Anthropic lowers valuation range after withdrawing IPO pricing filing, bearish sentiment rises in prediction markets
Anthropic withdrew its IPO pricing filing on November 9, one day before it was originally scheduled to launch the investor roadshow, and subsequently updated it to $70 to $78 per share, offering 825 million Class A shares, with proceeds reduced to $57.75 billion to $64.35 billion, and the maximum market value lowered from the initially rumored $2 trillion to $1.404 trillion. On Polymarket, the implied probability of betting that its first-day closing market value would be below $1.25 trillion rose from about 2.4% to about 36% within a week, and total bets on the related prediction market have exceeded $37 million. The article also mentions that Anthropic released Fable 5.2 and Mythos 5.2 during the same period, but the capital market is no longer willing to pay simply for stronger models.
Why read this
AI score 75
From the withdrawal of the pricing filing to the reversal of prediction market odds, this shows the rapid shift in bullish and bearish expectations in the Anthropic IPO pricing battle.
Topics
Body · AI translation
Anthropic's highly anticipated IPO, expected to make history, has seen more and more "surprises."
The latest surprise is that on November 9, the day before it was originally scheduled to launch its investor roadshow, Anthropic withdrew its IPO pricing document. That document would originally have set its maximum valuation at $1.441 trillion. Anthropic then updated the document, setting the valuation range at $70 to $78 per share, which means the maximum valuation was further lowered to $1.404 trillion, while the low end of the range was only $1.26 trillion.
The fierce pricing battle surrounding Anthropic's IPO has turned into a convoluted ongoing drama. From the initially rumored $2 trillion down to today's $1.4 trillion took less than half a month.
Meanwhile, in one corner of the internet—the prediction markets—a fierce battle between bulls and bears has also unfolded over Anthropic's post-IPO valuation performance, and for now the bears have won a sweeping victory. In theory, by betting against Anthropic, one could net a return of more than 10x within a week.
More and more people are betting against Anthropic
Over the past half month or so, Anthropic's IPO has set off a frenzy on public betting platforms. On Polymarket, the world's largest prediction market, more than a dozen prediction markets related to Anthropic's IPO, large and small, have appeared since October, with total bets exceeding $37 million. That does not even count several other mainstream betting platforms such as Kalshi and trade.xyz.
The so-called "prediction market" is a concept that has become very popular in cryptocurrency circles this year. Simply put, it lets users trade on predictions about real-world events: buying "yes" means betting that something will happen, and the opposite means betting that it will not. Each prediction is priced between $0 and $1, directly representing the probability the market assigns to the event occurring. What distinguishes it from ordinary sports betting platforms is that there is no house here—only users betting against one another.
Polymarket is a leading player among prediction markets. In the first half of this year, its average daily trading volume grew 4x year over year, and after its Series F financing at the end of October, the company's valuation had reached $21 billion. Because Polymarket has a large user base and all bets are made with real money, its prediction results are often regarded on social platforms as a kind of bellwether, especially for predictions about U.S. elections and sporting events such as the World Cup.
Finance is actually not a traditionally popular category on Polymarket, but this year more and more people have begun "trading stocks" on Polymarket. The last time Polymarket drew attention in the secondary market was during SpaceX's landmark IPO in the first half of the year. At that time, the debate over SpaceX's valuation was no less intense than the current one over Anthropic. And from early April to early May this year, the implied probability on Polymarket that SpaceX would "close above a $2 trillion market cap on its first day" rose from 45% all the way to 62%. In the end, SpaceX's first-day closing market cap did indeed exceed $2 trillion. Since then, many media outlets and analysts have begun to treat betting results on prediction platforms as a reference for price discovery ahead of major IPOs.
This time, Anthropic's IPO is drawing even more attention than SpaceX's. Reuters quoted a wealth manager as saying: "People's interest in taking part in predictions about Anthropic's IPO is like queuing up to snap up the newly released iPhone." And among the various predictions (listing date, amount raised, and so on), the one retail investors are most enthusiastic about, of course, is the traditional "first-day closing market cap" item.
And on this item, Anthropic's performance in recent days has been quite worrying, with the predicted market cap shrinking rapidly. Let's look at the odds. As of November 10, the odds on Polymarket for betting that Anthropic's first-day closing market cap would be below $1.25 trillion had fallen to 2.8x. Just a week earlier, those odds had been as high as 42x. At the same time, the odds for betting that Anthropic's market cap would be between $1.75 trillion and $2.0 trillion rose from 2.5x at the end of October to 5.6x now.
To put it simply, a week ago the betting market believed there was only about a 2.4% chance that Anthropic's first-day closing market cap would be below $1.25 trillion; now that has risen to about 36%.
This is not a simple vote—everyone making a choice has bet real money. Moreover, "prediction shares" on Polymarket can be traded in real time. That is to say, if one had bought the prediction "first-day market cap below $1.25 trillion" a week ago and sold it today, in theory one could earn a return of about 15x.
