Investors Balk at Anthropic’s $1.5 Trillion IPO Price as AI Sell-Off Puts November Listing in Doubt

Major investors in Anthropic’s blockbuster stock market debut are pushing back against the $1.5 trillion valuation the artificial intelligence company had been counting on, and some now doubt whether the offering can even begin its investor roadshow next week in time for a November listing, according to Reuters and The Wall Street Journal.

The resistance lands at the worst possible moment for the most anticipated listing of the year. Anthropic, the maker of the Claude chatbot, filed to go public on 26 October in an offering that some expect to reach $100 billion, which would make it one of the largest initial public offerings in history. But a brutal two-day sell-off in technology shares after Tuesday’s midterm elections has changed the arithmetic, and the buyers who were supposed to anchor the deal are now asking for a cheaper price.

Four people familiar with the investor discussions told Reuters that potential buyers are raising doubts about whether Anthropic can proceed with its planned roadshow next week, with some viewing the recently discussed $1.5 trillion valuation as being at the high end of expectations. The roadshow, the series of meetings with big investors that sets the price of a listing, had been expected to begin in the week of 9 November, with pricing the following week and trading the week before Thanksgiving. That timetable is now in question.

The Wall Street Journal reported on Thursday evening that major investors are increasingly sceptical the offering can proceed at a valuation of $1.5 trillion after the deep market sell-off tied to the unwinding of artificial intelligence bets. Three major investors who plan to take part in the offering told the Journal that the valuation will probably need to come down, with one saying it could be closer to $1 trillion. Anthropic declined to comment on the investor doubts, according to Reuters.

Anthropic’s executives and its advisers are expected to meet in the next few days to discuss the plans, the Journal reported, and the company could also decide to raise less money than anticipated, perhaps around $80 billion rather than the $100 billion some had expected. What is not yet decided is whether the company will accept a lower price, delay the listing until markets look more resilient, or press ahead and test whether demand is stronger than the sceptics claim.

The valuation gap widens

The argument over price has been running for weeks, but the direction of travel has been steadily downwards. At one point this year, investors expected the listing to take place in October and to seek a valuation as high as $2 trillion, according to Reuters. By early November, institutional investors lining up for the offering had told the company and its advisers that a valuation of about $1.5 trillion to $1.6 trillion was fair, well below the figures closer to $2 trillion that had circulated for months, the Journal reported on 4 November.

Now even that reduced figure is being questioned. The latest reporting suggests the gap between what Anthropic and its bankers hoped for and what the biggest buyers are willing to pay has widened again, and the speed of the shift is striking. In the space of a few weeks, the conversation has moved from whether the company is worth $2 trillion, to whether $1.5 trillion is fair, to whether $1 trillion is closer to the mark.

Investors are also waiting for a number Anthropic has not yet provided. The company left its October revenue figure blank in its updated stock market filing, and two people familiar with the discussions told Reuters that buyers are eager to see it before committing, to judge whether Anthropic has managed to restore growth after a slowdown in recent months. Until that figure appears, the buyers hold the leverage, because they know the company needs the money.

That need is real. Anthropic has disclosed purchase and lease commitments for computing power and infrastructure of about $895.4 billion, running as far out as 2045, according to its updated filing. About $447.5 billion of that total falls due by the end of 2030, and roughly $69.8 billion comes due in the twelve months after 30 September. Against those obligations, the company reported about $146.6 billion in cash, short-term investments and available credit. The gap between what it has promised to spend and what it holds is the reason the listing matters so much, and it is also the reason investors feel entitled to drive a harder bargain.

Some of those commitments can be reduced. About $213.8 billion of the total is conditional or can be terminated, according to analysis of the filing. But even allowing for that, the scale of the spending is enormous, and it is growing far faster than revenue. Analysts who pieced together the figures from the original prospectus had estimated about $800 billion of commitments, and an earlier reading of the 26 October filing put the figure at about $265.6 billion as of 30 June, which would imply more than $600 billion was added in a few months. The measurements differ by date and definition, but every reading points in the same direction: the obligations are swelling.

