IGI Logo

Below the Quote onAnthropic IPO, a trillion-dollar reality check for AI By Enrico Chinello, Next Generation Research Analyst, Julius Baer

Anthropic is marketing its IPO offering at USD 70–USD 78 per share, valuing the Claude maker at up to USD 1.4 trillion fully diluted. While these are impressive figures, they are well below the USD 2 trillion reported earlier this year by the press. Beyond its record size, the deal gives public markets a first full look at frontier artificial intelligence economics. Demand growth, intelligence prices, and compute capacity commitments remain the key areas to watch. We remain Constructive on the Cloud Computing & AI theme.

Anthropic has set terms for its Nasdaq listing: 825 million shares at USD 70USD 78, raising up to USD 64 billion in the base deal and around USD 90 billion including a USD 26 billion private placement with chip and cloud partners. With these figures, the implied valuation would be between USD 1.1 trillion and USD 1.4 trillion. This is well below the reportedly target valuation range of USD 1.5 trillion to USD 2 trillion earlier this autumn. Pricing is due on 18 November. Anthropic has been able to demonstrate formidable growth so far. Revenue reached USD 33 billion in the first nine months this year, a 13-fold annual increase. The annualised revenue run-rate still appears on track to reach USD 65 billion to USD 67 billion this year. On our calculation, the top of the range equals roughly 21x the 2026 annualised revenue run-rate. As we wait for the public market reaction, Anthropic’s listing arrives at a moment where three key areas are top of mind for artificial intelligence (AI) investors. First, demand: all indicators signal continued strength, suggesting that enterprise adoption is real and broadening, which supports the scale of AI investments. Second, pricing power: cheaper models from American and Chinese competitors have pushed the average intelligence cost down, implying that volumes must keep outrunning price deflation to continue to post revenue growth. Third, capital intensity: AI model providers have generally signed large compute capacity agreements, predicated on impressive revenue growth to sustain these expenses. The political angle, where regulators will need to find a balance between potentially stricter data centre regulation and the need to continue to support AI as a national strategic priority, introduces another complexity to the debate. For investors, the deal sets a public, market clearing price for frontier AI, as well as an expectation baseline for OpenAI’s listing expected in 2027. The target valuation range below prior indications, in our view, does not mean the AI wheel is broken, but it signals that markets are starting to discriminate between growth and returns. We believe these are healthy developments, which could contribute to better re-alignment of what is a sustainable, long-term growth trajectory for the AI ecosystem. We remain Constructive on Cloud Computing & AI. 

Above views are of the author and not of the website kindly read disclaimer