Everything About AI
NextFin News - Investors are lining up to buy shares in Anthropic's initial public offering before the company has even set a price, and they are doing it at a steep premium to the $965 billion valuation the AI company secured in a private funding round just five months ago. Prospective buyers have discussed IPO valuations as high as $2 trillion - more than double the May private round - even as a broader AI stock selloff pulled that conversation down to roughly $1.2 trillion to $1.3 trillion. The tension between those two numbers is the story: early demand is real, but the premium investors are willing to pay has been shrinking by the week.
Investors Are Bidding Before a Price Exists
Anthropic PBC, the maker of the Claude AI models, may begin formally marketing its IPO as early as Tuesday, November 10, 2026, according to people familiar with the matter. The date is not arbitrary: companies that file a draft registration statement confidentially must make the filing public at least 15 calendar days before starting a roadshow, and Anthropic's public Form S-1 landed on October 26. Fifteen days later is November 10.
The company has not set an offering price, a price range, or a share count. The cover of the registration statement still reads "between $ and $," with both numbers blank. What investors are shopping for, then, is not a priced deal but an indication of where the market will clear. And the early indications point to a valuation well above the last confirmed private price.
That anchor is precise. On May 28, 2026, Anthropic closed a Series H financing of $65 billion at a $965 billion post-money valuation. The round was co-led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital, with participants including Capital Group, Coatue, D1 Capital Partners, GIC, ICONIQ, XN, Baillie Gifford, Blackstone, Brookfield, Fidelity, Lightspeed, T. Rowe Price and Temasek. Amazon rolled roughly $5 billion of previously committed hyperscaler investment into the round. Shares of Series H-1 preferred stock were sold at $589.0095 each, according to the company's filing.
Against that $589 reference price, the IPO conversation has run hot. Early reports put the discussion near $2 trillion, or roughly $1,221 per share - a 107% premium. By early November, after a post-election selloff in AI stocks, prospective buyers were talking about $1.2 trillion to $1.3 trillion, with some as low as $1.1 trillion. Even at the bottom of that range, the implied price of roughly $732 to $793 per share still represents a premium of 24% to 35% over the May round.
The market has moved fast in both directions. On November 2, institutional investors told the company and its advisers they were working with valuations of $1.5 trillion to $1.6 trillion. Three days later, three prospective investors said the price likely had to come down, with one floating roughly $1 trillion. By November 6, the conversation had settled around $1.2 trillion to $1.3 trillion. Anthropic declined to comment.
The timing pressure is real. The earliest the roadshow can start is November 10; the earliest the price could be set is November 13; and the earliest trading could begin is November 16, ahead of the Thanksgiving holiday on November 26. Goldman Sachs and Morgan Stanley are leading the offering, with JPMorgan heading the U.S. retail sales effort. The company is weighing earmarking 15% to 20% of the deal for individual investors, though discussions remain ongoing and nothing has been filed.
What the Premium Is Actually Buying
The first question is what investors are paying for when they agree to a premium over a private round that is only months old. The answer is not current financial performance. It is a bet on where the company will be when the public market reprices it.
Anthropic's own filing shows a business growing at a pace few companies have ever seen, but also one whose growth is decelerating. Third-quarter 2026 revenue was $16.96 billion, up from $11.56 billion in the second quarter - 47% sequential growth, down from 144% in the second quarter, according to analysis of the filing. Annualizing the third quarter gives a run rate of about $67.8 billion. Trailing twelve-month revenue, adding the fourth quarter of 2025 through the third quarter of 2026, comes to roughly $35.3 billion.
At $965 billion, the company trades at 14.2 times that annualized run rate, or 27.4 times trailing revenue. At $1.2 trillion, the multiple rises to 17.7 times annualized and 34.0 times trailing. At $1.5 trillion, it is 22.1 times annualized and 42.5 times trailing. At the $2 trillion figure that circulated in early October, the multiple reaches 29.5 times annualized revenue and 56.7 times trailing.
