Billionaire investor Howard Marks lays out why valuing AI companies like Anthropic is so tricky

Howard Marks fair highlighted among the hardest parts of investing in AI companies.

Value investors akin him try to project a company's earnings five years from now to gauge whether a inventory is attractively priced today, he stated on "Bloomberg: The China Show" on Monday.

"I don't think we can do that alongside much of AI," Marks said.

He gave the example of Anthropic, the AI business rearward Claude that recently submitted to go public in what could position among the biggest IPOs in history.

"Can we really fool ourselves into thinking we cognize what it volition do five years from now?" Marks asked. "And yet, that doesn't average we don't desire to be in it."

Anthropic's cofounders wrote in its IPO prospectus that they "will be continually seeking to reinvent our business, releasing new classes of products whose aim is to dwarf the financial value of our existing ones."

They noted that advances in AI enabled chatbots in 2022 and coding agents final year, and this twelvemonth have been gradually making it viable to use agents for broader cognition work and specialized discovery.

Marks, the billionaire cofounder and co-chairman of Oaktree Capital Management, stated that in six decades of investing, he's never dealt alongside item akin AI "in conditions of its inscrutability, its novelty, its speed of progress, its autonomy, and its unpredictability."

However, he noted the AI boom's latest "stumbles" — specified as investors dumping Alphabet and Amazon inventory final week on concerns concerning their AI spending — have "somewhat burst the bubble of infallibility."

The Nasdaq 100 slumped by approximately 1,400 points, or nearly 5%, complete the way of Thursday and Friday final week. But the tech-heavy indicator is motionless up concerning 13% this calendar year, and has soared by about 160% since the end of 2022, from under 11,000 points to fair below 28,400 points.

Marks, who's been sharing knowledge and direction in his memos for complete 35 years, stated it's "very healthy" that investors are trying to authority their excitement and be a small additional "thoughtful" and "skeptical."

The distressed-debt capitalist stated that alternatively of worrying concerning how a inventory is performing on a sole day, investors should "put together a portfolio of things that volition activity in the lengthy run, and remain alongside it."

"There is no reward in the funding endeavor for hyperactivity," he said.

One of Marks' key influences, Warren Buffett, has championed long-term investing for decades. The conglomerate he built, Berkshire Hathaway, completed construction ample stakes in Coca-Cola and American Express in the mid-1990s, and hasn't touched them since.

Buffett, who stepped downward as Berkshire's CEO at the end of 2025 and as its chairman this September, issued a similar caution concerning the AI surge in a July interview.

"It's durable to discover values whenever everybody is preferring gambling," the legendary agreement hunter said. He additionally emphasized that so-called hyperscalers specified as Meta and Microsoft are spending hundreds of billions of dollars on microchips, data centers, and another AI infrastructure, describing that as "real money."

Another notable value investor, Michael Burry of "The Big Short" fame, has often warned that AI is a bubble destined to burst. He's made the case on his Substack that AI companies are overinvesting, and pointed to their assured accounting, hidden debts, and circular deals as red flags.