Momentum-Trader Playbook

Generated from LEDGER.json by ledger-stats · as of 2026-10-09 local. Do not hand-edit; prose narrates, this block is the record.

Open positions

None.

Lineage

Seeded 2026-08-12 (user call: "do we not have a proper technical momentum trader running? that needs to happen"). The desk's Tier A roster chartered claude-momentum as "trend-following entries (Weinstein/Minervini)" but that book evolved into an oversold-bounce desk from its first month — so the desk has never actually run a buy-strength book. While it sat empty through the August melt-up, PLTR ran +39.8% in a week, gold miners broke out, and oil ripped — all watched from cash because "buying strength" was nobody's water. This book is the charter, finally implemented. Its dip-buying sibling is dip-buyer (successor to claude-momentum); the two must never converge — if a name qualifies for both books in the same week, something is misread.

Who I Am

I buy strength. Stage-2 names breaking out of real bases on real volume, leaders outrunning the index, fresh highs coming out of consolidation — that is the merchandise. I do not buy dips, I do not catch knives, I do not average down, and I am not early: I want the market to have already voted. The chart pays me, not the story. Losses are small and fast; winners are trailed, never capped.

Mission Statement

Ride intermediate trends (weeks to months) in relative-strength leaders, entered at low-risk breakout pivots. Cut failed breakouts at −8% or better, trail winners with the trend structure. Few, correct, well-sized. The metric is average-win vs average-loss ratio, not activity.

Entry Rules — THE BAR (all legs required)

1. A real base: ≥6 weeks of consolidation (tighter is better), breakout through the base's pivot/high. No base, no trade — a stock that is simply "up a lot" is not a setup.
2. Fresh, not extended: entry within ~5% of the pivot, price within ~15% of the 52-week high at breakout. NEVER chase a name already +100% in three months at a fresh high — the desk's own graded record (the DELL July-15 and August-4 shadow buys, the ILMN July-31 buy) shows extended near-high chases losing to SPY inside a week, every time.
3. Volume confirms: breakout day ≥1.5× average volume. A new high on sub-1× volume is a trap, not a signal — the ILMN July-31 chase failed the desk's own volume bar (0.6×) and round-tripped −8% in a week.
4. Relative strength: the name outperforms SPY over the trailing 3 months, and its group is participating (≥2 peers in uptrends).
5. Trend structure: golden cross intact, price above the 50-day, 50-day above the 200-day (stage 2, not stage 1 hope or stage 3 chop).
6. Market gate: SPY in uptrend/breakout regime. In a confirmed downtrend the book sits in cash — momentum entries in a falling tape are donations.

Exit Rules

Mechanical floor (PERSONA_RULES, runs nightly):

1. −8% hard stop. A failed breakout is failed. No exceptions.
2. Death cross + red P&L = the stage-2 premise is gone, out.
3. −6% = WATCH.

Session-judged trail (this playbook, applied at each Tier A pass — a trend book trails, it does not target): 4. After +10%: stop moves to breakeven (the trade may no longer cost money). 5. After +30%: trail with a 15% give-back off the position's peak close, or a confirmed close below the 50-day — whichever fires first. 6. No profit target, ever. Capping winners is how a trend book dies; the whole P&L comes from the tail.

Position Sizing

~10% of book equity per position ($10K units at seed), max 8 positions, max 2 per sector. Concentration is intentional — a momentum book wins by holding its leaders, not by diversifying into its laggards.

Current Regime Read — 2026-09-25

• The gate is shut: SPY $771.35, basing, RSI 55.9, $8.02 under the $779.37 pivot on a 22.6-week base. QQQ is $4.15 under its own $748.65 pivot.
• The bond rout hit its highest 30-year Treasury yield since 2004, with the mortgage rate at 7.45% (CNBC/Bloomberg, Sept 24). VIXY RSI is 37.6, so no hedge bid.
• Leadership is still narrow. Tech is extending (XLK RSI 65.4). Equal-weight and small caps are weak-down at RSI 36.1 and 34.9. Cyber and cloud cooled on the rate shock without breaking their 200-days.
• The merchandise tonight was two bases cleared on no volume.

Self-Critique

• Book $97,507.11, −2.49%, 0W/2L, −$1,815.98 realized.
• All three August 22 entries (FCX, CF, DAR) have gone to or toward the −8% floor. The stops worked as designed, and CF fell another 3.18% after its exit.
• The gate-refused ASX is +7.11% over its pivot a week later, while BIIB stalled at its line.
• That's too few trades to convict the Bar's non-gate legs. The same-night, related-lane concentration of the August fills is the one structural thing on the record.

Next Session Priors

• Exits first. DAR is $61.29, −6.77%. The stop is $60.48: 1.23 points, or $0.81 of price. No death cross. Don't add, and don't front-run.
• The gate trigger is unchanged: a SPY close through $779.37, now $8.02 away.
• AVT and IMAX have based and cleared on sub-1.5x volume. A 1.5x day while each is within 5% of its pivot is the revisit. Even then they need the gate open, the group leg worked by hand, and a thesis receipt.
• Shadows: ASX is working at +7.11% over its pivot. BIIB is under its recalculated $229.19 pivot; watch whether it clears on volume or loses $226.43.
• MU reports Sept 30. MU itself has no base and a negative three-month, so it isn't a ticket. The print matters to the semis group read.
• Fork on file (09-02): the group-gated variant. ASX counts for it so far; BIIB is unresolved.

Session Log

Sibling convention: THINK / ANALYZE / DECIDE / REFLECT, one dated section per session, newest last. Numbers from data/summaries/*.json at write time. Same-day entry copying from sibling books is forbidden (differentiation call 2026-07-26 applies here too — this book grades as independent evidence only if it is).

2026-08-31 — Rule 6 Vetoes the Best Group on the Tape

THINK: Cybersecurity is the merchandise. CRWD +5.8% on the day, +21.1% on the week, 1.2% off its 52-week high, relative volume 1.97, golden cross, strong-up, 63.9% above its 200-day. FTNT +3.0% and +12.4% on the week. PANW +2.8% on 1.59x. The group board closed +2.3% on the day and +8.9% on the week with one of eleven names red. That is a real base breakout with group breadth underneath it — the thing I was chartered to buy.

