This earnings season is a chance for AI hyperscalers to get their mojo back

Investors can be attracted to themes and seduced by narratives. But there’s nothing as irresistible as a line on a chart that goes up very quickly. That simple fact may help explain the magnitude of the rotation within AI-linked stocks, where investors are eager to hitch their wagons to pockets of accelerating growth driven by supply shortages that force hardware prices skyward.

This week, much of Big Tech will deliver quarterly results, headlined by Meta and Microsoft on Wednesday.

If you’re a portfolio manager who came into 2026 wanting to maintain the same amount of AI exposure while focusing on the pockets within that theme with the best improvement in prospective earnings growth, well, that likely means you’ve been lightening up on the Magnificent 7 heavyweights.

Their size, high margins, and dominant positions in fast-growing markets have helped the hyperscalers outperform most US companies over the past decade, and that’s what allows for persistent capex outlays on such a grand scale.

One issue? Not all of hyperscalers’ AI spending supports immediate moneymaking opportunities. A lot of compute is still used for training The Next Great Model Update, rather than to support internal products or cloud capacity that can be sold to customers.

Earnings season could serve as fuel for a rotation back into megacap tech — whether that’s thanks to its successes or the inability of these new market hot spots to live up to sky-high expectations. Look no further than Intel, which fell 17% on Friday, for an example of what happens when a parabolic rally is followed by anything less than perfect results.

The Takeaway

Hyperscalers need more “hype,” or to deliver a clear sign about their “scale,” to continue producing earnings through this spending binge.

There’s a welcome sign that the Mag 7 is poised to rebound: last week, the cohort entered (and exited) “oversold” territory, based on the 14-day relative strength index, judged to be a positive technical development. The last time the group had been this washed out, based on this metric, was after onerous tariffs kneecapped the market in April 2025.