On April 24, this journal argued the Magnificent Seven were spending like industrials but priced like software. Five earnings reports later, the thesis was half right. The market is repricing — but the sorting function is revenue visibility, not capex level.
On April 24, this journal published The Industrial Turn — a thesis with three specific claims. The Magnificent Seven were spending like capital-intensive industrials. They were still priced like software companies. And the suppliers of that industrial capacity were the better bet.
Five companies have now reported. The data is in. Here is the score.
The Results
Microsoft reported $82.9 billion in revenue, up 18 percent year over year. Capital expenditure hit $31.9 billion for the quarter, with full-year 2026 guidance at $190 billion. Azure grew 40 percent. The AI business surpassed $37 billion in annualized recurring revenue, up 123 percent. The capex-to-revenue ratio: 38 percent.
Meta reported $56.3 billion in revenue, up 33 percent. It raised its 2026 capital expenditure guidance to $125 to $145 billion, up from the prior $115 to $135 billion range. It disclosed no quantified AI revenue metric. The stock fell roughly 7 percent after hours. Capex-to-revenue ratio on the new guidance: approximately 60 percent annualized.
Amazon reported $181.5 billion in revenue, up 17 percent. Capital expenditure reached $43.2 billion for the quarter, $147.3 billion trailing twelve months. AWS grew 28 percent with $14.2 billion in operating income. Free cash flow collapsed 95 percent to $1.2 billion, down from over $25 billion a year earlier. Capex-to-revenue ratio: 24 percent.
Alphabet reported $109.9 billion in revenue, up 22 percent. It raised capital expenditure guidance to $180 to $190 billion. Google Cloud grew 63 percent. Cloud operating margin nearly doubled year over year, from 17.8 percent to 32.9 percent. The stock rose roughly 7 percent after hours. Capex-to-revenue ratio on the new guidance: approximately 42 percent annualized.
Apple reported $111.2 billion in revenue, up 17 percent. Services revenue hit a record near $31 billion. Capital expenditure remained far below the industrial-scale commitments of its peers. The stock dipped roughly 1 percent after hours on an iPhone miss.
What the Thesis Got Right
The first claim — spending like industrials — is confirmed. Four of five companies are guiding $125 to $190 billion each in annual capital expenditure. Combined, the Magnificent Seven are committing roughly $650 to $715 billion. Microsoft explicitly disclosed that $25 billion of its capex increase came from component pricing alone. Meta cited elevated component pricing as a factor in raising guidance. These are utility-scale capital programs.
The second observation — free cash flow compression — is confirmed. Amazon's trailing twelve-month free cash flow fell from over $25 billion to $1.2 billion while revenue grew 17 percent. This is the industrial trap in real time: revenue grows, but free cash flow evaporates into capital expenditure.
The third claim — suppliers benefit — is partially confirmed. The $25 billion Microsoft component cost increase and Meta's elevated pricing are direct transfers to NVIDIA, TSMC, and Broadcom. The capex numbers validate the supplier thesis on the cost side.
What the Thesis Missed
The thesis predicted the market would sort between suppliers and spenders. The market did something more precise. It sorted within the spenders themselves.
Alphabet reported the largest cloud growth of any hyperscaler — 63 percent — with operating margins that nearly doubled in a year. The market rewarded it with a 7 percent after-hours move. The cloud revenue is quantified, growing, and profitable. The capex is justified by a visible return.
Meta raised its capital expenditure guidance by $10 billion at the midpoint and disclosed no quantified AI revenue metric. The market punished it with a roughly 7 percent decline. The capex is enormous and the return is unquantified.
Same week. Same magnitude of capital commitment. Opposite market reactions. The difference is not capex level. It is revenue visibility.
Microsoft sits in the middle — $37 billion AI ARR growing 123 percent is quantified and real, but $190 billion in annual capex makes the market uneasy about the ratio. Amazon's AWS grew 28 percent with strong operating income, but the 95 percent free cash flow collapse overshadowed it.
The sorting function the thesis identified — industrial spender versus supplier — exists. But the market's primary sort is finer-grained: industrial spender who can prove the AI revenue is arriving versus industrial spender who cannot.
Apple as Control
Apple is the control case. It is not spending at industrial levels. Its capital expenditure remains a fraction of its peers'. The market still punished it — not for capex, but for an iPhone miss. Services revenue crossing $30 billion is Apple's version of revenue visibility. The market demands segment-level proof of strength regardless of capital intensity.
The Revised Thesis
The Industrial Turn was right that the Magnificent Seven have crossed into industrial territory on capital expenditure. It was wrong about what differentiates within industrials.
The market learned to read these earnings like industrial earnings in a single week. The question is no longer whether the capex is too high. The question is whether the revenue attached to it is quantified. Cloud revenue with disclosed growth rates and expanding margins justifies capital programs. Advertising revenue without an AI attribution layer does not.
Testable prediction: Alphabet outperforms Meta over the next six months. The sorting function — revenue visibility on AI capital expenditure — will persist as the market completes the repricing from software multiples to industrial multiples with a revenue-visibility premium.
The thesis was half right. The half it missed is where the next twelve months of returns will come from.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)