What the amended filing revealed

The updated filing that arrived alongside the valuation debate gave investors plenty to argue about. Anthropic reported third-quarter revenue of $16.96 billion, up 47 per cent from the previous quarter and about 1,127 per cent higher than a year earlier, according to coverage of the filing. The headline looks strong. The trend underneath it is less comfortable.

Sequential growth had been 132 per cent in the first quarter and 144 per cent in the second, so the third quarter broke a run of accelerating growth that had become central to the case for the listing. Anthropic gave three reasons for the slowdown: it did not have enough computing capacity, business activity was seasonally slow over the summer, and it restricted customers who broke its terms of service. It did not say how much each factor contributed. Price cuts tied to new model versions, plus contractual discounts, reduced revenue by about $600 million. Analysts have pointed out that adding that money back only lifts sequential growth to about 52 per cent, still far below the second quarter.

Profitability also moved in the wrong direction. An operating profit of $974 million in the second quarter became an operating loss of $567 million in the third. The company blamed roughly $5 billion of computing power bought under expensive short-term contracts and a $2 billion non-cash charge for matching employee donations. Compute costs reached 58 per cent of revenue, up from 48 per cent the quarter before. The $23.29 billion net loss reported for the quarter looks alarming, but most of it comes from non-cash changes in the value of convertible notes, and the company reported an $80.7 billion net loss for the first half of 2026 on a similar basis, a charge that grew mainly because Anthropic’s own valuation rose.

There are bright spots in the numbers too. The company’s annualised revenue run rate topped $80 billion at the end of October, up from the prior month, according to a person familiar with the matter cited by the Journal. Net revenue retention is about 900 per cent, meaning existing customers are spending far more than they did a year ago. Some 79 per cent of third-quarter revenue came from customers outside the top five, up from 64 per cent in the first quarter, so the business is less dependent on a handful of big accounts. More than 11,900 customers now generate at least $100,000 a year, up from about 1,500 at the end of 2025.

Valuation specialists are split on what all of this is worth. Professor Aswath Damodaran of New York University, one of the most widely followed independent valuers of technology companies, trimmed his equity value for Anthropic to $1.45 trillion from $1.475 trillion after the third-quarter numbers, calling the quarter the most telegraphed third-quarter result of all time and saying the slowdown was already priced into his model. He sees the assumption of $1 trillion in revenue within ten years as the weakest link in the case for the company. If revenue falls short, he warned, margins would shrink, contracted computing power could sit idle and the risk of default would rise. Matt Kennedy of Renaissance Capital said the results were not worrying but expected the pressure on growth and margins to show up in the final valuation.

Measured against trailing twelve-month revenue of about $35.3 billion, a valuation of $1.5 trillion to $2 trillion works out to roughly 42 to 57 times sales. Secondary-market trading, where existing shareholders sell to each other, recently implied a valuation of about $1.41 trillion, according to Nasdaq Private Market, while another estimate put it at about $1.57 trillion as of 12 October. Anthropic raised $65 billion in May at a $965 billion post-money valuation, so even a listing at $1.5 trillion would be a large premium to the last private round. The question is how large a premium the public market will pay for a company that is growing fast, spending faster, and still losing money on an operating basis.

A brutal week for AI stocks

The valuation debate cannot be separated from the market Anthropic is trying to list into. Trading was anything but smooth on Thursday, fuelled by the outcome at the polls but also by corporate developments and increasing market anxiety, as Reuters put it. The Nasdaq Composite fell 5.5 per cent on Thursday, entering correction territory, while the Nasdaq 100 dropped 6.3 per cent and the S&P 500 lost 4.4 per cent. The Philadelphia Semiconductor Index plunged about 12 per cent in its worst session in six years. More than 50 technology stocks in the S&P 1500 dropped by 10 per cent or more in a single session, and Vanda Research said retail selling was the strongest it had been since the Covid-19 pandemic.

