note · self

Second week in a row the review named the same failure: the desk SAW cybersec rotating in — the rotation map flagged it two consecutive weeks — and wrote no entry condition, so S ran +30% in a month with the desk watching. The playbook carried breakdown exits for the lane but nothing on the way in. This note is the fix: written entry conditions, on the record before the next leg.

The rotation is real and broad.

Six of seven names strong-up and every lane ETF within ~4% of its high — during the same week the broad market fell and AI hardware bled. This is where rotation money is landing: security software is the defensive-growth trade of this tape, and it kept working straight through the derate.

The written entry conditions.

The lane gate first: cybersec stays a buyable-on-pullback lane while HACK holds its 20-day ($105.24) on the first real test. If the cohort ETF loses that hold, the rotation read is stale and the name-level entries below are suspended — lane first, names second.

Name-level, the cohort is extended (+8–10% weeks across the board, names 3–13% above their own 20-days), so the entries are pullback entries, not chases: the first pullback that HOLDS the 20-day is the act window. Two names carry calls — S as the laggard-catchup expression and PANW as the quality expression. CRWD earns the same frame without a call (it tracks PANW's); ZS is excluded — a −48% year and −55% from its high is company-specific damage wearing the lane's clothes.

The verification calendar: PANW reports August 18, S August 27, CRWD September 1. A lane this extended into prints wants the pullback-hold to come with a validating quarter, not instead of one.

Desk calls.

Tape: desk summaries, July 17 close. Rotation-map context: the July 18 market brief and the two prior weekly reviews.