- Next leg of AI-driven gains could belong to the companies building and operating the data centres rather than just the ones supplying the components.
The sell-off in US semiconductor stocks over the past two weeks isn’t a warning sign for the broader market — it’s evidence that the rally is finally broadening out.
That’s the message from Morgan Stanley, which told clients in a note that the AI trade is undergoing a significant rotation, one that could shift investor capital away from the red-hot chip sector and into a wider set of beneficiaries.
At the centre of Morgan Stanley’s thesis are the so-called hyperscalers — the deep-pocketed tech giants that have been pouring billions into data centre expansion to power the next wave of artificial intelligence.
Companies like Alphabet, Amazon, and Meta Platforms have committed staggering capital expenditures to scale up their AI infrastructure, a spending spree that has, until recently, disproportionately lifted semiconductor stocks.
But Morgan Stanley’s note, authored by chief US equity strategist Michael Wilson and his team, suggests the dynamic is changing. The brokerage flagged that there could be “more capex discipline in the near-term,” and pointed out that hyperscaler stocks have already absorbed their period of underperformance.
In other words, the market may have over-corrected on these names, and the next leg of AI-driven gains could belong to the companies building and operating the data centres rather than just the ones supplying the components.
How the numbers stack up
The rotation is already visible in the data. The Philadelphia SE Semiconductor Index surged 11 per cent in June, riding a wave of AI enthusiasm that appeared almost unstoppable. But in the two weeks since, the chip benchmark has shed over 11 per cent, giving back all of those gains and then some.
Meanwhile, the Roundhill Magnificent Seven ETF — a concentrated bet on Wall Street’s largest tech titans, including the hyperscalers — have clawed back some of the ground it lost during a heavy sell-off in June.
The timing is notable. Alphabet, Amazon, and Meta all faced sharp selling pressure in June, even as chip stocks hit new highs. That divergence has now inverted, with semiconductor names under pressure while the broader mega-cap tech complex stabilises. For Morgan Stanley, this is exactly what a healthy rotation looks like.
Beyond the chips: Rate expectations and oil
The shift out of semiconductors isn’t happening in a vacuum. Morgan Stanley also pointed to macroeconomic tailwinds that are reinforcing the rotation. Markets have been scaling back expectations for additional rate hikes from the Federal Reserve, a development that tends to favor growth-oriented sectors and rate-sensitive industries. At the same time, crude oil prices have been falling, easing input cost pressures across the economy and improving the outlook for consumer-facing businesses.
These twin forces — a friendlier rate environment and lower energy costs — are creating a more favourable backdrop for sectors that had been overshadowed by the semiconductor frenzy.
Where the money goes next
Morgan Stanley identified several areas that could attract capital as the rotation unfolds. Consumer discretionary stocks stand to benefit from easing inflation pressures and still-resilient household spending. Transportation shares, which are closely tied to economic activity and sensitive to fuel costs, could get a lift from lower oil prices.
And biotechnology, a sector that thrives in lower-rate environments due to the long-duration nature of drug development returns, is also on the brokerage’s radar.
Crucially, Morgan Stanley is not calling an end to the AI story. The firm’s research arm has previously estimated that nearly $3 trillion of AI-related infrastructure investment will flow through the global economy by 2028.
The point is not that AI is fading, but rather that the market’s approach to pricing AI exposure is maturing. The early, narrow rally concentrated in a handful of chip names is giving way to a broader recognition that value will be created across the AI stack.
What investors should watch
The key question for the months ahead is whether the hyperscalers can deliver the kind of revenue growth and margin expansion that justifies the spending. Wall Street has yet to see clear, consistent evidence that AI products will generate returns commensurate with the enormous capital being deployed.
As Morgan Stanley noted, greater capex discipline from the hyperscalers could actually be a positive catalyst — a signal that management teams are focused on returns, not just scale.
