AI's rising bill of expectations

We will learn much more about the AI theme this earnings season, especially how impactful the technology will ultimately be. But the backdrop for mega-cap tech stocks is starkly different from just four months ago. After narrowing considerably earlier this year, the earnings growth gap has now flipped for the first time since 2022, with the S&P 493 outpacing the “magnificent seven” at roughly 29% versus 18%. See the chart. The bar remains high. Consensus expects S&P 500 earnings to grow about 30% – marking a third consecutive quarter of growth above 20% – according to LSEG data. Meanwhile, scrutiny of the AI theme has grown. That shifting backdrop contributed to our decision to downgrade U.S. equities to neutral last month. Our conviction in AI fundamentals, however, remains intact.

Last week’s earnings highlighted a growing tension between strong demand for AI-linked goods and services and the rising cost of meeting it. Hyperscaler capital spending has surged to about $649 billion over the past year, from about $357 billion in the prior-year period, according to data from Bloomberg, leaving little cash surplus after capex. And while hyperscalers’ operating margins have edged up to 27.1% from 26.9% in the previous four quarters, incremental margins have fallen to 27.9% from 40.4% over that same period. In other words, each additional dollar of revenue is generating less additional profit, raising questions about whether returns can keep pace with soaring investment. Still, our analysis finds that companies with higher AI adoption outside the S&P 500 technology and communication sectors are on track for 3% higher earnings growth this quarter than lower adopters, based on evidence of AI use in regulatory filings. Cloud revenue growth at Microsoft and Alphabet also points to sustained momentum in AI demand. But whether that demand will be enough to justify the scale of investment is still an open question, one we plan to debate at our 2027 internal Outlook Forum this week.

Funding capacity in focus

As the AI buildout becomes increasingly debt-financed, leverage and differences in funding capacity are coming into sharper focus. Credit markets are becoming an important signpost of how much additional AI investment can be financed and at what cost. The Treasury rally alongside the tech sell-off this week also fits our view that the ebb and flow of AI investment expectations will drive competition for capital and bond yields around structurally higher levels.

Within equities, we’re selective with our AI exposures. We favor companies and listed infrastructure poised to benefit from the critical bottlenecks in the AI buildout, a key reason for our move back to overweight EM equities in September. We see that opportunity playing out in Asia. Recent results from chipmakers point to tight supply and strong AI demand. But staying nimble is crucial, as our EM positioning shows: we upgraded at the start of the year, took profits midyear and returned to overweight in September. Brazil’s post-election selloff bears watching, but it does not change our EM overweight.

Our bottom line

AI faces a tougher near-term test as investor scrutiny intensifies. But our conviction in AI fundamentals remains intact, keeping us selective on where we take AI exposure while favoring opportunities tied to scarcity in the buildout.

The Nasdaq Composite capped its worst month since March 2025 as concerns over returns on massive AI spending intensified. The key question now is the economic value of the AI buildout: what returns it generates and who captures them. We see a range of possible outcomes for growth, profitability and returns on this investment. U.S. Treasuries extended their rally last week, with 10-year yields falling nearly 6 basis points to 4.93% after the Fed kept rates unchanged, as expected.

We focus on October nonfarm payrolls for signs of a tightening labor market and what that could mean for inflation. We’d expect stronger wage pressures than in recent jobs reports, potentially adding to core services inflation, making the October jobs data key to assessing our view of price pressures.