Although it has not yet listed, Anthropic's share price on Polymarket has already well and truly collapsed. If Anthropic's first-day market cap really does fall below $1.25 trillion, that would mean a broken IPO even at the low end of its pricing range. This is a world of difference from the way SpaceX's IPO in May was enthusiastically embraced by bettors on Polymarket.
Why did Anthropic withdraw its pricing document?
Just as prediction markets swung sharply bearish, a seemingly bungled withdrawal of an offering document has cast another shadow over Anthropic's IPO journey.
Reuters reported that Anthropic sent a document to investment banks on the evening of November 9, US time, showing a planned offering of about 867 million shares at a price range of $70 to $80 per share, raising about $69.3 billion at the top of the range. But less than two hours later, Anthropic sent another email saying the pricing information previously provided was inaccurate and that it would update the document later.
Anthropic then updated the document, lowering both the price range and the fundraising amount, ultimately settling on issuing a total of 825 million Class A shares at an estimated price range of $70 to $78 per share, with proceeds falling to between $57.75 billion and $64.35 billion.
To take a conspiratorial view, perhaps Anthropic wanted to test the market's reaction before deciding whether to adjust the terms, and after poor investor feedback it withdrew and redesigned the offering structure. Of course, this may also be overreading things; the world really is a slapdash operation, and this was simply an internal communication error—the wrong version of the document was sent.
Before October, when Anthropic's prospectus had not yet been published, the expected valuation floated was $2 trillion. Over the past 10-plus days, that number has shrunk dramatically. In early November, media reports said $1.5 trillion was regarded as the upper limit of Anthropic's pricing range.
Then came the November 5 stock market crash, when the Nasdaq Composite fell 5.56% in a single day and about $2.5 trillion in Nasdaq market value was wiped out, the largest single-day market value drop in its history. Tech stocks, chip stocks and AI-related companies were almost all routed. That day, The Wall Street Journal reported: "Some IPO investors are skeptical that Anthropic can achieve a $1.5 trillion valuation." There were also rumors that Anthropic might decide to postpone the offering.
For now, Anthropic has been quite "willing to take advice," choosing to slash the offering price sharply and press ahead with the IPO rather than "wait for a more favorable market environment."
The fierce battle over Anthropic's IPO pricing has effectively become a general showdown in the capital markets between bullish and bearish attitudes toward AI. Faced with the same prospectus, one camp of analysts considers it "better than expected," or even "injecting new vitality into a struggling AI industry"; the other camp focuses on the Q3 growth slowdown and argues that AI's grand narrative has "peaked."
As the village's most outstanding university student, Anthropic carries too much on its shoulders.
For VCs in the primary market, Anthropic's IPO is of great significance. Because of its and OpenAI's unprecedented massive financing needs, many established VCs have been forced to make "decisions that go against their ancestors," having to sharply expand fund sizes in order to get on board. After entering 2025, they even issued a large number of controversial SPVs specifically to invest in these two giants.
The problem now is that Silicon Valley VCs hold a large amount of paper gains, but over the past 10 years they have not returned much cash to LPs, so it is hard for them to contribute more to larger future funds. If Anthropic lists successfully, hundreds of billions of dollars in massive capital will flow back into Silicon Valley's VC ecosystem, and all of this will no longer be a problem.
For the entire AI industry, Anthropic's IPO is even more of a case where one move affects the whole situation. On the one hand, its share price will directly become the valuation anchor for other AI companies. On the other hand, Anthropic has on its books up to $895.4 billion in future computing power and infrastructure spending commitments, of which $447.5 billion will come due before 2030. If Anthropic's performance falls short of expectations, or its cash flow falls short of expectations, the entire AI infra sector will need to be revalued.
Anthropic's products remain excellent. Just as various unexpected dramas were unfolding around the IPO, Anthropic released its new-generation flagship model Fable 5.2 on November 9, and simultaneously launched Mythos 5.2 based on the same underlying model. Fable 5.2 continues to lead the world in performance, and while being more powerful it has also sharply lowered its price, even becoming cheaper than OpenAI's secondary model Sol 6.
But in stark contrast to the first half of the year, the capital markets are now finding it hard to pay simply for "stronger models."
It is worth mentioning that on November 10 SoftBank released its latest quarterly report (July-September). Two pieces of information in the report caught the eye: first, a net loss of 1.12 trillion yen for the quarter; second, Elon Musk turned out to be right—SoftBank really did default on a batch of orders for power generation equipment. This equipment was originally intended for a data center in Ohio, USA, and that data center is part of the "Stargate" project jointly built by OpenAI, Oracle and SoftBank.
From Anthropic to SoftBank, saying the bubble has burst is an exaggeration, but the great ship of AI is undoubtedly rocking.
English text is being prepared.