Starting tomorrow, five of the Mag 7—Microsoft, Alphabet, Meta, Amazon, and Apple—will release their Q1 earnings within three days.
This is the most critical earnings week of 2026, bar none.
The Numbers First
Here is how much the four hyperscale computing giants plan to spend on AI infrastructure this year:
| Company | 2026 Capex | 2027 Estimate |
|---|---|---|
| Microsoft | ~$146B | $170B |
| Alphabet | $175-185B | ~$200B |
| Meta | $115-135B | ~$142B |
| Amazon | $200B | ~$209B |
| Apple | $13.5B | $15.4B |
Combined, the four largest spenders will invest $645 billion in 2026, a 56% year-over-year increase.
Apple is in a smaller tier, not primarily a hyperscaler. But the annual capital expenditure of the top four already exceeds the combined 2024 defense and education budgets of the entire United Kingdom.
What Makes This Time Different
In the past two earnings seasons, analysts asked "how much will AI cost?"—worrying about insufficient spending, Nvidia supply constraints, and slow data center construction. That was Phase One.
Then came "can we afford this AI spending?"—concerns about excessive capex and long payback periods. That was Phase Two.
A Saxo analyst put it bluntly:
The market can still forgive heavy spending. But the market may not forgive vague spending.
That is Phase Three. Investors want to see how capex turns into revenue, margins, and monetization. Storytelling is no longer enough; they need data.
What Each Company's Investors Are Watching
Microsoft is the most critical. Its annualized AI revenue target is $25 billion. That means every $1 of AI capex must generate $0.17 in AI revenue. Azure growth is expected at 38%, and Copilot enterprise adoption is the key metric.
If Copilot adoption data is again vague, Microsoft's stock is likely to suffer on earnings night.
Alphabet has a more complex story. Its Search business is being cannibalized by its own AI Overviews. YouTube's AI-driven ad monetization must step up, and Cloud is expected to grow 50% in Q1. But the $185 billion capex burden means free cash flow will keep shrinking, making the return on capital look unattractive.
Meta is spending $115 billion and must show a clear improvement in ad conversion rates. The market expects ad prices to rise 12% and ad impressions to increase 16%. If both targets are missed, analysts will ask a direct question: where is the marginal return on this capex?
Amazon relies on AWS. The market expects AWS's AI workload contribution to accelerate while its operating margin holds steady. This quarter, AWS's operating margin has become a proxy for AI capex conversion efficiency.
Apple is in its own category. Its AI investment is an order of magnitude smaller. The market cares about whether Apple Intelligence is actually driving iPhone sales. China iPhone sales data will be scrutinized under a microscope.
The Real Inflection Point
This week's earnings will not change the direction, but they will calibrate the pace.
If all four hyperscalers deliver "capex up, AI revenue sharply up," the AI infrastructure narrative can continue into next year.
If any one of them shows signs of "capex up, revenue not keeping pace," the entire valuation anchor for AI stocks will loosen. The Mag 7 now account for nearly 35% of the S&P 500's weight. Even a small shift represents trillions of dollars.
Nvidia's earnings are last (late May), but its fate is already being decided this week. If the top five are unwilling to keep spending, everyone downstream will have to adjust.
The next 48 hours are exam time.
Sources: Mag 7 Earnings Preview April 2026 (Saxo); Mag 7 earnings preview: Can Big Tech turn AI spending into earnings growth? (FXStreet); CocoLoop; Big Tech faces key earnings test on AI spending and cloud growth (Yahoo Finance)