Fry cautions against rushing to buy in, and his warning starts with an unlikely analogy: single-number roulette.

In roulette, a player picks a number from 1 to 36 and hopes the ball lands on it, for a 35-to-1 payout. The odds look almost fair until you count the two green slots, “0” and “00,” which belong to the house. Those extra slots give the casino an edge on every spin, and the longer you play, the harder it is to overcome — which is why novelist Mario Puzo, of The Godfather fame, once said, “If you are looking to cheat the house or beat it legitimately, forget about roulette. It’s basically a sucker’s game.”

Fry’s point is that even correctly identifying the next revolutionary company does not guarantee a good investment — the price paid matters. In December 2000, three Wharton School professors studied 48 publicly traded e-commerce companies that had captured investors’ imaginations during the dot-com boom. The results were grim: six had filed for bankruptcy, fifteen were trading below $1, Amazon was down 90% from its peak, and only one — 1-800-contacts — was trading above its IPO price. One winner out of 48.

The picture was worse among private dot-coms. A separate Wharton analysis counted roughly 1,100 reasonably large e-commerce businesses operating during the bubble; only 20 were still around by 2015, and two — Amazon and eBay (EBAY) — had come to dominate the industry. The internet did change the world as its promoters predicted, but the overwhelming majority of the companies chasing that opportunity disappeared.

Anthropic is no Pets.com, Fry concedes: its revenue reached $16.8 billion in the third quarter, up from almost nothing two years earlier, and it has a legitimate shot at becoming one of the dominant technology companies of the next several decades. But that says nothing about whether the stock will be a good investment. Most people forget how modest Amazon’s valuation was even at the height of the dot-com boom: adjusted for inflation, its market value peaked at $36 billion in December 1999 — roughly what Dassault Systèmes (DASTY), a French software company, or Okta (OKTA), a San Francisco cybersecurity company, is worth today. That left room for 80-fold gains even for investors who bought at the very top; Amazon is now worth almost $3 trillion.

At a $36 billion IPO valuation, Fry says he would be buying Anthropic eagerly, and he might take a chance even at $100 billion. But the company is instead going public at $1.5 trillion to $2 trillion. Repeating Amazon’s 80-fold return from there would require a valuation of $120 trillion to $160 trillion. That $160 trillion figure is twice the value of the entire U.S. stock market today, twice the value of all real estate in America — housing, offices and farmland — and more than the combined net worth of every American household. The company could succeed beyond anyone’s expectations and still leave its investors with disappointing returns, because a price that already assumes a spectacular future leaves almost no room for anything less. At least roulette guarantees a 35-to-1 payout if you win.

Fry’s approach is to look where most investors are not. The lesson he took from the dot-com boom was not to avoid technology stocks, but to avoid paying prices that assume a company’s wildest dreams come true. In 2000 he looked past the obvious internet winners: Barron’s published his warning about Corning (GLW), one of the boom’s hottest stocks, which fell 87% within a year, while his recommendation of Humana (HUM) — a health insurer using the internet to make its existing business more efficient — eventually gained more than 4,000%, even as many internet infrastructure builders struggled. He is applying the same discipline to AI, and sees a split coming that he calls the “AI Separation”: on one side, the AI Builders spending enormous sums developing the technology and infrastructure, some of which will become extraordinary businesses but may still deliver disappointing stock returns at extravagant valuations; on the other, the AI Appliers — established businesses using AI to cut costs, improve productivity and raise profits, some of which still trade at valuations that leave room for pleasant surprises. He believes this separation will begin on November 4, the morning after the midterms, just weeks before Anthropic’s shares start trading.

Anthropic may become one of the great companies of the 21st century — Fry says he would not bet against that. But at $1.5 trillion or $2 trillion, investors would be paying an enormous amount for a future that has not happened yet. He would rather own companies where the future does not have to be quite so spectacular for the investment to work. As Puzo put it, sometimes the smartest bet is to find a different game.

Industry overview based on InvestorPlace.

What it means for investors

The math is the story here: at a $1.5 trillion to $2 trillion IPO price, Anthropic would need to reach $120 trillion to $160 trillion to repeat Amazon’s 80-fold run from its bubble peak — more than the entire U.S. stock market or all American real estate. That does not mean Anthropic is a bad company; its revenue jumped to $16.8 billion in the third quarter from almost nothing two years ago. It means the entry price does most of the work in determining returns, and at these levels the margin for error is razor-thin. Fry’s framework is the practical takeaway: separate the AI Builders, where huge spending and huge valuations can produce great businesses but poor stocks, from the AI Appliers — established companies using AI to lower costs and lift profits at valuations that still leave room for upside. He expects that split to start showing on November 4, the morning after the midterms, so investors should watch whether money rotates from AI infrastructure names toward AI users in the weeks around Anthropic’s debut.

What it means for employees

There is no direct employee angle in this story. It is about IPO valuation and investment strategy, not jobs, wages or workplaces, so there is nothing here to translate into advice for workers.

What it means for customers

There is no direct customer angle in this story. It is about whether Anthropic’s stock is a good investment at its rumored valuation, not about Claude’s pricing or features, so there is nothing here about what customers will pay or receive.

Regards,

Eric Fry

Editor, Smart Money

Note: Investment calculations and projections are for illustrative purposes only. Market conditions change frequently. Consult a qualified financial advisor before making investment decisions.