The largest IPO in history is a few weeks away.
Anthropic filed confidentially with the SEC back on June 1st. Word is the public prospectus drops after Labor Day, listing maybe late September, maybe early October. Morgan Stanley, Goldman and JPMorgan are running it.
The number people keep saying out loud is two trillion dollars.
I’ve been doing this 25 years and I’ve never watched a company grow like this one.
January 2024, the run rate was $87 million. By that December it was a billion, nine billion by the end of last year, and it went past sixty-five billion in July.
That’s not taking the steps up…it’s a ROCKET SHIP!
Now for the thing nobody’s talking about.
There’s a rule about what happens to an industry when its biggest player goes public, and it’s held since 1983. It does the exact opposite of what you think it does.
If you own anything in AI, that rule decides what happens to your money the day that prospectus hits.
Two Yale professors already did this work, so I don’t have to.
Spiegel and Tookes went through every US IPO from 1983 to 2011 and asked one question. What happens to everybody else in the industry?
Their answer: after a firm goes public, every competitor sees profits per unit of market share go down, and their customers get easier to steal.
Read that again.
You probably think a hot company files, everybody gets excited about the sector, and all boats rise. That’s the instinct, and it’s wrong.
What happens instead is that one company walks out of the process with a war chest, a public stock they can use to buy things, and a marketing budget nobody else can match.
The IPO doesn’t lift the industry. It arms one guy and points him at everybody else.
Now flip it over
If going public hurts the competition, then not going public should help them, right?
That’s exactly what happens.
The Bank for International Settlements looked at China, where regulators just suspended listings outright. The companies most exposed to whoever was sitting in that IPO queue popped hardest on the announcement day.
Two things happened at once. The threat of a better-funded rival evaporated, and so did all those new shares that would’ve been competing for the same investor dollars.
And here’s the tell. The weak companies gained the most while the strong ones barely moved, which tells you this is about who was genuinely scared and not about sentiment.
Two times you can go look this up
Goldman Sachs filed August 24th, 1998 and pulled it a month later. Asian crisis was tearing through everything, Nikkei down 39%, and they said thanks but no thanks. They refiled the following March and finally listed in May of ’99.
Then there’s WeWork, which everybody remembers.
S-1 came out August 14th, 2019 at $47 billion. Three weeks later the number being whispered was twenty to twenty-five, and the deal got pulled in September. Cost Goldman eighty million bucks just for showing up.
Now write this one down, because it’s the number nobody quotes.
Of companies that withdraw an IPO, about 13% ever come back and do it. Eleven percent go bankrupt and forty-two percent get bought.
Pulling an offering isn’t a delay, folks. For most of them it’s the funeral.
So what does this mean for you
If it lists, your AI positions just got a competitor with a printing press.
Two trillion dollars of public currency, and every one of your names now has to compete with somebody who can outspend them, outhire them, and buy whatever they can’t build. That doesn’t show up next Tuesday. It shows up in margins, over years, quietly, until one day you’re wondering why the thing you own stopped working.
If it gets pulled, the whole thing runs backwards. The names most exposed get a reprieve, and the weakest of them get the biggest one.
Either way, September 7th is the date.
We just watched a version of this with SpaceX, which came public in June at the biggest raise on record and has spent eleven weeks arguing with itself about what it’s worth. The rule had things to say about that one too.
I’ve got a lot more to say about this in the coming days.
Your only option,
Mark Sebastian