Friday brought a partial recovery, with the Nasdaq Composite rebounding 1.09 per cent and the semiconductor index jumping 4.2 per cent, led by Advanced Micro Devices, which surged 11 per cent, and Intel, which gained more than 7 per cent, according to Investopedia and the Schwab Network. But the damage to confidence was done. Technology shares have underperformed the S&P 500 by 10.7 per cent over the past month, the most extreme one-month underperformance since May 2002, according to Keith Lerner, chief investment officer at Truist Advisory Services. Lerner described the selling as a buyers’ strike in technology, adding that it was uncomfortable but not abnormal, while warning that the rubber band is getting stretched.

The proximate cause was political. Democrats won control of both chambers of Congress on Tuesday, and voters in many states backed ballot measures limiting data-centre development, crystallising political opposition to the infrastructure the AI boom depends on. The fear among investors is that the new Congress will tighten scrutiny of AI data centres and the huge capital spending behind them, and that state and local resistance will slow the build-out. Rick Meckler, a partner at Cherry Lane Investments, a family investment office in New Vernon, New Jersey, told Reuters that valuations have risen for a long time in the private market and that investors will probably see some pullback, while insisting the companies themselves are not going away. “It’s not like these companies are not going to be a part of the future,” he said. “They’re going to become public. Investors are still going to have a lot of interest in them.”

Corporate news added to the pressure. Microsoft tightened oversight of some data-centre projects from the smaller cloud providers known as neoclouds, Nebius, IREN and Nscale, Bloomberg reported on Thursday, according to Reuters. Nebius, IREN and CoreWeave were among the biggest decliners on Thursday, and each has lost about half its market value over the past month. Bloomberg also reported on Friday that Microsoft is winding down a 2023 arrangement to rent AI computing capacity from Oracle, after the older generation of Nvidia chips in the servers stopped meeting its requirements, and that Microsoft is reassessing more than $80 billion in other outside cloud contracts. Oracle shares fell about 4 per cent on the news. Credit default swaps covering CoreWeave and SpaceX alike hit records that imply an increasing expectation they could default on their obligations in coming years, according to Reuters.

The jobs report on Friday gave the market a second shock. The United States added 165,000 jobs in October, far above the 59,000 forecast, and the unemployment rate fell to 4.1 per cent, according to the Labor Department. After the data, traders raised the odds of a Federal Reserve rate rise in December to 78 per cent from 70 per cent, according to LSEG data cited in market coverage. Higher rates tend to weigh on growth stocks, and the strong jobs numbers complicated the rally even as technology shares bought back some of Thursday’s losses.

The OpenAI parallel

Anthropic is not the only AI company finding that investors have become harder to please. The Journal reported that OpenAI, Anthropic’s great rival, has raised less than half of the $30 billion it had initially targeted for its latest funding round. NEXA reported on 1 November that investors had committed just $11 billion and that OpenAI had cut its fundraising target to $20 billion. The parallel matters because the two companies have been the twin symbols of the AI boom, and both are now discovering that the private market’s enthusiasm has limits when the public market turns hostile.

The contrast between the two companies is instructive. OpenAI is raising money privately, where it can negotiate quietly and adjust its target without a public price. Anthropic is trying to do the same thing in public, with a prospectus, a roadshow and a ticker, which means every disagreement about price becomes news. The $100 billion Anthropic hopes to raise is far larger than OpenAI’s round, and the valuation at stake is roughly fifty times larger. That scale is precisely what makes the listing so important, and so fragile.

There is also a political shadow over the deal. Senator Elizabeth Warren wrote to Anthropic’s chief executive, Dario Amodei, asking how the company justifies its valuation, with a deadline for a reply of 3 November. The letter does not block the listing, but it signals that the deal will be examined in Washington as well as on Wall Street, and that the company’s answers about its numbers will be read by people who are not buying shares.

