The most consequential earnings season of the AI era kicks off next week, as Alphabet, Microsoft, Meta, Apple, and Amazon all report quarterly results between July 22 and July 30. Hanging over every call is a single number: the roughly $725 billion in combined capital expenditure that Amazon, Google, Meta, and Microsoft plan to pour into AI infrastructure this year — up about 77% from last year's record $410 billion. After a bruising start to the season elsewhere in tech, investors want proof that AI revenue is growing fast enough to justify the biggest corporate spending wave in history.
The Reporting Calendar
Alphabet opens the marquee week on Wednesday, July 22, after the US market closes. Microsoft — reporting its fiscal fourth quarter — and Meta follow on July 29, with Apple and Amazon closing things out on July 30. Nvidia, the company absorbing an outsized share of all this spending, reports later on August 26. Options markets are already pricing implied post-earnings moves of 5% or more for several of these stocks, a sign of how much uncertainty surrounds the results.
The mood music is tense. The season begins just days after IBM shares dropped more than 20% on an infrastructure earnings miss, sharpening scrutiny of every AI-adjacent revenue line. And on July 17, a surprise release from Chinese AI startup Moonshot — whose new Kimi K3 model reportedly rivals top offerings from OpenAI and Anthropic — rattled AI and semiconductor stocks worldwide, reviving memories of last year's DeepSeek shock and renewing questions about whether Western labs' enormous spending advantage is durable.
Where the Money Is Going
The capex commitments disclosed through recent quarters are staggering in aggregate:
- Amazon is projecting roughly $200 billion in capital expenditures for 2026
- Microsoft is tracking toward about $190 billion for the calendar year, including an extra $25 billion attributed to higher component pricing
- Alphabet has guided to $180 billion to $190 billion, and has told investors to expect a significant further increase in 2027
- Meta raised its full-year range to $125 billion to $145 billion, citing pricier components and additional data center costs
Analysts at Evercore and Bank of America now place 2027 combined capex above $1 trillion, and Goldman Sachs projects hyperscaler spending from 2025 through 2027 will reach $1.15 trillion — more than double what the group deployed across 2022 to 2024.
The Monetization Scoreboard
The bulls point to cloud momentum. In the first quarter, Google Cloud grew around 63% year over year and reached $20 billion in quarterly revenue, while its contract backlog swelled to $460 billion — roughly double the level at the end of 2025. Amazon Web Services posted $37.6 billion in quarterly revenue, and Microsoft's Azure-driven cloud segment reached $34.7 billion. Jefferies analyst Brent Thill has argued that "the AI economy is healthy" and dismissed the pessimists bluntly, saying "the bear thesis is garbage."
The bears counter with free cash flow. Because capital spending is dramatically outpacing revenue from still-nascent AI products, analysts warn Big Tech free cash flow could fall by as much as 90% in 2026. Amazon is projected to swing to negative free cash flow of $17 billion to $28 billion, while some estimates see Alphabet's free cash flow collapsing from $73.3 billion in 2025 to $8.2 billion this year. The debt markets are feeling it too: Alphabet has raised over $85 billion in debt over the past year, pushing its total balance past $100 billion, and Morgan Stanley and J.P. Morgan estimate the tech sector will need roughly $1.5 trillion in new debt over three years to fund the buildout.
Why It Matters
This earnings season is effectively a referendum on the AI economy. First-quarter reports showed the market is no longer grading on revenue beats alone — Meta fell 6% after raising capex guidance, while Alphabet jumped nearly 7% on cloud strength — meaning payback timelines, not spending ambition, now move stocks. The outcome will ripple far beyond the Magnificent Seven: chipmakers, power utilities, data center builders, and the startups raising money on the assumption of ever-cheaper compute are all downstream of whether these four companies keep their foot on the accelerator. A confident quarter entrenches the buildout; a wobble could mark the first real check on AI capex since the boom began.
What to Watch
Watch three things on each call: cloud growth rates against the Q1 benchmarks of 63% for Google Cloud and roughly 31% for Azure, cloud operating margins as depreciation from new data centers starts to bite, and any revisions to full-year capex guidance. With 2026 spending estimates already creeping toward $800 billion to $900 billion in some analyst models, even the direction of the revision may matter more than the number itself.
