Anthropic Investors Push Back on $1.5 Trillion IPO Valuation as Tech Selloff Hits

Investors in Anthropic’s planned stock market listing are pressing the artificial intelligence company to cut its proposed $1.5 trillion valuation, with at least three major investors telling bankers the figure is too high after a sharp selloff in technology stocks, according to the Wall Street Journal.

The pushback has become one of the biggest sticking points in preparations for one of the most closely watched initial public offerings of the year. Anthropic, the maker of the Claude AI models, had been expected to pursue an offering of around $100 billion as early as November. The Wall Street Journal reported that the company could now seek around $80 billion instead, and that one investor suggested a valuation closer to $1 trillion.

The valuation debate comes as investors await further details about the company’s financial performance. Anthropic’s amended filing with the Securities and Exchange Commission showed that its revenue reached $17 billion in the third quarter. Its revenue run rate, a measure used to estimate annual revenue based on current performance, exceeded $80 billion at the end of October, according to the Wall Street Journal. The company’s October revenue figure was left blank in its updated filing, making it another number for investors to assess.

Reuters reported that some prospective investors are waiting for Anthropic’s October financial results before pushing for a lower valuation. The roadshow was initially expected to begin during the week of November 9, with pricing anticipated the following week, Reuters previously reported. Anthropic did not comment on the valuation reports.

Compute spending draws scrutiny

Investors are also looking beyond revenue growth to examine the costs of developing and operating increasingly sophisticated AI models. The filing showed that Anthropic has committed approximately $895.4 billion towards AI computing power, chips and other infrastructure required to develop and operate its models.

Michael Burry, the investor known for his bets against the US housing market before the 2008 financial crisis, has expressed reservations about participating in the IPO. In a Substack post on Wednesday, Burry pointed to the computing commitments in the amended filing and argued that they could become difficult to justify even if the company achieves the revenue growth anticipated by optimistic investors. He said bullish investors expect Anthropic’s revenue to reach around $100 billion in the coming year, a figure that remains far below its long-term infrastructure commitments.

“Not only is the company priced for optimism,” Burry wrote, “it is betting its entire existence on that optimism.” He also raised concerns about take-or-pay agreements with major cloud providers, under which Anthropic could be required to pay for computing capacity even when it does not need it. Burry said he has no plans to pursue shares of Anthropic, and does not expect to be a shareholder anytime soon. In April, he had described Anthropic as “eating Palantir’s lunch” and suggested that the market was moving towards a plug-and-play system for integrating AI models.

A weak market for AI stocks

The market backdrop has also deteriorated. A sharp selloff in technology stocks has raised questions about whether investors will support a $1.5 trillion valuation for an AI company at a time when enthusiasm around the sector is weakening. The Global X Artificial Intelligence and Technology ETF fell 6.2 percent during Thursday’s session. Democratic victories in the US midterm elections have raised concerns about potential resistance to new data-centre developments, and analysts at Jefferies cut their forecast for new data-centre capacity in 2030 by approximately 8 percent in a research note on Thursday.

Malik Ahmed Khan, senior equity analyst at Morningstar, does not believe the current market weakness will derail the IPO. “We don’t expect any near-term impact on the IPO,” Khan wrote in a research note on Thursday. He added that concerns surrounding data-centre development could become a dynamic to watch over the next couple of years.

Others are less certain. Jim Lebenthal, chief market strategist at Cerity Partners, said in a Barron’s report that the company could reconsider its plans if the market environment prevents it from achieving an attractive valuation, and that it could instead raise capital in private markets if the public-market valuation fell too far. HSBC’s Max Kettner, who recently reduced his equity rating from maximum overweight to neutral, said a valuation closer to $1 trillion could be more acceptable to investors than the previously discussed $1.5 trillion figure. He cautioned that the company would need to avoid a sharp decline immediately after listing.

The pullback in AI valuations is not limited to Anthropic. OpenAI also cut its funding target to about $20 billion at a $1.25 trillion valuation, according to the Financial Times. A lower valuation for Anthropic could improve investor appetite and reduce the risk of a weak market debut, but it could also mean raising less capital than first planned.

Reporting based on coverage by Invezz, the Wall Street Journal and Reuters.