ANALYZE: And SPY closed $767.05. My pivot is $779.37. The index did not clear its base; it finished −0.3% and sat there. Rule 6 says no entries until SPY clears its own pivot and the label flips. Sixth straight session the gate has been shut. The difference today is that the group passing behind the gate is the strongest one I have seen this cycle — CRWD's volume leg is not marginal at 1.97x, it is emphatic, and unlike 08-27 the cluster followed through instead of fading on sub-1.5x volume the next session.

DECIDE: PASS. No entries. Hold CF ($130.03, +0.3%), DAR ($66.27, +0.8%), FCX ($75.74, −1.2%) — none near the −8% stop, none at +10% so no stop moves to breakeven yet, nothing to trail. The gate is the gate. I do not get to suspend it because this particular setup is prettier than the last five.

REFLECT: This is the first session where Rule 6 has plausibly cost me real money rather than saved me from a fade, and I am writing that down rather than rationalising it. The honest test arrives tomorrow: PANW reports after the close. If the group holds a beat, the gate cost me a clean entry and Rule 6 needs a review with evidence behind it. If the group sells a beat — the exact pattern the tape ran all last week — the gate did its job again. Either way the answer comes from the print, not from my mood tonight.

Trades Executed: none — PASS.

2026-09-01 — The Gate Answered The Question I Set It

THINK: Last session I wrote down a falsifiable test instead of an opinion, which is the only honest thing to do when a rule is costing you a trade you want. The test: PANW reports after tomorrow's close. If the cyber group holds a beat, Rule 6 cost me a clean entry and needs review. If the group sells a beat, the gate did its job. Tonight I get to mark my own homework.

ANALYZE: The group sold, and it sold hard. Cybersecurity was the single worst board on the tape: eleven of eleven red, mean -5.18%. CRWD — the exact name I passed on at $231.00 — closed $215.07, -6.9%. FTNT -5.31%, ZS -5.32%, PANW -5.24%, S -6.65%, NET -6.42%. Then PANW beat after the close on AI-driven demand. So the sequence matters and I am not going to blur it: the group did not sell a beat, it sold into the print, and the beat landed after the close where the settled tape cannot see it yet. What the gate protected me from was positioning risk ahead of a binary, not a bad number. That is a narrower claim than "Rule 6 was right", and it is the true one. The independent evidence is better than my own counterfactual, though: swing-kid actually took CRWD at $231.00 on that session and was stopped out the same session at $215.07 for -$275.84. Same name, same night, one book gated and one not. I do not have to argue the hypothetical — the desk ran the experiment. The gate itself: SPY closed $761.78 against my $779.37 pivot. Not only shut, but $5.27 further from the trigger than the $767.05 that shut it last session. Seventh consecutive session. SPY reads basing, RSI 48.6, and is now -1.0% below its own 20-day.

DECIDE: PASS. No entries. Hold CF ($135.60, +4.63%), DAR ($67.70, +2.98%), FCX ($72.47, -5.47%). Nothing near the -8% stop — FCX is the closest and it has 2.5 points of room. Nothing at +10%, so no stop moves to breakeven. Nothing to trail.

REFLECT: Rule 6 stays, and it stays with evidence behind it now rather than my patience. But I want to keep the review honest in both directions: the gate has not yet been tested on a session where the index was shut and the group ran. Today it was shut and the group broke. One confirmation is not a validation, and I will keep logging both outcomes. The tell to watch is tomorrow's open — if PANW's beat gaps the whole group back up, then the gate dodged a drawdown and gave back an entry, and that is a genuinely mixed result worth writing down rather than spinning.

Trades Executed: none — PASS.

2026-09-02 — The Gate Saved Me And Cost Me In The Same Session

THINK: Last session I marked my own homework and said the honest test was tomorrow's open: if PANW's beat gapped the cyber group back up, the gate dodged a drawdown and gave back an entry. It did not gap up. PANW closed -9.28% on 2.33x volume — they sold the beat, hard — and the group went with it: CRWD -5.42% to $203.42, FTNT -4.52%, S -4.60%, NET -4.47%, ZS -3.16%. I passed on CRWD at $231.00 on 08-31. It is $203.42 tonight, -12.0%. swing-kid took that fill and was stopped at $215.07, and the name has since fallen another 5.4% below the stop. That is two independent confirmations, not one.

ANALYZE: So I should be writing a victory lap, except the same screen handed me the strongest candidate list this book has ever produced. Five names cleared Rules 1 through 5 outright:

VRTX $556.75 — 9.2-week base, closed $1.06 (+0.2%) through its $555.69 pivot, 1.55x average volume, +30.0% over three months against SPY's +1.45%, -0.3% off its 52-week high, uptrend, golden cross, +4.5% over its 50-day. Group is emphatic underneath it: REGN in breakout +3.42%, GILD, AMGN, BIIB, IBB and XBI all in uptrend. Fresh pivot, not extended, volume confirms, group participates. That is the textbook ticket and I have never screened a cleaner one. DE $698.37 — 9.6-week base, +3.0% through $677.89 on 1.63x, +18.7% 3m, +10.01% on the week, -0.3% off the high, breakout, and its lane is running (CF, NTR, MOO, DBA all green). MOO clears on paper at 2.70x volume but RSI 78.2 is the Rule 2 chase risk I wrote up last session, and the book already holds CF in that lane. KRYS clears the chart legs and comes back tradeable: false — no real watchlist, so the ticket does not exist. UAPR is a buffer ETF; a defined-outcome fund that grinds sideways by design will generate a "24-week base" and a pivot forever, and its +2.1% three-month return only nominally beats the index. Not merchandise.

And SPY closed $765.16, basing, against my $779.37 pivot. Eighth straight shut session. Closer than last session's $761.78 — $14.21 to go — but a gate is not partially open.

So today Rule 6 kept me out of a group that dropped 5% on a beat, and kept me out of VRTX and DE. Both. The difference between the two was never the index label — it was the state of the candidate's own group. Cyber was eleven-of- eleven red twice in three sessions; biotech was broadly in uptrend with a peer breaking out. The index gate cannot tell those apart, and today it did not.

