Key Theme of U.S. Earnings Season: Tech Giants' AI Capex Intensity and the Battle Over Returns
Nashnova编辑部
As Q2 earnings season wraps up, Big Tech's AI capital spending has surged into the hundreds of billions of dollars — but the market's one question remains: can revenue growth outrun the spend?
Same massive spending — why did the market reward Microsoft but punish Alphabet?
Alphabet raised its 2026 capex outlook to as high as $205 billion and signaled a "significant increase" in 2027. Its stock fell over 7% in a single day.
Microsoft held 2026 capex flat and guided a measured increase for fiscal 2027; Amazon said full-year capex could reach $220 billion due to rising memory costs. Both stocks rallied over 15% post-earnings.
This means → The market isn't against spending. It's against spending without a clear story on where the returns come from. Microsoft and Amazon matched their capex narrative to revenue growth; Alphabet did not.
How do analysts read this divergence?
D.A. Davidson head of tech research Gil Luria noted Alphabet's numbers were "very, very good" — but free cash flow turned negative while the company pledged even more capex. "They assumed good results would be enough. Clearly the market disagreed."
His verdict on Microsoft and Amazon was the opposite: "Investors felt these companies are responsible — spending is generating real returns, it's sustainable, and the party can continue."
In plain terms = The market's logic isn't "how much you spend." It's "can you convince me the money is well spent?" Companies that told a convincing return story saw their stocks rise; those that didn't got sold off regardless of headline results.
Where exactly is the market's tolerance line for AI capex?
Luria laid out a clear framework: revenue growth faster than capex growth + margins still expanding — meet both conditions, and the market gives room to keep spending.
He pointed to Microsoft Azure: cloud revenue grew 43% year-over-year in Q4, with next-quarter guidance at 45%. "As long as capex growth stays below that level, it's responsible investment."
This means → It's not the absolute dollar amount that sets the market's mood — it's the relative rate of growth. Revenue outpacing capex gets a green light; revenue lagging capex triggers a red flag, no matter how big the top line looks.
Free cash flow has turned negative — how risky is that?
Both Amazon and Alphabet posted negative free cash flow in the latest quarter. Luria warned: "As long as spending is funded almost entirely by operating cash flow and cash on hand, it's fine. Once companies lean more on debt financing, risk rises."
JPMorgan strategist Dubravko Lakos-Bujas struck a more optimistic note: although most hyperscalers are expected to remain free-cash-flow-negative in fiscal 2027, backlog-to-capex ratios and book-to-bill ratios are both rising.
In plain terms = Cash flow on paper is negative, but customers are lining up faster. This reflects a monetization pace that may be catching up with the spending pace — not a runaway cash burn.
When will AI investment dollars reach the real economy?
Laird Norton Wetherby CIO Ron Albahary argued the market's next checkpoint is the "handoff" — from hyperscaler capex to companies embedding AI in their own business models. He cited pharma as a test case: investors need to see AI spending show up in margin improvement and faster drug-development cycles.
Freedom Capital Markets' Paul Meeks noted the handoff is already underway: cloud-infrastructure firm CoreWeave jumped 19% post-earnings; rival Nebius surged 34%, with total contract value nearly quadrupling in its latest quarter.
This means → The AI investment cycle is moving from "giants building infrastructure" to "ecosystem companies monetizing AI." Whether this handoff succeeds is the core test for whether tech valuations can hold in the second half of the year.
Content is for reference only, not financial advice.