J.P. Morgan's Stephanie Aliaga on Why the AI Cycle Is Still Early and the Sectors She Likes

Stephanie Aliaga, a global market strategist at J.P. Morgan Asset Management, joined Phil Rosen on Full Signal to discuss where the AI cycle stands, the selloff in neocloud stocks, hyperscaler moats, frontier lab economics and why AI isn't showing up in productivity data yet.

Her core takeaway is that the AI boom is still early and far from a bubble bursting, but the opportunity set is widening well beyond the infrastructure winners of the past three years.

"We are still early, believe it or not," Aliaga said, though the strategy that worked two or three years ago may not work best next year.

The US hyperscalers and infrastructure names have led so far, and she said "they no longer have a monopoly on the AI opportunity set."

Strong demand and a largely cash-financed build-out by highly profitable firms give this boom "a real air of resiliency," she said. For a different read on bubble risk, see Sparkline Capital's Kai Wu on the AI bubble.

Rosen pointed to neocloud stocks like CoreWeave and Nebius, which he said are down about 30% in the last month or so. "In this AI race, not all balance sheets are created equal," Aliaga said.

Without the hyperscalers' massive revenue streams, neoclouds borrow heavily, and higher rates "do challenge the economics" of those players, she said.

"These companies are riskier," she said, though she's unsure how much of the swift move reflects fundamentals versus sentiment.

The hyperscalers are expected to grow earnings by 60% this year yet have underperformed the broad market, Aliaga said, with valuations about on par with the market or at a slight discount.

She sees a lot of demand visibility. The combined cloud backlog for the top three hyperscalers is now $1.7 trillion, she said, compared with roughly $800 billion of annual capex this year for the top five.

Asked if their moats will last, she said "I think so," at least for the foreseeable future, because "you have to be big to play in the world of AI."

Rosen said he has personally gotten a little more bearish on Anthropic and OpenAI as businesses. Aliaga said pricing is still being discovered, and said the labs' inference gross margins look "really solid," at around 70% to 80% based on financials her investment team has.

Cheaper models expand the total addressable market for model usage, she said. And just as the world has a reserve currency, "we may have a world reserve model," with US labs holding a moat on where enterprises spend their money.

Real-economy adoption is "a lot more pedestrian," Aliaga said, citing a KPMG report that found barely one in 10 large US companies have scaled agents over multiple business functions.

About 60% of S&P 500 firms are talking about AI on earnings calls, she said, "but just 2% can point to an actual tangible profit impact." In her words, "That doesn't mean AI is failing. It means we're early."

J.P. Morgan's 2027 Long-Term Capital Market Assumptions expect AI-assisted work to grow from about 6% of US work hours today to 40% of tasks by 2040. That translates to a long-run labor productivity forecast of 2.2%, she said, versus about 1.5% before the pandemic.

"I think technology still shines bright," Aliaga said, though investors need to be "buckled up for volatility." She sees semiconductors as well positioned given their visibility into future demand.

She also likes infrastructure beneficiaries such as utilities and suppliers of networking or cooling equipment. Among adopters, she pointed to healthcare and financials, and to software, where she said "we're in the very early innings" of some names becoming critical enablers of corporate adoption.

On the less risky end, she said "fixed income looks really attractive these days." Abroad, she said "we think Europe and Japan scream really attractive," and called international value "the place to look."

Earnings back up the trade, she said, with the market expecting 30%-plus earnings growth this year. "We're not investing on hopes and dreams here," she said.

Aliaga's team publishes its AI research at JPMorgan.com/AIhub, including its paper on AI-conscious asset allocation.

Tune in to the full conversation with Stephanie Aliaga on YouTube, Spotify and Apple Podcasts.

For more on the AI trade, read Nomura Investment Chief Andy Goldberg on the AI Boom and Sectors to Watch, and subscribe to the free Opening Bell Daily newsletter for a daily briefing on stocks, macro and Wall Street.

Enjoying Full Signal? A quick rating on Apple Podcasts helps more investors find the show.

This article is for informational purposes only and is not investment advice. Views and stock opinions are those of Stephanie Aliaga and market figures are as stated by the guest at the time of recording.