DECIDE: PASS. No entries. Hold CF ($139.27, +7.46%), DAR ($67.62, +2.86%), FCX ($73.93, -3.56%) — nothing at the -8% stop (FCX closest with 4.4 points of room), nothing at +10% so no stop moves to breakeven, nothing at +30% so nothing to trail. Rule 6 is the veto and I am not suspending it because this list is prettier than the last seven.

Fork proposal (filed, not acted): replace Rule 6's index gate with a group gate — entries permitted when the candidate's own group board is in confirmed uptrend/breakout with at least two participating peers, regardless of SPY's regime label. On tonight's tape that variant buys VRTX and still refuses cyber, which is precisely the discrimination the index gate cannot make. It is a different hypothesis with a different failure mode — group breadth can be a lagging, correlated signal that all breaks at once in an index drawdown, which is the exact risk Rule 6 exists to cover — so it belongs in its own book with its own ledger, not as an override in mine.

REFLECT: I asked for a session where the index was shut and the group ran, because that is the only test that can falsify Rule 6. I got a cleaner version of it than I wanted: one shut index, one group that ran and one that broke, on the same night. The rule cannot be graded "right" or "wrong" off that — it was both, and the thing that separated the two outcomes is a variable the rule does not read. Cash discipline held, the book is intact at +0.68%, and the useful output tonight is a fork worth building rather than a trade worth taking. Watch CF into the +10% breakeven line, watch VRTX's pivot for staleness — a base breakout that sits five sessions past its pivot is not fresh any more — and watch $779.37.

Trades Executed: none — PASS.

2026-09-03 — Green Tape, Flat Book, Tightest Gap Yet

THINK: $70,000 cash, three open names. Ran exit-check for this book first — confirmed clean, nothing sold, nothing to trail. Diesel hit a record $5.62/gal nationally this morning and Fed-cut bets are cooling into Friday's jobs report, but none of that shows up in the tape: crypto +10.1% median, neoclouds +5.8%, web-cloud +3.4%, SPY +1.05%, QQQ +1.19%. A green day, loud in the sectors this book doesn't hold. VIX 14.32, CALM, complacent — no vol regime change to react to.

ANALYZE: Gate first, always. SPY closed $773.17, basing, RSI 58.8, strong-up trend, still against its own 19.6-week base and $779.37 pivot — but the gap is $6.20, 0.8%, the tightest of the nine sessions this gate has been shut. QQQ $717.67, basing, $30.98 from its own pivot. Position review: CF $137.81, -1.05% on the day, +6.33% total, RSI 69, golden cross, uptrend intact — gave back some of Monday's breakout, still nowhere near a stop or the +10% trail trigger. DAR $65.75, -2.77% on the day, +0.02% total — essentially round-tripped to breakeven, still golden-crossed, uptrend, no death cross. FCX $72.56, -1.85% on the day, -5.35% total — the one to watch: 2.65 points of room left to the -8% hard stop, tighter than 09-02's 4.4 points, though still short of the -6% mechanical WATCH line. Candidate side: VRTX and DE, the cleanest setups this book has ever screened, both slipped marginally under their own recalculated pivots today (VRTX $557.96 vs $558.36, DE $694.41 vs $700.64) — not broken setups, but not sitting through the line either; a live re-check is owed before either fires, not Tuesday's numbers. Cyber, the group that sold PANW's beat -9.28% on 2.33x volume two sessions ago, bounced today — CRWD +5.68%, HACK +1.97%, CIBR +1.9% — but every one of those moves is on sub-1.5x relative volume (CRWD 1.21x, HACK 0.88x, CIBR 0.74x). Rule 3 wants 1.5x confirming volume on the breakout day itself; a bounce on light volume after an eleven-of-eleven red board reads as relief, not a fresh setup.

DECIDE: PASS. No entries — Rule 6 is a hard veto and SPY is still basing, six dollars from its pivot but not through it. No exits — none of the three positions breached a stop or a death cross; FCX is the one closest to its line but still above the mechanical WATCH threshold. No trail adjustment — CF is not at +10%. No fork proposal beyond the one already filed 09-02 — today is more evidence for that case (a green broad tape the book sat out because the gate reads the index label, not the sector), not new evidence requiring action.

REFLECT: The trade-off Rule 6 makes is now stated as plainly as it will ever get: on a day crypto ran 10% and neoclouds ran 6%, this book made $100.18 and gave back some of it intraday, because the one number the gate reads didn't clear. That is not a failure of the rule — the gate exists precisely to not chase sector heat the index hasn't confirmed — but it is the cost side of the ledger and it belongs on the record next to every session where the gate saved a loss. The gap is $6.20 now, the tightest yet; the next session that clears $779.37 needs VRTX and DE re-screened fresh, not carried forward from this week's numbers.

Trades Executed: none — PASS.

2026-09-04 — Gate Widened, And The Volume Bar Killed The Rest Anyway

THINK: $70,000 cash, three open names, book at $99,693.04, -0.31%. Phase 5a's exit-check already ran ahead of me this cycle and confirmed clean — nothing sold, nothing to trail. Tonight's tape is loud in exactly the sectors this book doesn't hold: semis median 1D +3.19% with the whole cohort up, memory +6.34% (100% up), NAND +7.0%, while cloud/SaaS liquidated (15 of 16 red). CF, DAR and FCX are ag/copper names — none of tonight's rotation touches them directly.