How the deal is structured

Despite the doubts, the mechanics of the listing are well advanced. Anthropic plans to list its Class A shares on the Nasdaq Global Select Market under the ticker ANT. Goldman Sachs is the lead-left underwriter and Morgan Stanley sits on the right, with twenty more banks in the syndicate, including JPMorgan, Citigroup and Barclays. The amended filing sets aside part of the offering for individual investors through Charles Schwab, Fidelity, Robinhood, SoFi, E*TRADE from Morgan Stanley and Trade-PMR, with JPMorgan leading sales to individual investors in the United States.

Reports of the deal’s size vary. One account says Anthropic aims to sell about $100 billion of shares. Others cite a range of $60 billion to $100 billion and stress that the filing leaves the share count and the price range blank. On timing, a roadshow could begin as soon as the week of 9 November, with trading possible the week of 16 November, just before Thanksgiving. Unlike SpaceX, which set a single price for its June listing, Anthropic is expected to propose a price range. SpaceX rose 30 per cent on its first day in June, a precedent the bankers will have in mind when they decide how aggressively to price Anthropic.

The structure reflects a deliberate choice. According to the Journal, some investors trust Anthropic’s chief financial officer, Krishna Rao, to set a conservative valuation and give priority to delivering IPO investors a first-day gain in the stock price. They believe that even if Anthropic debuts at a roughly $1.5 trillion valuation, it could jump to $2 trillion by the end of the first trading day. That is the optimistic reading: price low, let the market do the work, and let the headline valuation arrive on the first day rather than in the prospectus. The pessimistic reading is that the buyers are not bluffing, and that a company which needs tens of billions of dollars for computing power cannot afford to discover on its first day of trading that the market will not pay what it hoped.

Anthropic’s backers include Amazon and Alphabet, both of which would be well rewarded by a listing at anything near the figures under discussion. The company’s governance is unusual: control is held firmly by insiders, and the filing describes a company that is hard to value, with costs tied up in contracts that run for decades. That combination of insider control and very long-dated obligations is part of what makes the pricing argument so difficult. Buyers are being asked to pay a premium for growth, to fund commitments that stretch to 2045, and to accept that the people running the company will keep control after it lists.

Analysis: Why It Matters

The IPO is a referendum on the AI boom

Anthropic’s listing is the single most important test of the artificial intelligence boom since the market turned. The company is not a start-up with a prototype; it is a business with tens of billions of dollars in annual revenue, a product used by millions of people, and backing from two of the largest technology companies in the world. If buyers will not pay $1.5 trillion for that, the question becomes what they will pay for anything in the sector. The repricing is already rippling outwards: secondary-market valuations, private funding rounds and the share prices of chipmakers and cloud providers all take their cue from the price the market sets for the best-known AI company. A delay or a sharp cut would not just be Anthropic’s problem. It would be read as evidence that the whole edifice of AI valuations, built in a few short years on the promise of future revenue, is being re-examined at once.

The buyers have leverage, and they know it

The striking feature of this negotiation is how openly the buyers are using their position. In most large listings, the argument about price happens behind closed doors, and the public sees only the final number. Here, major investors have told the Journal and Reuters, on the record through people familiar with the talks, that the price must come down and that the timetable may slip. That is not an accident. Anthropic needs the money more than the buyers need the shares, because its computing commitments are already signed and the bills fall due whether or not the listing happens. The buyers know it, the bankers know it, and the company knows it. The result is a negotiation conducted partly in public, in which each leak is a move. The risk for the buyers is that they push too hard and the deal is pulled, leaving them without the shares they want. The risk for Anthropic is that it accepts a price that defines its value for years, in front of the whole market, at a moment when the market is at its most sceptical.

