Inside the Shapeshifting AI Trade That Has the Stock Market Heading Higher Again

Nvidia CEO Jensen Huang and Alphabet CEO Sundar Pichai at a U.S. Senate artificial-intelligence forum in September 2023. (Andew Caballero-Rrynolds/AFP/Getty Images)

We’re going to need a few more nails.

Just one week ago, the death of the artificial-intelligence trade was the talk of the markets. The debate really began in October, when Meta Platforms released better-than-expected earnings but also increased the amount of money it would spend on data centers and other AI projects, causing the stock to slump.

Nvidia’s results after the close on Nov. 19 should have been the final nail in the coffin. The stock opened 5.1% higher the next day, after the company blew past Wall Street’s expectations, but finished down 3.2% on concerns that expected growth was dependent on cash that might not materialize.

But beneath all the Sturm und Drang, the AI trade had already started to reconfigure itself. Alphabet stock, unlike shares of its peers, rose on Oct. 30 after the company reported results, largely because it was actually making money by spending money—a novelty among the hyperscalers.

Since then, Google has released an update of its Gemini AI model, one that appears to have leapfrogged the competition, and investors have focused on Alphabet’s tensor processing units, the name for its custom chips, which have powered its AI tools. The stock surged this past week to become the third-largest company by market value in the U.S.—and may even be set to pass Nvidia and Apple for the top spot.

The trade, though, is about far more than Nvidia vs. Alphabet. Mizuho analyst Daniel O’Regan notes that the trade of the week had been to buy stocks linked to Alphabet’s AI efforts at the expense of those linked to Nvidia and OpenAI. The former include Broadcom, which works with Google to design its TPUs; TTM Technologies, which makes circuit boards; technology solutions company Celestica; and Lumentum Holdings, which makes optical and photonic products for the cloud and networking. They have gained an average of 16% this past week, while the four connected to OpenAI— Advanced Micro Devices, CoreWeave, Nvidia, and Oracle —have advanced 2.7%.

Whether that pair trade continues remains to be seen. “Consensus is that we are not out of the woods and there is going to be a lot of AI-related drama,” O’Regan writes. “HOWEVER, this trade is NOT a zero-sum game and there should be room for multiple winners (and losers).”

The market might get evidence of that sooner than it thinks. Melius analyst Ben Reitzes points out that large language models have a way of leapfrogging each other. So, while everyone is focusing on Google because of the launch of Gemini 3, they also have a new Grok from Elon Musk’s xAI and then an update of ChatGPT, both of which will use Nvidia chips, to look forward to.

“What should happen is that new, great models not only drive increased usage (great for inference), but also drive a lot of FOMO [fear of missing out] at competitors, who learn from it and spend even more to top it,” Reitzes writes. In other words, expect the arms race to continue.

Of course, the market has already spotted its designated tech losers—software stocks. On Wednesday, Workday, down 8.7%, was the worst performer in the S&P 500 despite beating earnings expectations. The reason: slower-than-expected subscription growth, even though AI seemed to provide a boost. Others in the sector, from Salesforce to Adobe, have also suffered under the weight of expectations—and the assumption that AI will slowly eat away at their businesses. Salesforce has dropped 32% this year, while Adobe has fallen 28%.

Many “software as a service” names—Salesforce, Workday, Docusign, and Atlassian among them—trade at steep discounts to their fundamentals, according to KeyBanc analyst Jackson Ader, with valuations that imply that their free-cash-flow growth will be in line with gross domestic product, not the double digits Wall Street expects. He thinks that is far too pessimistic. “This, in our opinion, is unlikely and presents a framework for upside moving into 2026,” Ader writes.

But not all software companies are cut from the same cloth. Goldman Sachs recently highlighted a number that could benefit from AI due to high labor costs, including Pegasystems and Teradata. “The combination of continued corporate AI adoption and growing concerns about the AI infrastructure complex has increased recent investor focus on the next beneficiaries of the ever-expanding AI trade,” writes Goldman strategist Ryan Hammond.

And even beyond them. As questions about the AI trade have grown, investors have rotated not just within the tech sector but outside it, too. Since mid-August, the biggest beneficiary of that has been the healthcare sector. The Health Care Select Sector SPDR exchange-traded fund has gained 10% in November, even as the Technology Select Sector SPDR ETF has slumped 5.4%. DataTrek Research co-founder Nicholas Colas chalks up the sector’s outperformance to concerns about AI, fears that the Federal Reserve wouldn’t cut interest rates in December, and the fact that healthcare had underperformed by more than 24 percentage points for the 100 days ended Aug. 19.

The sector has now recouped a large chunk of that underperformance over the past three months. And with AI making a comeback and the Fed likely to cut on Dec. 10, it could be set for a pause, Colas explains. “Now that the market has regained some of its confidence in a December Fed rate cut and Alphabet has rejuvenated investor interest in AI, healthcare is very likely to take a breather through year end,” he writes.

Let’s hope it isn’t a final breath.

Write to Ben Levisohn at ben.levisohn@barrons.com