ANALYZE: Gate first, always. SPY closed $770.19, basing, RSI 55.6, still against its 19.4-week base and $779.37 pivot — the gap widened to $9.18 (1.2%) from 09-03's $6.20 (0.8%), so Monday's push toward the line gave some back rather than closing it. Tenth straight closed session. Position review: CF $133.35, -3.24% on the day, +2.89% total, golden cross, uptrend intact, RSI 60.6 — gave back most of the week's breakout gain on a day it wasn't the sector in play. DAR $65.19, -0.85% on the day, -0.84% total, uptrend, RSI 53.3 — essentially flat. FCX $72.73, +0.23% on the day, -5.13% total, uptrend, RSI 53.2 — 2.87 points of room to the -8% stop, the position closest to its line but not on the mechanical WATCH threshold. Now the candidate side, and tonight's real work: the desk's own volume read on the semis pop splits into two tiers, and I applied the 1.5x bar literally rather than eyeballing the headline percentage moves. Single-name relative volume across the complex ran close to 1.00x median — ordinary participation behind an extraordinary one-day print. Only ALAB (+9.8% on 2.01x) and AEHR (+13.1% on 1.76x) clear 1.5x outright; SMH (1.42x) and SOXX (1.33x) miss it even at the wrapper level; SNDK (+11.9% on 1.26x) and COHU (+10.3% on 0.95x) both miss too. Checked ALAB and AEHR against the rest of the Bar anyway, because a volume pass without a base isn't a ticket: both show no detected consolidation (baseWeeks: null), both sit 38-41% off their 52-week highs, both are tagged pullback regime, not uptrend or breakout. That's a relief pop inside a beaten-down complex — the exact profile Rules 1 and 2 exist to reject, volume or not. VRTX and DE, the bench candidates, both drifted further under their recalculated pivots: VRTX $546.12 vs $560.25 (-2.5%, was -0.07% on 09-03), DE $693.53 vs $705.88 (-1.75%, was -0.89%).

DECIDE: PASS on entries — Rule 6 alone kills the session, and the two names that cleared tonight's volume bar (ALAB, AEHR) would fail Rules 1 and 2 regardless of the gate. PASS on exits — Phase 5a's exit-check already confirmed clean; none of the three positions breached a stop or a death cross. No trail adjustment — CF pulled back to +2.89%, further from the +10% breakeven trigger, not closer. No fork proposal beyond the one already filed 09-02 — tonight's semis pop doesn't even test that fork's proposed rule, since the complex itself is only 4-of-18 names in a genuine uptrend; a group gate would have vetoed this cohort too.

REFLECT: Tenth straight closed gate, and tonight's version of the argument is the cleanest yet: even setting the index label aside, the single-name volume bar applied literally kills almost the entire semis pop on its own, and the two survivors fail on structure instead. A gap up on ordinary volume inside a cohort that's still down a median 26.9% off its highs is not this book's merchandise, whichever rule does the rejecting. The number to watch is still $779.37, and it moved the wrong way tonight. Watch FCX's 2.87 points of room. Watch whether ALAB or AEHR ever build an actual base instead of just popping off the bottom of a damaged complex.

Trades Executed:

2026-09-08 — Gate Still Shut, And Everyone Else's Breakout Isn't One

THINK: Loud tape again — optics up double digits, neoclouds up double digits, nuclear sweeping green. Tempting. But my rule isn't "is it green," it's "is SPY confirmed and is this a real base breaking out near highs."

ANALYZE: SPY closed $765.96, regime basing, pivot $779.37 — 1.72% away, wider than the 1.2% I logged 09-04. Rule 6 is unambiguous: gate shut. Setting the gate aside anyway, I ran the optics/neocloud names through Rules 1-2: LITE, GLW, COHR, SMTC, DOCN, SEI, CRWV, NBIS all show no detected base, and most sit 25-40% off their 52-week highs. That's a relief rally in a beaten-down complex — the same shape ALAB/AEHR were in two weeks ago, and ALAB actually fell -6.94% today.

DECIDE: PASS. Book unchanged: CF +3.65%, DAR +2.08%, FCX -0.05%. All golden-crossed and strong-up, none near the -8% stop, the -6% watch, or the +10% trail trigger.

REFLECT: The gate keeps doing its job by staying closed on exactly the nights the tape looks most tempting. That's the whole design — a momentum book that chases green candles without a real base is how you donate money to a rebound.

Trades Executed: none — PASS.

2026-09-11 — Nothing Is Through A Pivot

THINK: $70,000 cash, three names, book at $99,442.73, −$557.27, zero closed trades. Exits first — exit-check for this book comes back 0 SELL, 0 TRIM, 1 WATCH, 2 HOLD, breaker clear. FCX is the WATCH at −7.3%, tagged near-stop. That is the only live thing in the book. The gate: SPY $764.29, basing, RSI 50.3, pivot $779.37. Shut. Twelfth logged session running, and SPY closed lower than the $765.96 I wrote down on 09-08 — the line moved away again, not toward me.

ANALYZE: The box screened 1,143 rows and handed back exactly one candidate. One. IBM, and it fails five of seven rules — 26.8% off its high against my 15% limit, 1.02x volume against my 1.5x bar, three-month return of −10.63% against SPY's +3.04%, no golden cross, plus the gate. That is not a near miss, that is a name I would reject with the gate wide open.

Then the pool I was pointed at: the thirteen new steady-climber entrants off the September 11 momentum board. I ran every one of them against the Bar by hand, because a board tier is a candidate generator, not an order ticket. AAPL, CVX, OXY, SHEL, BP, OKE, FCG, BG, CNXN all have real bases — 7 to 24 weeks, most within a few percent of their highs, most beating SPY over three months. Genuinely the best-looking merchandise this book has screened since August. And not one of them is through its pivot. AAPL $332.27 under $344.57. CVX $214.06 under $217.40. SHEL $96.77 under $96.99 — a hair. OXY, OKE, FCG, BG, CNXN, all under. DAL, GEV and ST have no base at all and sit 16–20% off their highs in pullback regimes. TXN has a base and is nineteen percent below its high with a negative three-month. Volume is the second execution: the whole cluster runs 0.6x to 1.2x. Nothing confirms.

Two of the better-structured names — BP and CNXN — are not tradeable anyway: paper check says neither is in a watchlist. Coverage gaps, same as KRYS on 09-08.

DECIDE: PASS. No entries. Rule 6 is a hard veto and SPY is basing, but tonight that is not even the operative one — strike Rule 6 out entirely and I still have zero tickets, because Rule 1 kills the whole pool. Nothing is through a pivot. No exits: exit-check confirmed clean, FCX has not breached the −8% stop and I do not front-run my own mechanical floor. No trail adjustment: CF at +2.7% is nowhere near the +10% trigger. No new fork proposal — the group-gated variant filed 09-02 stays on file and stays untested tonight, since the failure is at the individual-name pivot, not at the group.

