Research firm questions Anthropic's $2tn IPO valuation, estimates lower worth
Independent analyst New Constructs says Anthropic's planned IPO is unlikely to justify a $2 trillion valuation, projecting a potential market value as low as $144 billion.
Independent research firm New Constructs released a report on Oct. 6 calling Anthropic's upcoming public offering the "most ridiculous IPO of 2026" and estimating the AI company's equity value could be as low as $144 billion, far below the $2 trillion target the firm has reportedly set.
Using a reverse discounted cash flow model, New Constructs calculated that achieving a $2 trillion market cap would require Anthropic to post a 10% after-tax operating margin immediately and sustain it for ten years, with revenue growth of 1,000% in the current year, 100% in 2027 and 56% annually from 2028 to 2035. The report notes the company posted a negative 175% operating margin for 2025 and would need after-tax operating profit of $358 billion by 2035 to meet the target, a figure it compares with Nvidia's and Apple's trailing profits.
In its lowest-case scenario, the firm assumes margins improve to 20% over a decade and growth slows to 20% annually from 2028, yielding a 2035 after-tax operating profit of $89 billion and a market value of $144 billion. A middle scenario with 36% annual growth results in a valuation of $1.1 trillion, still below the $2 trillion goal.
"In short, we don’t think Anthropic has a viable business," the report states, citing competition from open-source models and higher costs for Anthropic's Claude Fable 5 and Mythos 5 compared with alternatives.
PitchBook senior analyst Harrison Rolfes examined the multiples, noting that at a $2 trillion valuation Anthropic would be priced at about 435 times its projected 2025 revenue, but the multiple falls to under 44 times based on second-quarter revenue and under 31 times using the July run-rate. He added that the company's reported third-quarter revenue of $16.8 billion exceeds the $16.3 billion threshold but does not alone justify the $2 trillion figure.
The draft IPO filing reportedly dedicates roughly 80 of its 261 pages to risk factors, including warnings that AI could pose catastrophic or existential threats and that models might develop self-preserving behaviors. New Constructs highlighted these risks as a primary red flag, referencing Anthropic's own research on potential blackmail behavior by its Claude model.
Bloomberg quoted Slow Ventures partner Sam Lessin suggesting that the U.S. government could require AI labs to hold $1 trillion in escrow for future remediation, estimating societal liabilities could reach $2-$4 trillion.
Anthropic has not responded to inquiries, and the company’s confidential prospectus has not been publicly released.