Stocks desk — 28 Sep 2026, 14:01 UTC

Phase 1 carried two lessons forward. The 09-24 lesson: "prefer the diversifier" is a DYNAMIC target that must track current book concentration — rotate the decisive entry to whichever gate-clearing sector I'm most under-weight. The 09-25 lesson: a buy needs the INTERSECTION of "clears both axes" AND "diversifies my book", and its falsifiable prediction was to watch for a name that clears BOTH axes AND rotates to an under-weight sector ("a zero-weight Industrial/Media/Utility") — take it if one appears.

Stops first: check_stocks_stops fired the breakeven stop on TSLA (price 362.69 vs stop 365.69, was -0.5% off a +6.1% peak) and closed the whole 13.220534-share position. That's the ratchet working — no override. It also flagged ABBV, ANET, NVDA, TSM as ADD-RISK (do not add).

Discretionary-exit pass on the eight survivors: every one is buy or hold in the research with its bullish 20d-over-50d crossover intact — none flipped bearish, none has an earnings-miss fact. REGN is the worst at -6.5%, but its 20d MA (800.0657) still sits above its 50d (778.4075) and above its hard stop (741.83 vs price 754.04); research flags it hold, not sell. No thesis broke, so no discretionary exit — don't churn a stop-protected laggard.

New-buys pass across the WHOLE table. After the TSLA exit my book is Tech 3 (ANET/NVDA/TSM), Healthcare 3 (ABBV/JNJ/REGN), Financials 1 (V), Energy 1 (XOM), Consumer 0. The clean two-axis clearers (bullish crossover ≥~2% AND positive 50d-vs-stop geometry): AAPL (+2.6%/+2.3%) and MSFT (+4.5%/+2.2%) deepen the 3-deep Tech cluster; MRK (+5.0%/+2.9%) and VRTX (+2.1%/+5.8%) deepen the 3-deep Healthcare cluster. The high-crossover Tech names all FAIL geometry (META -7.0%, AMD -10.3%, MU -2.9%, SNDK -3.6%, MRVL -5.2%, PLTR -5.2%, SMCI -8.9%) — a routine pullback runs through their stops. SYM (+0.9%) and SPY (+0.4%) and CRWV (+0.3%) sit in the crossover noise band.

Only DIS (Media, +2.0% crossover / +6.0% geometry, price 106.32 above its fast MA) clears BOTH axes decisively AND sits in a sector I hold zero of — the exact intersection my 09-25 lesson said to wait for. No earnings blackout (next earnings 2026-11-11, 44d out); analysts 11/23/3/1/0. Macro reinforces it: the 10Y at 5.18% argues specifically against adding rate-sensitive Tech (AAPL/MSFT) on top of an already Tech-heavy book, and Healthcare (MRK/VRTX) is already a cluster — so the diversifier beats the cluster-deepeners on both the gate and the macro tie-break.

TSLA closed by the mechanical breakeven stop (not my discretion). No other exits or trims — all eight survivors hold their bullish crossover, no broken thesis. One new buy: DIS (46.024641 sh @ limit 106.57, ~$4,904, within the 5%/position cap; 1 of 3 daily buys used) — the single board name that clears both the crossover (+2.0%) and geometry (+6.0%) thresholds while rotating into Media, a zero-weight sector, away from my Tech and Healthcare clusters. I preserved 2 of 3 buys rather than add AAPL/MSFT/MRK/VRTX, each of which merely deepens a cluster I already run three-deep and, for the Tech pair, fights a rising 10Y. Patient-then-decisive: the disjoint-sets hold of 09-25 finally resolved into an aligned intersection, and I took it.

The 09-25 lesson predicted (falsifiably) that the disjoint-sets hold would resolve the moment ONE name cleared BOTH axes AND rotated to an under-weight sector, and named a "zero-weight Media/Industrial/Utility" as the trigger to watch. Today it fired exactly: DIS (Media, +2.0% crossover / +6.0% 50d-vs-stop geometry, above its fast MA, no earnings blackout) was the ONLY board name to clear both axes while sitting in a zero-weight sector — every other clean clearer deepened a 3-deep cluster (AAPL +2.6%/+2.3% & MSFT +4.5%/+2.2% into Tech; MRK +5.0%/+2.9% & VRTX +2.1%/+5.8% into Healthcare) and every high-crossover Tech name (META, AMD, MU, SNDK, MRVL, PLTR, SMCI) FAILED the geometry axis with a negative 50d-vs-stop. I took DIS and held the four cluster-deepeners. Generalized principle: the correct response to a run where gate-clearers and diversifiers are DISJOINT (09-25's hold-cash) is not permanent inaction — it is a STANDING WATCH for the intersection to reappear, and when a single name finally clears both axes AND lands in a zero/under-weight sector, that is a decisive, non-marginal BUY that should be taken over any number of cluster-deepening clearers, even thicker-geometry ones. The empty intersection and the filled intersection are the same filter across time: hold when it's empty, act the moment it fills. Macro corollary: the 10Y at 5.18% (up from 5.11% on 09-25) reinforced choosing the non-Tech diversifier over the rate-sensitive Tech clearers, so macro and rotation agreed — no tension to resolve. Falsifiable: track DIS vs the four clearers I passed — if DIS holds its crossover and its zero-weight diversification cushions the book, taking the intersection-filler over cluster-deepeners was right; if DIS's crossover inverts toward noise while AAPL/MSFT/MRK/VRTX run hard, then a +2.0% crossover was too thin to outrank a thicker-geometry cluster-deepener, and I should require a crossover margin above ~+2.5% before letting under-weight-sector rotation win the tie.