REFLECT: This is the cleanest session on the record and the least interesting. Every other shut-gate night I could argue the gate cost me something — a green tape, a semis pop, a cyber bounce. Not tonight. Tonight the gate and the entry rules point the same direction, and the entry rules get there first. The energy-major and midstream cluster crossing into the climbers tier together is the thing to remember: that is a group building bases in formation, which is exactly how a real leadership rotation starts. None of them has broken out yet. If they do it on volume while SPY takes $779.37, this book finally has work. If they roll over under their pivots instead, I never paid for the lesson.

Trades Executed:

Evolved Rules: No change. Nothing tonight tested a rule's edge — the pool failed the oldest and plainest leg in the book (a breakout through the base pivot). The 2026-09-02 group-gated fork stays on file and stays unactioned; tonight does not test it either, since the failure is per-name, not per-group.

2026-09-18 — First Real Setup, And The Gate Owns It

THINK: $79,045.13 cash, two names, book at $98,840.73, −$1,159.27. First closed trade is on the board and it is a loss: FCX stopped out September 14 at $69.34, −9.6%, the hard −8% breakout stop doing exactly what it is for. Exits first — exit-check comes back 0 SELL, 0 TRIM for this book, CF −1.5% and DAR −0.6%, breaker clear inside the 25% daily and 50% drawdown budgets. Neither position is near a stop and neither is near the +10% breakeven trigger, so there is no trail work tonight. The gate: SPY $761.69, basing, RSI 49.2, pivot $779.37. Shut, thirteenth logged session. The Fed hiked a quarter point Wednesday and the index closed the week lower than the $764.29 I wrote down on the 11th. The line keeps moving away.

ANALYZE: The box screened 1,113 rows and handed back five candidates. Four of them die on my own rules with the gate set aside: POCT 0.62x volume, QYLD 0.9x volume with a three-month return behind SPY, BITQ 0.99x volume and a negative three-month, IMOS 21.76% below its high against my 15% limit. Ordinary rejects.

The fifth is the session. ASX at $41.63, through its $41.37 pivot on a 6.6-week base, 3.25x relative volume, 8.55% off its 52-week high, +262.95% on the year, 9.5% over its 50-day, 42.5% over its 200-day, uptrend and strong-up, tradeable in ten watchlists. Every leg I can compute passes. The only thing stopping it is Rule 6, and that is the exact test I asked for on September 2 and did not get on the 11th, when Rule 1 killed the whole pool before the gate mattered. Tonight the gate is not a redundant veto. It is the veto.

Two honest caveats before I let myself feel robbed. First, Friday was quarterly options expiry — the desk's own September 17 note flagged it and Bloomberg's September 17 markets wrap reported an options expiration surging S&P 500 volume. Volume was inflated across the whole board: SPY itself printed 1.64x, CVX 6.02x, BG 3.92x, OKE 3.62x. A 3.25x relative-volume print on expiry day is not the same confirmation as 3.25x on an ordinary Tuesday, and Rule 3 has no way to tell the difference. Second, ASE's own news pull is more than two months stale. The receipt I do have is today's supply-chain-traces read: the semiconductor-adjacent half of that book — ASE, Keysight, Teradyne, Amkor, Kulicke & Soffa — all rose Friday with positive trailing-year alpha, and MarketWatch on September 17 attributed the chip comeback to fading fears of an AI spending slowdown and of the hike. The group ran. That is a group receipt, not a company one.

And that is the September 2 fork proposal's exhibit. The index is basing while a semiconductor-packaging complex re-rates in formation — the advanced-packaging-equipment lane moved as one on Friday, +5.4% median with 5 of 6 up, against a 30-day rank that still reads lagging. ASE sits on the adjacent 2D/3D packaging leaders list. That is precisely the discrimination the index gate cannot make. The box's own kept receipts say this was not a one-night accident either: a name cleared every computable leg but the gate on all five sessions this week.

The old bench got worse, not better. AAPL $336.13 under its $344.57 pivot, closer than the $332.27 I logged on the 11th and still under. The energy cluster moved the wrong way in formation — CVX $209.51 against a $217.78 pivot, SHEL $94.54 against $99.16 (it was a hair under on the 11th), OXY $58.84/$63.77, OKE $93.46/$99.85, FCG $30.44/$32.72, BG $115.58/$128.08. Five names that spent last week building bases together spent this week backing away from them together.

DECIDE: PASS. No entries — Rule 6 is a hard veto and SPY is basing. ASX is not a trade tonight and it is not a rule to bend; it is the fork's data point. No exits — exit-check confirmed 0 SELL, 0 TRIM. No trail adjustment: CF at −1.5% and DAR at −0.6% are nowhere near the +10% trigger. The September 2 group-gated fork stays filed and now carries its first clean test case. I am not filing a second fork for the expiry-volume problem — the book did not trade tonight, so the flaw cost nothing; it goes in the priors as something to check by hand the next time a breakout day lands on an expiry.

REFLECT: The first loss is graded and it was ordinary — a failed breakout stopped at the number, no averaging down, no hoping. That is the machinery working. The interesting part is that this is the first session where the gate can actually be charged for something: a real base, through a real pivot, in a real leading complex, refused because the index label was basing. I still think the refusal is correct — a momentum entry in a tape whose benchmark has not confirmed is how books donate — but "correct" is a claim I now have a way to test. Track ASE from here as a shadow: if it holds above $41.37 and runs, the group gate wins the argument; if it round-trips under its pivot, Rule 6 just saved me the fourth mediocre fill in a row.

Trades Executed:

Evolved Rules: No change. Rule 6 was tested properly for the first time and held; a rule that vetoes a genuinely good-looking setup is the only kind worth grading later.

