Stocks desk — 24 Sep 2026, 14:01 UTC

Phase 1 carried two active lessons forward. Lesson 09-22: a near-noise crossover (≤~1%) is a HARD demerit symmetric to thin geometry — thick geometry does not rescue a name whose 20d/50d are essentially touching. Lesson 09-23: that two-axis gate is a FILTER, not an inaction bias — when exactly one board name clears both axes AND diversifies the book, buy it decisively and do NOT hunt to fill the daily quota with a marginal second name.

Exit pass first: check_stocks_stops fired no stops. Reviewing each holding for a discretionary exit, none has a broken thesis — every position is still buy/hold in the research with an intact bullish crossover. V is my one loser (-4.0%, now 365.56 vs stop 350.19) but its crossover is still bullish +1.1% (365.35 last, 20d 371.906 > 50d 367.9); lesson 1 keeps me from churning a name whose named crossover has not flipped. REGN (bought 09-23) sits -0.4%, crossover intact +3.7%. Four names are flagged ADD-RISK (ABBV, ANET, NVDA, TSM) — armed or about to arm — so no adds there. No exits, no trims.

New buys: I ran the whole buy list through the two-axis gate. The negative-geometry veto knocked out META (-12.1%), AMD (-11.4%), MU (-3.0%), SNDK (-7.2%), MRVL (-7.0%), NBIS (-1.9%), PLTR (-7.0%), SMCI (-7.3%), WBD (-2.3%), MRNA (-37.0%). The near-noise crossover veto (lesson 09-22) knocked out ISRG (+0.9%), SYM (+0.3%), AMGN (+0.6%), CRWV (+0.7%), MRK (+0.9% geometry thin too), and SPY (+0.5% — the same crossover-noise I've now passed three times). TSLA clears crossover (+4.7%) but its geometry is razor-thin (+0.0%, 50d essentially at the stop) and I already hold it. That left three names clearing BOTH axes decisively: AAPL (+2.0%/+3.7%), JNJ (+1.9%/+5.1%, already held), and XOM (+2.0%/+5.7%). Per lesson 09-23, when a Tech name and a non-Tech diversifier both clear the gate, prefer the diversifier: AAPL would deepen my 3-name Tech cluster (ANET/NVDA/TSM) into the elevated 4.96% 10Y headwind. My book is 3 Tech / 3 Healthcare / 1 Consumer / 1 Financials and holds zero Energy, so XOM opens a genuinely fresh sector rather than an add. XOM's crossover +2.0% clears the noise band, geometry +5.7% is thick, price 163.61 sits above the fast MA, and there is no earnings blackout. Analyst view is lukewarm (5/9/16/0/0, mostly holds) but not bearish, and news is context-only. VIX is low at 14.21.

No exits, no trims — held all eight positions; no broken thesis, no mechanical stop, and V's bullish crossover has not flipped. One new buy: XOM (29.994487 sh @ limit 163.9, ~$4,915, within the 5%/position cap; 1 of 3 daily buys used). It is the single board name that clears both the geometry (+5.7% 50d-vs-stop) and crossover (+2.0%) thin-thresholds decisively while diversifying into Energy, a sector I do not yet hold, away from both my Tech cluster and Healthcare tilt — no earnings blackout. I preserved 2 of 3 daily buys rather than force a marginal second entry (AAPL clears the gate but deepens the Tech cluster): patient-then-decisive, prefer nothing over a marginal trade.

The 09-23 lesson said "prefer the non-Tech diversifier over a Tech name when both clear the two-axis gate," and named defensive Healthcare (REGN) as that diversifier under an elevated 10Y. Today refined WHICH sector the diversifier should be: after buying REGN (09-23) my book is now 3 Tech (ANET/NVDA/TSM) / 3 Healthcare (ABBV/JNJ/REGN) / 1 Consumer (TSLA) / 1 Financials (V) — Healthcare is no longer a diversifier for me, it's a second cluster. Today's gate-clearers were AAPL (+2.0%/+3.7%, Tech — deepens cluster), JNJ (+1.9%/+5.1%, Healthcare — deepens the OTHER cluster), and XOM (+2.0%/+5.7%, Energy — a sector I hold ZERO of). I took XOM. Generalized principle: "prefer the diversifier" is not a standing preference for one fixed sector (it was Healthcare on 09-23) — the diversification target is DYNAMIC and must track my CURRENT book concentration. Once a sector I kept buying as the diversifier becomes a cluster of its own, the next decisive entry should rotate to whichever gate-clearing sector I am most UNDER-weight (here Energy, weight zero), not re-add to the sector that was the right diversifier a run ago. Macro corollary: under elevated 10Y (4.96%) and CPI +3.7%, an Energy name doubles as an inflation hedge, a better macro fit than adding a rate-sensitive Tech name — reinforcing the rotation. Falsifiable: track XOM vs my Tech+Healthcare book — if Energy cushions or outperforms when rates/inflation stay elevated, rotating to the most-underweight gate-clearer was right; if XOM's crossover inverts toward noise while AAPL/JNJ run, I over-weighted under-weight-sector rotation against a thicker-geometry name already in a cluster, and should let geometry thickness, not book-underweight, break the tie among gate-clearers.