Skeptical Investors Are Hunting for Ways to Short the AI Frenzy -- WSJ

Dow Jones18:30

By Gregory Zuckerman

AI bears are stirring after a three-year slumber.

An obsession with artificial intelligence powered the stock market to dizzying heights. Now, it is pushing skeptical traders to search for ways to profit from what they feel will be an inevitable shakeout.

One hedge-fund manager is preparing for Nvidia's chip sales to fall. Another has an off-the-books wager against OpenAI, which is privately held. Some are shorting shares of Oracle, one of the tech giants in the race for AI dominance.

Yet others are betting against the mounting debt of big tech companies including Amazon.com and Alphabet, which are investing as much as $670 billion this year to build out their AI infrastructure. Bears see this as a safer route than betting against stocks, which are more prone to sudden price spikes that could squeeze their short positions.

"People are more comfortable shorting the hyperscalers now because they're sacrificing their free cash flow," says Michael O'Rourke, chief market strategist at JonesTrading, who counts himself among the AI bears. "That's a major shift and a major risk."

The bets come amid rising anxiety around how the AI buildup might ultimately play out, with investors torn between fears the technology won't live up to hype -- or worse, that it will succeed beyond imagination and devastate the economy. For bears, the bet is simply that tech giants won't ever see sufficient profits to justify their massive AI spending commitments and sky-high valuations.

Among the high-profile naysayers is Michael Burry, who foresaw the subprime mortgage crisis and recently compared the frenzy around AI to the dot-com bubble.

One way bears are surfacing is with bets against Oracle, which had more than 2% of its shares shorted as of Jan. 30, according to FactSet, up from about 1.5% a year earlier. That is still relatively modest compared with the 3.6% average for S&P 500 companies, but reflects worries about the company's plans to raise as much as $50 billion from investors this year to fund its AI build-out.

Investors are also skeptical about Oracle's ties to OpenAI, which is facing growing competition from rivals like Anthropic. Oracle has a $300 billion deal to sell computing power to OpenAI.

"An Oracle short is a bet against OpenAI," says O'Rourke.

Bank of America strategist Michael Hartnett has been urging clients to short the bonds of hyperscalers like Oracle, Meta Platforms and Microsoft. Some traders prefer to target AI debt because fewer individual investors play in that market, helping shield it against meme-stock-like surges that would squeeze bearish positions.

For now, bets against AI giants remain relatively modest, in part because there aren't as many ways to establish the type of big, bearish wagers that produced windfalls when past investment frenzies ended.

After suffering massive losses in the 2008 housing crisis, banks became wary of taking the other side of big bets, such as the ones on risky mortgages that netted hedge-fund manager John Paulson $15 billion. Investors also fear AI stocks could shoot up in price, crippling any short bets.

"I don't have the cojones for it," says Jack Ablin, chief investment strategist at Cresset Capital. "I'm not prepared for one to run up in my face in response to a positive headline."

Mark Spiegel, who runs New York hedge fund Stanphyl Capital Partners, recently got a taste of the risks. He shorted shares of Nvidia because he thinks its chip sales will slow as investor worries grow about the huge sums hyperscalers are investing in AI.

"It's inevitable that they cut back on spending," says Spiegel, who predicts the mere hint of such a shift will crash Nvidia's stock.

Spiegel ended his short on Monday after a small loss, but he is eyeing a new short position in the stock.

Some bears are finding other ways to place their bets.

Jim Chanos, the well-known short seller who closed his hedge fund but still trades and advises clients, has been shorting shares of Ormat Technologies. The renewable-energy company inked a recent deal to sell geothermal power to Google's growing operations in Nevada.

"They will most likely lose money on this deal," Chanos told his clients last week, citing the company's high costs. An Ormat representative didn't respond to requests for comment.

There aren't any shares to short yet for OpenAI, which is expected to go public later this year after setting off the AI frenzy with its ChatGPT in late 2022. The company was recently valued at $830 billion in its latest fundraising round.

That hasn't stopped Benn Eifert, who runs QVR Advisors, a San Francisco hedge fund. He has entered into personal wagers with tech professionals and others about OpenAI's eventual valuation, complete with legal contracts.

Eifert noted the amount of cash OpenAI is burning and the competition it is facing.

"There will inevitably have to be a big pullback in data center spend," he says.

If OpenAI's valuation tops $300 billion a year after the initial public offering, Eifert will lose millions. If the valuation is less, he wins millions -- the exact amount depends on how low it falls.

Write to Gregory Zuckerman at Gregory.Zuckerman@wsj.com

 

(END) Dow Jones Newswires

February 25, 2026 05:30 ET (10:30 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.

At the request of the copyright holder, you need to log in to view this content

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

We need your insight to fill this gap
Leave a comment