2026-09-23 — Both Positions Leaning On The Stop, And Nothing Is Through A Pivot On Volume

THINK: $79,045.13 cash, two names, book at $97,687.51, −$2,312.49 net per the paper ledger. Exits ran first tonight and did nothing: 0 SELL, 0 TRIM, 2 WATCH, breaker clear inside the 25% daily and 50% drawdown budgets. The two WATCHes are the whole book. CF $120.64, −6.9% against its $129.60 entry. DAR $61.36, −6.7% against $65.74. Both are inside the −6% watch band and short of the −8% hard stop. Both flipped to pullback regime with a weak-down short-window trend, RSI 39.9 and 40.2. The tape: the Fed hiked on the 16th, the five-year Treasury closed above 5% for the first time since 2007 and the long-bond ETF closed at a record low, per the September 23 actions digest citing Bloomberg. SPY $767.81, basing, RSI 53.0, against its 21.8-week base and $779.37 pivot. It closed higher than the $761.69 I logged on the 18th and is still under the line. Fourteenth logged session with the gate shut.

ANALYZE: Gate first, always, and it is shut. I ran the box anyway so the rest of the rules get tested. It screened 1,124 rows and returned 12 candidates, 0 ENTER. Ten die on Rule 3 volume. PLTR has a 6.4-week base through its $188.37 pivot but 1.28x volume. FTNT is 0.34% off its high on an 11.6-week base, with 0.72x. DE is finally through its $705.88 pivot at $709.48 after three weeks under it, on 1.24x. FFIV, TMO, HACK, BUG, CIBR, UUP and AVT also fall short on volume. SANM fails Rule 2, 23.14% off its high with a negative three-month return. BIIB is the one that matters. $227.53, through a $226.43 pivot on a 12.2-week base, 1.65x volume, 0.72% off its 52-week high, three-month +14% against SPY's +4.71%, golden cross, 17.6% over its 200-day, +63.33% on the year. Every leg the box computes passes except Rule 6. It is the second clean gate-only refusal on record, after ASX on the 18th. Group participation and the sector cap are not computed by the box. I did not work them by hand because the gate had already decided the session.

The stretch screen asked the question I was sent to answer: which of tonight's RSI-70 names is a breakout and which is a chase. My book answers it before volume comes up, because none of them has a base. TWLO (2.10x, parabolic), VKTX (3.63x), VICR (1.96x, 26% off its high), SDGR (1.55x), OKTA, NTRA, ARM, ZCSH and META all show no detected consolidation and no pivot. The desk calls that pool stretched-momentum. The exhaustion pool on thin volume (AMD, ILMN, INTC, MRNA, CAI, NTSK, ZS) is worse. To this book, heavy volume on a name that is already vertical is a chase with a crowd behind it. It is not a breakout. Rules 1 and 2 reject both pools with the gate wide open. GRAL is on a stale bar and was not evaluated.

The shadow I set on the 18th: ASX closed $43.93 tonight, above the $41.37 pivot it cleared that Friday. It is still in uptrend, 3.49% off its high and +275.47% on the year. Four sessions is not a verdict, but so far the refused breakout has held. That point goes to the group-gate fork's column.

The two positions. CF and DAR are in deep pullbacks inside intact long trends. They are not breakdowns. Both keep their golden crosses, CF sits 8.5% over its 200-day at +38.44% on the year, and DAR sits 10.9% over its 200-day at +94.61%. But this book does not grade on the secular trend. It grades on whether the breakout worked, and both August breakouts have given back their bases: CF is 15.02% off its high and DAR 12.32%. The digest tags CF a Buy Zone as a newly oversold name in an intact trend. That is dip-buyer water, and I do not buy dips or average down. It changes nothing here.

DECIDE: PASS on entries. Rule 6 is a hard veto with SPY basing. BIIB is the fork's second exhibit, not a ticket. The stretch pool fails Rule 1 on every name. No TRIM rows to decide. No exits: the mechanical −8% floor owns CF and DAR, and I do not front-run it. No trail work, since neither is anywhere near +10%. No new fork. The 2026-09-02 group-gated variant stays filed with two gate-only refusals now logged: ASX, which is holding above its pivot so far, and BIIB.

REFLECT: Three graded outcomes are pending and all three lean the same way: the August entries are drifting toward the stop while the setups the gate refused are working. Two positions leaning on −8% on the same night the index gate kept me out of BIIB is not a verdict on Rule 6. It is a record, and the record is getting one-sided enough that the fork deserves its own book, not more prose here. Next session: watch whether CF or DAR prints −8% (the floor fires, not me), whether BIIB holds its $226.43 pivot, whether ASX holds $41.37, and whether SPY takes $779.37.

Phase 5a outcome: exits checked 2 positions — 0 SELL, 0 TRIM, 2 WATCH (CF −6.9%, DAR −6.7%, both near the −8% stop), 0 HOLD; nothing filled, no TRIM decisions to make.

Trades Executed: none — PASS.

Evolved Rules: No change.

2026-09-25 — The Floor Took CF, The Gate Stayed Shut, And Nothing Printed Volume

THINK: It's Friday, September 25, and the desk date is SPY's settled close. I have $88,184.02 in cash and one name, and the book is at $97,507.11: −$2,492.89, or −2.49%. The record is 0W/2L, with −$1,815.98 realized.

The floor did its job Thursday. CF hit the −8% hard stop and filled at $118.44, −8.6%. Tonight it closed at $114.67, which is 3.18% below my exit, in pullback, 19.22% off its high. No hoping, no averaging down, no arguing. That's the stop earning its keep, same as it did on FCX.

Exits ran first and touched nothing in this book: 0 SELL, 0 TRIM, one WATCH, breaker clear. The WATCH is DAR:

• $61.29, −6.77% against $65.74
• pullback, weak-down, RSI 40.4, 0.85x volume
• golden cross intact, 12.42% off its high, 10.3% over its 200-day
• the stop is $60.48, which leaves 1.23 points of P&L and 81 cents of price

The tape is mixed. The bond rout ran a third cycle: CNBC and Bloomberg reported on September 24 that the 30-year Treasury yield closed at its highest since 2004 and the 30-year mortgage rate hit 7.45%. VIXY sits at RSI 37.6, so there's still no hedging bid.

SPY is $771.35, basing, RSI 55.9, 1.03% off its high. It's still on its 22.6-week base, $8.02 (1.04%) under the $779.37 pivot. QQQ is closer, $4.15 under its own $748.65 pivot, but Rule 6 reads SPY. Breadth is the ugly part:

• equal-weight: RSI 36.1, −4.95% over 30 days, weak-down
• small caps: RSI 34.9, −5.67% over 30 days, weak-down

The gate is shut for a fifteenth logged session.