The slowdown is real, even if the growth is too

It is easy to read the filing as a story of a company in trouble, and that would be wrong. Revenue of nearly $17 billion in a single quarter, growing more than eleven-fold in a year, with customers spending nine times more than they did a year ago, is an extraordinary business by any standard. The problem is not that Anthropic is failing. The problem is that the story investors were told, of growth accelerating quarter after quarter, has broken, and the costs of that growth have risen faster than the revenue. Compute now eats 58 cents of every dollar of revenue, and the company had to buy expensive short-term capacity because it did not have enough of its own. That is a fixable problem, and Anthropic says the capacity shortage is one of the three reasons growth slowed. But it is also the reason the company needs to raise $100 billion, and it is the reason investors are asking whether the business model works at the scale Anthropic is promising. A company that must spend $895 billion to earn $35 billion a year is making a bet that the future will be very much larger than the present, and the buyers are entitled to ask what happens if it is not.

Politics has joined the pricing model

A year ago, the price of an AI company was a question of revenue growth, margins and market share. This week it became a question of politics as well. The midterm results, the ballot measures limiting data centres, the tightening of oversight by Microsoft, and the letter from Senator Warren all point in the same direction: the political environment that allowed the AI build-out to proceed at full speed is changing. Data centres need power, land and water, and the communities that host them are increasingly asking what they get in return. Congress, under new management, is expected to ask harder questions about the concentration of computing power, the cost of the build-out, and who benefits from it. None of this makes the technology less useful. But it does make the business of building it slower, more expensive and less certain, and that uncertainty is now being priced into the most important listing of the year. Anthropic’s valuation is not just a number. It is a judgement about how much political friction the AI boom can absorb.

The first-day pop is not a strategy

There is a seductive argument circulating among the optimists: price the deal conservatively, let it rise on the first day, and let the market deliver the $2 trillion valuation that the prospectus could not. SpaceX did something similar in June, rising 30 per cent on its debut. But a first-day pop is a consolation prize, not a strategy, and it carries its own risks. A company that prices at $1 trillion and watches its shares rise to $1.5 trillion has left $500 billion on the table, money it could have used to pay for the computing power it has already promised to buy. A company that prices at $1.5 trillion and watches its shares fall has a different problem: a broken listing, angry investors, and a valuation that haunts every future fundraising. The bankers’ job is to find the price at which the shares change hands smoothly and stay there, and the fact that the argument is being conducted in public suggests they have not found it yet. The meeting between Anthropic’s executives and its advisers in the next few days will be the moment that decides which of these risks the company is willing to take.

What to Watch Next

1. Whether the roadshow begins in the week of 9 November, as originally planned, or is pushed back. A delay of even a week would signal that the buyers have won the first round of the negotiation, and would put the November listing in doubt.
2. The October revenue figure, which Anthropic left blank in its updated filing. Investors are waiting to see whether growth reaccelerated after the third-quarter slowdown, and the number will set the tone for the entire pricing discussion.
3. The meeting between Anthropic’s executives and its advisers in the next few days. The decision on whether to cut the valuation, raise less money, or delay the listing will come out of that room, and the market will read it as the company’s answer to the buyers.
4. How technology shares trade next week. The Nasdaq’s partial recovery on Friday showed that buyers will return after a sharp fall, but another leg down would make it very hard to launch the largest offering in years into a falling market.
5. What happens to the wider IPO pipeline. Companies weighing their own listings are waiting to see how Anthropic trades. A smooth deal could open the way for a wave of new issues early next year, while a weak one would deepen worries about AI exposure across the market.

Sources

• Reuters, via The Hindu BusinessLine: Will AI backlash delay Anthropic IPO? Investors raise doubts
• The Wall Street Journal: Major Investors Are Skeptical Anthropic IPO Can Achieve $1.5 Trillion Valuation
• The Wall Street Journal: Anthropic Investors Push for Cautious Valuation Closer to $1.5 Trillion
• Investopedia: Markets News, Nov. 6, 2026: Nasdaq, S&P 500 Rise as Chip Stocks Bounce Back from AI Tailspin
• Schwab Network: Closing Bell: Tech Rebounds as Strong Jobs Report Lifts Rate-Hike Odds
• Aistify, reporting Bloomberg: Microsoft Winds Down Oracle AI Cloud Deal as Older GPUs Lose Their Fit