Show-Me Time — Crunch Hour for the Other Five of the Mag 7

A thesis read for the first night of the show-me window — companion to this morning's Morning 10, which has the tape and the four-print calendar. This is the frame behind it.

The market stopped paying for ambition

For two years the AI trade rewarded a story. Announce the datacenter, name the model, raise the capex line, and the stock followed. That reflex is gone. The market has quietly switched from paying for AI ambition to demanding a credible path to monetize the spend — and the tape has already handed out grades, well before a single one of this fortnight’s prints.

Set Tesla and Nvidia aside; the story is the other five. Two of them own the customer and got paid. Three are caught between spending like an infrastructure company and watching agentic AI question their own core business, and they were sold. This is what the twelve months into Tuesday’s close look like.

Read the seven one at a time

Apple is the calm one. It owns the customer and carries no capex spree to justify, so as long as no new interface displaces the smartphone and regulators do not break the App Store open, it can monetize services on top of installed reach. The market has paid that quiet certainty: +53.5% over twelve months. Alphabet is the same archetype under threat — it owns reach, and the debate is whether answer-AI disrupts search economics faster than Gemini and Cloud monetize the audience. It is the only crunch name that has already shown the monetization: Google Cloud grew 63% in the April quarter to $20.0B in revenue, with operating income of $6.6B against $2.2B a year earlier and backlog roughly doubling to over $460B. The tape paid it — +81.9% over twelve months — but cooled to +8.0% over six. The re-rate is done; now it has to keep proving.

Microsoft sits in the double squeeze: it spends like an infrastructure company while agentic AI questions per-seat enterprise software, its own core. It posted the worst tape in the group — −20.7% over twelve months, −22.7% over nine, roughly 48 points behind the Nasdaq-100 in a year — and it got there despite the numbers, not for lack of them. Azure grew 40%, Microsoft Cloud crossed $54B in the quarter, the AI business runs above a $37B annualized rate, and the stock was sold anyway, five of its last ten prints down more than 3%. Amazon carries a two-front version: AWS capex on one side, a destination storefront on the other that answer-AI could hollow out if agents do the shopping and skip the visit. AWS re-accelerated to 28% — the fastest in fifteen quarters, $37.6B in the quarter, a $150B run rate — so the cloud is fine; the question is the other half of the company. Up 8.8% over twelve months, 18 points behind the index.

Meta is the one with no articulated path at all. Ads integration is real, but it does not carry a $125–145B capex guide on its own, and the market is waiting to hear management name the rest. Three candidates are on the table: push AI deeper into the ad machine (real, but already priced), rent out the overbuild and become its own capacity landlord, or build the personal-assistant business at three-billion-user scale. Any one of them, named and costed, would be an answer. Until then the tape reads −8.4% over twelve months and −12.0% over nine.

The market pays the vendors, not the spenders

Here is the line that ties the seven together. Every one of them — Nvidia’s +24.3% included — is dwarfed by the semiconductor index over the same window: SOXX +128.0% in twelve months. The market pays the vendors of the capex spree, not the spenders. And the spree is not slowing: Alphabet’s 2026 guide sits at $180–190B, roughly double 2025’s $91.4B, with the CFO flagging a significant 2027 increase; Amazon points to about $200B; Meta to $125–145B. The four big hyperscalers together run near $725B in 2026, up about 77% from roughly $410B in 2025, with street estimates crossing a trillion combined for 2027. The inversion — vendor paid, spender doubted — holds until a spender shows the revenue that turns that outlay into a return. That is what show-me time means, and the grading window is nine days: Alphabet reports tonight after the close, consensus near $2.87 EPS on about $116.5B of revenue, though the capex line is the number that matters; Microsoft and Meta land July 29, Amazon and Apple July 30. Five prints, one question asked of each — not how much are you spending, but what does it earn.

C · members block

Into tomorrow

Tonight is the first show-me verdict, and the book is watching it from a half-built position. The two-part trigger is half in ink: the technical half printed at Tuesday’s SOXX close of 552.69, well above the 530–532 band the diary set as the line. The fundamental half is Alphabet’s capex guide tonight. A guide at or above the $180–190B path with a 2027 number attached completes the pair and converts the tactical starter into a full add at Thursday’s open. A guide-down does not stop the book — it means the add comes lower rather than at the open, because the semiconductor bid is a claim on the spend regardless of which spender earns the return first.

No adds today. The diary does not trade into the print; it reads the print and acts at the next open. What it is listening for tonight is narrow: the capex line against the $180–190B path, and whether management attaches a 2027 figure to it. Everything else in the release is context. The grading window runs nine days and the book only needs the first grade to size the rest.