Those multiples only make sense if buyers believe the current growth rate is the floor, not the ceiling. The filing offers one strong supporting argument: compute costs fell from 172% of revenue to 57% year over year. That is operating leverage in its rawest form - each dollar of revenue now requires far less capital expenditure than it did a year ago. If that trend holds, today's losses compress quickly and the company can reach profitability without slowing investment.
But there is a counterweight. Revenue growth decelerated sharply in the third quarter. A company that grew revenue 144% sequentially in the second quarter and 47% in the third is still expanding at an extraordinary rate, but the direction of travel matters more than the level. Buyers at $1.5 trillion are implicitly assuming the deceleration stops. Buyers at $1.2 trillion are pricing in more of it.
The Premium Is Compressing, Not Expanding
The most important fact in this story is not that investors are willing to pay a premium. It is how much smaller that premium has become in a matter of weeks.
In early October, the conversation ran as high as $2 trillion - roughly double the May private round. By November 6, it had fallen to $1.2 trillion to $1.3 trillion, with some buyers as low as $1.1 trillion. That is a compression of roughly $700 billion to $800 billion in implied value, or about 35% to 40% off the early-October high, in less than a month.
The catalyst was a market-wide repricing of AI risk. The Nasdaq 100 index plunged 6.4% on Thursday, November 5, wiping out $2.3 trillion in market value - a record for a single day. Over the broader post-election selloff, the index fell about 14%. Investors who had planned to anchor their IPO bids to the valuations of already-public AI names found those anchors falling beneath them.
This is the mechanism that turns a premium into a trap. A pre-IPO buyer pays a fixed price and receives illiquid shares, often with a lock-up. If the public market opens lower than the IPO price, the buyer is immediately underwater with no way to exit. The premium that looked like conviction at $2 trillion starts to look like overpayment at $1.2 trillion if the comparable companies keep falling.
The compression also reflects a simple arithmetic reality: the IPO is not happening in a vacuum. Anthropic's pricing will set the template for the rest of the private AI cohort, including OpenAI, which has been discussed at valuations above $1.4 trillion. Bankers and investors know that the first large AI listing will establish the multiple that every other private AI company gets measured against. That creates pressure to price the deal correctly on the first attempt - because a failed debut would damage the entire sector's access to public capital.
Why the Premium Exists at All
Given the compression, why is there still a premium? Three forces are at work.
First, scarcity. Anthropic is one of only two companies in the world building frontier AI models at scale that is close to a public listing. There is no exact comparable. Microsoft, Alphabet and Amazon are diversified conglomerates whose AI businesses are bundled into much larger revenue bases. A pure-play on frontier model demand, with audited financials, is a new asset class. Scarcity commands a premium even in a weak market.
Second, the growth trajectory remains unlike anything public markets have priced before. Revenue went from roughly $9 billion at the end of 2025 to an annualized $67.8 billion by the third quarter of 2026 - more than sevenfold in nine months. No large public software company has ever grown that fast at that scale. Buyers are not underwriting the current quarter; they are underwriting the curve.
Third, the investor base has broadened beyond venture capital. Demand is coming from retail investors, family offices, private funds and private equity firms, according to one IPO adviser who has worked on more than 275 offerings.
Buyers are pricing in where they believe the company will be, not just where it is today.
That adviser, who expects the deal to be heavily oversubscribed, sees a valuation close to $2 trillion as achievable.
That view is not universal. One investor, speaking to a major financial newspaper, argued that if Anthropic is growing 800% a year, even at the low end it should trade at 30 times revenue - a calculation that would value the company at $3 trillion. But that is a bull-case multiple, not a base case, and it assumes the 800% growth rate persists.
The Second-Order Question the Market Is Not Asking
The first-order question is: what will Anthropic's IPO valuation be? The second-order question is more consequential: what happens to the private AI market if the IPO prices at the low end of the current range?