ANALYZE: Here is the box receipt, pasted:

momentum-trader entry box — desk date 2026-09-25
  book: equity $97,507.11 · 1 open (DAR)
  screened: 1125 rows -> 2 candidate(s)
  PASS 2:
    volume-confirms:fail ×2: AVT IMAX
  NOT COMPUTED: group-participation, sector-budget

Two chart-perfect breakouts, and neither has volume behind it:

• AVT $103.08: 7.4-week base, through its $102.70 pivot, 1.43% off its high. Three-month +19.35% against SPY's +5.81%. Stage 2 is clean, but volume is only 0.79x.
• IMAX $56.05: 7.6-week base, through its $55.57 pivot, 1.11% off its high. Three-month +36.44%. Volume is 1.14x.

Both fail Rule 3 and Rule 6. A pivot cleared on sub-1.5x volume is the ILMN lesson: a trap, not a signal. The group and sector legs the box can't compute were never reached, because both names die on volume and on the gate first. The thesis-receipt step was never reached either, since there's nothing to admit.

The rate shock pushed last session's bench back into its bases:

• PLTR flipped from pullback to uptrend, but at $189.67 it's under a recalculated $194.68 pivot on 0.64x.
• FTNT is $173.46, under $181.37, on 0.53x.
• DE is $690.46, under $721.22.

Now the shadows.

• ASX is $44.31, 7.11% above the $41.37 pivot it cleared on the 18th. It's uptrend, 2.66% off its high, +286.65% on the year. The box no longer sees a base under it because it left the base. That's what a working breakout looks like.
• BIIB is $227.60, 0.52% above the $226.43 pivot it cleared on the 23rd. It printed a higher high inside the week, and the base recalculated to a $229.19 pivot. So it now sits 0.69% under the new line on 1.00x. It held, but it didn't follow through.

That's one working and one stalled. In the fork's column, ASX scores a point and BIIB has no score yet.

The August class is now fully graded or close to it:

• FCX, stopped at −9.6%
• CF, stopped at −8.6%
• DAR, at −6.77% and leaning on the line

All three were bought the same night, all three are commodity-linked, and all three cleared every leg of the Bar. I still can't separate "the Bar's non-gate legs are weak filters" from "three ordinary failed breakouts in a lane the tape stopped paying." One thing is on the record, though: three entries on one date in related lanes, sliding together, behave more like one bet than three. That's an observation, not a rule change.

DECIDE:

• Entries: PASS. Rule 6 is a hard veto: SPY is basing and $8.02 under the pivot. AVT and IMAX would also fail Rule 3 with the gate wide open.
• Trims: none on this book.
• DAR: HOLD. The mechanical −8% floor owns it at $60.48. There's no death cross, and I don't add or average down.
• Trail: nothing to trail. Nothing is near +10%.
• Forks: no new one. The group-gated fork filed 09-02 stays on file, and tonight doesn't test it. Both candidates fail on volume, not on the gate alone.

REFLECT: Tonight is the machinery working, not a verdict. The stop took CF at the number and saved another 3.18% the next day, which is exactly why it isn't negotiable. The entry side keeps delivering the same message:

• nothing at all the night the index is shut
• two bases cleared on volume nobody showed up for
• a refused breakout (ASX) that has worked for a week

That last one is still one data point. BIIB stalling right at its line is the counterweight.

The record is 0W/2L on three August fills. That isn't a sample, but it is a pattern worth writing down: same night, related lanes, same slide. For next time, watch four things:

• DAR against $60.48, which the floor will fire, not me
• a 1.5x day on AVT or IMAX while each is still within 5% of its pivot
• BIIB against both of its lines
• SPY against $779.37

Phase 5a outcome: exits checked 1 position: 0 SELL, 0 TRIM, 1 WATCH (DAR, −6.8%, near the −8% stop). Nothing filled and there were no TRIM decisions. CF was already stopped out on 09-24 at $118.44 (−8.6%) by the mechanical floor.

Trades Executed: none. PASS.

Evolved Rules: No change.

2026-10-02 — Flat Book, Shut Gate, And The Distributors Finally Brought Volume

THINK: Friday, October 2. The desk date is SPY's settled close. The book is all cash: $97,359.56, −2.64%. The record is 0W/3L and −$2,640.44 realized. The breaker is clear, 3.48% off the high-water mark. The August class is fully graded. DAR hit the −8% floor on September 29 and filled at $60.32, −8.2%. Tonight it closed at $62.94, back above my exit. I don't care. A failed breakout is failed, and the stop doesn't owe me a refund.

Exits had nothing to do: no positions, no TRIM, no WATCH. On the tape, payrolls came in at 29,000 against 84,000 expected and unemployment rose to 4.2% (CNBC, October 2). October hike odds dropped, but yields still climbed into the close. TLT is at RSI 23.5 with a death cross. SPY is $769.64, basing, RSI 54.5, still on a 23.6-week base, $9.73 under the $779.37 pivot. That is further away than the $8.02 I logged last Friday. QQQ is $0.93 over its own $748.65 pivot. Breadth is still ugly: equal-weight is weak-down at RSI 37.1 and small caps are weak-down at RSI 39.6. VIXY at RSI 41 shows no hedge bid. The gate is shut, the sixteenth logged session in a row.

ANALYZE: Box receipt, pasted:

momentum-trader entry box — desk date 2026-10-02
  book: equity $97,359.56 · 0 open
  screened: 1252 rows -> 37 candidate(s)
  PASS 37:
    fresh-not-extended:fail ×15: AMKR FN KLAC MPWR SMTOY AIXA.DE AOSL STM STRL CLFD ON VSH …
    volume-confirms:fail ×14: TSM SMH PAPR QYLD EWT QQQM QQQ VGT XLK FTEC SOXX BHE …
    market-gate:fail ×8: BOCT POCT UOCT CIBR AIS ARW AVT DAC
  NOT COMPUTED: group-participation, sector-budget

Semis are the hot group, and the box says they don't fit this book. Fifteen names are too extended or have no fresh base. TSM and SMH have real bases through their pivots, but on 0.99x volume. Even the chip funds sit near RSI 69 without a breakout bar. A rally that already ran is not merchandise.

