Nvidia will dominate Big Tech’s earnings season again

The world’s largest tech companies should at least be getting used to that by now. More than a year into the so-called AI revolution, the most eye-popping changes have come to their cash-flow statements. Microsoft, Amazon.com, Google parent Alphabet and Meta Platforms, which owns Facebook and Instagram, parted with a combined $106.2 billion in capital spending during the first half of this year. That is up 49% from the same period a year earlier, with most of the increase going for chips from Nvidia and other infrastructure to power generative artificial intelligence services.

All of the aforementioned companies signaled plans three months ago to keep their foot on the gas. Thus, Wall Street expects their combined capital bill to top $60 billion in the quarter ended September, up 56% year over year, according to consensus estimates from Visible Alpha. Another sizable double-digit jump is expected for the December quarter, which would bring the year’s total outlay to about $231 billion, some 49% higher than 2023.

Expectations for next year aren’t exactly down to earth either. “In 2025, Cloud Capex spend will equate to the entire Apollo Space program in real terms," Morgan Stanley analysts predicted in a report earlier this month.

This particular moonshot is great news for Nvidia. The company’s data-center revenue is projected to surge 97% year over year to $28.6 billion for its fiscal quarter ending later this month. But a major return on that investment for Nvidia’s largest customers still seems in a galaxy far, far away. Only Microsoft has made limited disclosures about how generative AI services have contributed to revenue growth. But even those may not mean much this time around given changes to its financial reporting that were announced in August and take effect in the company’s fiscal first quarter report on Oct. 30.

“Investors’ models are not likely to be consistent, but that may be the plan in a difficult period when opaque results may be beneficial to Microsoft," wrote John DiFucci of Guggenheim in a report last week. He added that he believes demand for the company’s M365 Copilot, which uses generative AI to help users do things like analyze spreadsheets and create PowerPoint presentations, has been “lackluster."

The other companies have their own issues going on. A growing number of analysts think Amazon’s spending on an ambitious satellite program could bring the company’s operating income below Wall Street’s stated targets for this year and next, thus curbing the operating margin expansion the company has been delivering of late.

Google and Meta are believed to be enjoying a strong online advertising market, but Google is now under the cloud of the U.S. government’s efforts to break up the company, while Meta could be more exposed to a slowdown in ad spending by Chinese e-commerce companies like Temu and Shein. Colin Sebastian of Robert W. Baird also believes investors aren’t fully accounting for how blowout capital spending will raise depreciation charges against future earnings; he expects Alphabet, Amazon and Meta’s operating income for the fourth quarter to come in 1% to 4% below Wall Street’s consensus forecasts.

Apple, meanwhile, is grappling with what looks like a disappointing launch of its first AI smartphone, with several analysts estimating weak initial demand for the iPhone 16 family that went on sale last month. And while Apple isn’t engaged in the same capital expenditure race as the others due to its hardware-centric business model, it too sees generative AI as a massive opportunity and is spending accordingly. Apple’s research and development expenses are expected to hit a new record of $31.5 billion for the fiscal year ended September, up nearly 6% from the previous year despite an estimated revenue gain of less than 2% in that time, according to FactSet estimates.

Wall Street is still largely positive on the potential for AI to boost big tech’s fortunes; Microsoft, Amazon, Alphabet and Meta average buy ratings from 91% of analysts covering those stocks, according to FactSet. And a survey of more than 400 companies by Morgan Stanley found about 40% of the respondents who have adopted generative AI solutions say the returns on their investments are exceeding expectations.

“Nonetheless, few GenAI adoption projects have yet moved from proof of concept to production," the Morgan Stanley report read.

It remains to be seen if Big Tech’s AI investments will be a gold mine or money pit.

Write to Dan Gallagher at dan.gallagher@wsj.com