If Anthropic lists at $1.2 trillion to $1.3 trillion, roughly 18 to 19 times annualized revenue, the private market will have to mark down every other late-stage AI company that was valued on the assumption of a $2 trillion anchor. Companies that raised at $965 billion on the expectation of a $2 trillion public comp suddenly find their markups underwater. Secondary sellers who held out for bids well above $1,000 per share will find fewer buyers. The lock-up math changes: early employees and investors who expected a pop on day one may instead face a flat or negative opening.
Conversely, if the deal prices near $1.5 trillion or higher and trades up on day one, the private market gets a green light. Follow-on rounds for OpenAI and other AI companies become easier. Secondary liquidity improves. The multiple that looked stretched at 22 times annualized revenue becomes the new floor.
This is why the order book matters more than the headline valuation. A heavily oversubscribed book at $1.2 trillion is a stronger signal than a thin book at $1.5 trillion. The first says demand is deep at a rational price. The second says the price was set above where real money wants to commit.
Is This Cyclical or Structural? And What to Watch
The compression in the premium is cyclical, not structural. It was driven by a specific, mean-reverting force - a sharp, event-driven selloff in AI stocks that followed the U.S. midterm elections and culminated in the Nasdaq 100's record one-day drop. Selloffs of that magnitude typically stabilize once the forced selling exhausts itself and earnings reality reasserts. If the AI growth story remains intact, the multiple will revert toward where fundamentals justify it.
What is structural is the underlying repricing of frontier AI as an investable asset class. Before 2026, no company had ever reached a $965 billion valuation while still private. The willingness of public-market investors to pay 15 to 20 times a forward revenue run rate for a pre-profit AI model builder is a regime change in how software gets valued - and it will not revert on its own. The cyclical leg is the width of the premium; the structural leg is the existence of the premium at all.
The base case is that Anthropic prices somewhere in the $1.2 trillion to $1.5 trillion range, a premium of 24% to 55% over the May private round, and lists in the second half of November. That range reflects both the scarcity value of a pure-play frontier AI asset and the reality of a market that has just repriced AI risk lower.
The upside case is a clean, oversubscribed book that pushes the price toward $1.5 trillion or higher, with first-day trading confirming the level. That outcome requires the Nasdaq 100 to stabilize, AI-exposed names to stop falling, and the roadshow to produce demand that exceeds the supply of shares on offer. A retail allocation of 15% to 20% could help absorb the deal, but it also introduces a new variable: individual investors are more prone to first-day momentum chasing and quicker to exit if the stock wobbles.
The downside case is a repricing below $1.2 trillion, or a delay. If the market selloff extends into the roadshow window, or if the order book comes in thin, the company faces a choice: price below the private round and signal weakness, or wait. Waiting has costs. The confidential filing has already been made public; the clock is running. Every week the deal sits unpriced, the market has another opportunity to talk the valuation down.
The signals to watch, in order:
- The amended S-1 with a price range. Until it lands, every valuation figure is a report, not a term.
- The quality of the order book during the roadshow - not just the size, but the type of investor committing.
- The Nasdaq 100 and the performance of already-public AI-exposed names during the marketing window.
- Whether the retail allocation is confirmed at 15% to 20%, and how retail demand compares to institutional demand.
The falsifying signal for the premium thesis is specific: if the final price comes in at or below $589 per share - the May Series H-1 reference price - the "big premium" narrative is wrong, and the market is telling investors that five months of AI enthusiasm has been fully given back.
The premium is real, but it is a shrinking premium. Investors are still willing to pay above the last private round for a scarce, fast-growing AI asset - yet the gap between early-October euphoria and early-November realism shows how quickly conviction thins when the broader market turns. The IPO will not tell investors whether Anthropic is a great company. It will tell them how much of that greatness was already priced in.
Explore more exclusive insights at nextfin.ai.
