Anthropic's IPO Needs Double Nvidia's Profit to Justify $2 Trillion

A research firm has put a number on what a $2 trillion Anthropic would have to become. In a note published Tuesday, October 6 (CNBC reported it on October 7), New Constructs estimated that Anthropic would need to earn double Nvidia's trailing-year profit to justify the valuation it is reportedly seeking in its Nasdaq listing. Nvidia's net income over the past four quarters topped $190 billion, according to the source. New Constructs values Anthropic at $150 billion, and called the offering the "most ridiculous IPO of 2026."

The Numbers Behind the Argument

The case rests on reported figures, not a look at the filing itself. Anthropic has not made its prospectus public, and New Constructs says it has not seen the S-1, the registration document a company files with regulators before going public. The firm relied on a leaked copy cited by Reuters, which reported that Anthropic's 2025 revenue was $4.6 billion against a net loss of $42 billion.

Against that sits a much brighter picture from the company's side. Anthropic claimed at the end of July that its annualized revenue run rate was up sevenfold year-over-year to $65 billion. The New York Times reported in September that it was on pace for $100 billion in annualized revenue by the end of 2026. Anthropic did not respond to CNBC's request for comment.

Why Run Rate and Valuation Don't Meet

Two terms explain the gap. A run rate takes a recent period's revenue and multiplies it out to a full year. It is a snapshot of momentum, not money actually collected, and it says nothing about cost. A market cap is the share price times the number of shares, and at $2 trillion it prices in a future in which the company earns enormous profits for years. New Constructs argues the reported losses and the rise of open-source models, whose weights anyone can download and run, mean closed-model vendors will struggle to turn revenue into profit. Its words: "We don't think Anthropic has a viable business." That is one firm's opinion, not a finding.

Who Is Exposed

New Constructs argues the offering's purpose is liquidity for the company's Wall Street backers rather than wealth for public investors. If that is right, the people holding the risk are the buyers on day one: index funds that must own the stock once it is large enough, retirement accounts, and retail investors. A listing at this scale would make one unprofitable AI lab a major weight in tech-heavy portfolios.

Operators who build on Anthropic's models have a quieter exposure. A company under pressure to show profit after listing may change pricing, usage limits or product priorities. That is an inference, not something the source says.

A Track Record Worth Weighing

David Trainer, New Constructs' founder, has a reputation as a bear on IPOs, and his record is mixed. The firm called WeWork the most ridiculous IPO of 2019; the offering was pulled six weeks later and WeWork filed for bankruptcy in 2023. It was bearish on Allbirds in 2021, which later sold its assets for an estimated $39 million. But it made the same call on DoorDash in 2020, and DoorDash's market cap has since grown from over $60 billion to $83 billion. Trainer himself has said the firm does not always get it right.

Questions You Should Be Asking

• Which figures in this debate come from the actual prospectus, and which come from leaks and press reports that no one outside the company can verify?
• What would Anthropic's earnings have to look like, and by when, for $2 trillion to be reasonable, and who has shown that math?
• If open-source models keep improving, what exactly stops customers from switching, and has Anthropic said so?
• If you depend on Anthropic's models, what is your plan if pricing or terms change after listing?
• Do you hold Anthropic exposure through an index fund or retirement plan, and do you know how much?

What To Watch Next

The signal is the public S-1. Once Anthropic files it, the leaked loss and revenue figures can be checked against audited numbers, and the offering's price range will show how much of the $2 trillion ambition survives contact with investors. Weak demand ended WeWork's attempt within weeks; strong demand would count against New Constructs.