BACK TO REPORTSPUBLIC PREVIEW
AI Infra Margin Watch: Where the Next Compression Starts
A public preview of how we map compute demand, margin pressure, and the names best insulated from it.
This sample report shows the structure of a full ANLY research note: a short thesis, a catalyst map, and the gating logic that unlocks the full operating model.
CHECKING ACCESS...
BROWSE REPORTSPublic preview onlyFull markdown content hiddenSign in to save and unlock
WHAT THE FULL REPORT UNLOCKS
The member version adds the operating model, catalyst calendar, and trade setup behind this preview.
Free windows for fresh reports
Permanent saves on open
Stamps for archived research
CONTENTS
4 SECTIONS
Executive summary
The next margin break in AI infrastructure is less about demand disappearing and more about the cost stack moving from obvious hardware bottlenecks into less visible areas: power, cooling, networking, and financing.
What matters now
- Compute demand is still expanding, but the mix is shifting toward buyers that negotiate harder on price and utilization.
- The cleanest upside remains in names with long-duration contracts, software-like recurring revenue, or exposure to power-constrained supply.
- The first margin compression often shows up in guidance language before it appears in revenue.
What the full report adds
- A ranked catalyst map for the next 30 and 90 days.
- A simple framework for separating temporary capex pauses from real demand loss.
- The peer set and operating metrics that are excluded from this preview.
Preview boundary
Sign in to open the full research note, save it to your archive, and unlock the sections that turn this thesis into a trade setup.