AI stocks face a tough earnings season as valuations fell sharply. Trivariate Research noted, “Only 5.3% of the time since 1999 have we seen more multiple contraction for the most expensive decile than we just saw in Q3.” Investors now worry about high financing costs and whether AI revenues will grow.
The market’s most expensive stocks—including some AI favorites—have experienced unusually sharp valuation declines, raising the stakes for earnings season.
Wall Street is expecting another quarter of strong earnings growth, but Trivariate Research warned that this growth doesn’t necessarily translate into stock gains, especially for the most expensive names.
The analysts found that the most expensive stocks on a forward price-to-earnings basis at the end of the second quarter saw multiple contractions roughly 100 times as large as those of the cheapest names. AI trade favorites like AMD, Marvell, Corning, and Bloom Energy were part of this group.
“Only 5.3% of the time since 1999 have we seen more multiple contraction for the most expensive decile than we just saw in Q3,” Trivariate Research said.
Valuation declines extend beyond the most expensive stocks, though they did see the most dramatic re-ratings. Despite this year’s strong earnings growth, the S&P 500’s forward price-to-earnings ratio fell to 19 at the end of last week from 22 at the start of 2026, according to Dow Jones Market Data.
One explanation for this dynamic is that investors’ faith in the AI trade is waning. Dauvin Peterson, 22V’s data and infrastructure analyst, said the stocks most exposed to AI capital spending are getting “very little credit” for their expected future earnings, translating to compressed valuations.
Melissa Brown, the global head of investment decision research at SimCorp, says that while concerns about the AI trade are mounting, high yields are the main driver behind recent valuation declines—and are just hitting AI-related companies harder.
“Rates will impact all companies’ ability to finance their operations, but AI companies are uniquely vulnerable as they currently have significant financing needs to build out their infrastructure,” the strategist explains.
“Questions are arising about when or whether revenues will be sufficient to cover these now higher expenses. And of course, since they started higher, they had a longer way to fall as compared with other companies.”
U.S. Treasury yields have climbed to multidecade highs. The 10-year yield hit its highest level since 2002, beating the previous record made in 2007.
With yields high, investors are less willing to pay a premium for future earnings. In other words, more speculative names are trading at a serious discount relative to their long-term earnings potential.
That hasn’t always been the case in this bull market. Russ Ivinjack, global CIO at Aon, says in recent years investors have been willing pay for future growth even if it meant looking through weaker near-term fundamentals. But persistently higher yields are making investors more demanding, he says.
“Companies can still command premium valuations, but they increasingly need to prove why. This broader dynamic is contributing to multiple compression across the most richly valued parts of the market.”
The bar is high this earnings season, especially for AI names. Companies will need to deliver strong results and convince investors that their growth prospects justify their valuations. After a massive run-up during this bull market, AI stocks have more to prove.