Eight names fail only on the gate. Three of them are fake: BOCT, POCT and UOCT are buffer funds on October reset day. A new series and a reset-day volume spike are not a base breaking out, and I won't count them as gate refusals.

The real five:

• AVT is the name I said to watch. Last week it cleared on 0.79x. Tonight it has 1.58x on an 8.2-week base, $106.94 against a $104.58 pivot, 1.56% off its high, three-month +30.13% against SPY's +3.34%, with the stage-2 stack clean.
• ARW is its distribution twin: 14.2-week base, through $233.94 at $242.01 on 2.1x, 2.12% off its high. Two peers breaking out together is the group leg I usually have to work by hand, and here it showed up on its own.
• DAC (6.2-week base, 1.6x), CIBR (11.4-week base, 1.51x) and AIS (6.7x, but 12.6% off its high) fill out the list.

Every one of them is refused by Rule 6 alone.

Shadows. This is where the group-gate fork's record gets honest:

• ASX is $47.45, about 14.7% over the $41.37 pivot it cleared September 18. It's strong-up and 1% off its high. The refused breakout worked, and it is still working.
• BIIB is $219.88. It lost the $226.43 line I said to watch, and the base recalculated to a $230.49 pivot. That breakout failed.
• IMAX is $52.72, back under its old $55.57 pivot, on 0.72x. Last week's no-volume clear went nowhere, which is the ILMN lesson again.

The fork scores one point for (ASX), one against (BIIB), and IMAX is evidence for the volume bar, not the gate. One-for-one is not a case for loosening anything.

DECIDE:

• Entries: PASS. Rule 6 is a hard veto. SPY is basing and $9.73 under its pivot. AVT, ARW, DAC, CIBR and AIS are fork exhibits, not tickets. The thesis-receipt step was never reached.
• Exits/TRIM: none. The book is flat.
• Trail: nothing to trail.
• Forks: no new one. The 09-02 group-gated variant stays filed with its ledger at ASX worked, BIIB failed. AVT and ARW are logged as tonight's paired gate-only refusals.

REFLECT: A flat book on a shut-gate night is the system working, even though it doesn't feel like it. Three August fills taught one lesson three times: related lanes bought on the same night slide together. Tonight's best setup is a peer pair, AVT and ARW, which is that same concentration risk pointed the other way, so if the gate opens, I take one of them, not both, until the sector leg is computed. Watch next session:

• SPY against $779.37. Nothing else unlocks this book.
• Whether AVT holds over $104.58 and ARW over $233.94, inside 5% of their pivots.
• Whether ASX keeps running.

The ## Evolved Rules section this playbook is supposed to carry doesn't exist. That needs fixing outside a session.

Trades Executed: none. PASS.

Evolved Rules: No change. The section header is missing from this playbook; flagged to the lead.

2026-10-09 — The Index Is Back At Its High, The Gate Still Reads Basing, And Last Week's Refusals Paid The Gate's Rent

THINK: Friday, October 9. The desk date is SPY's settled close. The book is all cash at $97,359.56, −2.64%. The record is 0W/3L, with −$2,640.44 realized. The breaker is clear.

Exits had nothing to do: no positions, so no SELL, no TRIM and no WATCH.

The index finally went somewhere. SPY closed at its first record since August on October 6 (Bloomberg, October 6). It finished the week at $778.57: +1.16% on the week, RSI 59.9, 0.39% under its 52-week high. But the regime label is still basing. The 50-day is flat and the 60-bar range is tight. The intraday high moved the 24-week base's pivot up to $781.62, so SPY is $3.05 under the line, not over it. Friday came on 0.48x volume.

QQQ is basing too: $751.27, under a recalculated $762.86 pivot. Underneath, chips gave ground after OpenAI told investors its annualized revenue was about $50 billion, below the widely cited figure (FT; CNBC, October 8). Software, cloud and cyber took the money.

Breadth is still the soft spot:

• equal-weight is weak-down at RSI 50.8, 4.65% off its high
• small caps are weak-down at RSI 38.5, 8.6% off theirs
• VIXY is at RSI 39.3, so nobody is buying protection
• TLT is in a downtrend

Bloomberg (October 9) has a "rate shock" spreading beneath the AI rally, and CNBC (October 9) says junk bonds are "flashing yellow". September CPI is scheduled for October 14.

The gate is shut, the seventeenth logged session in a row.

ANALYZE: Box receipt, pasted:

momentum-trader entry box — desk date 2026-10-09
  book: equity $97,359.56 · 0 open
  screened: 1262 rows -> 24 candidate(s)
  PASS 24:
    volume-confirms:fail ×13: ABBV QYLD WCLD IGV CLOU HACK CIBR XYLD HNGE CON BOX NTNX SEA
    market-gate:fail ×5: PLTR SKYY BUG HUM TEVA
    fresh-not-extended:fail ×4: AI ISRG TEAM ATZ.TO
    relative-strength:fail ×2: BAPR OGN
  NOT COMPUTED: group-participation, sector-budget

Tonight's merchandise is real, and it's all one lane. Software, cloud and cyber cleared nine-week ranges together. That's the group leg doing itself.

The clean ones, all refused by Rule 6 alone:

• SKYY $174.89: 8.8-week base, 1.0% over its $173.15 pivot on 3.32x, 0.03% off its high. Three-month +25.14% against SPY's +3.13%.
• BUG $50.08: 6-week base, 0.93% over $49.62 on 1.6x. Three-month +27.2%.
• PLTR $209.05: 6-week base, 2.25% over $204.44 on 1.62x, 0.25% off its high. Three-month +64.88%, under the +100% chase line. RSI 76.3.

WCLD (1.36x), CLOU (1.35x), HACK (1.2x), CIBR (0.95x) and IGV (0.76x) cleared their bases too, but without the volume. One close above a base is an attempt. Nobody has retested anything yet. No fund-specific cause for Friday's software move is on file.

Two gate-only refusals sit outside the lane:

• HUM jumped 11.56% Friday on 3.9x through a 17.2-week base's $428.88 pivot. Its own regime is still basing. The cause is not on file: its news was last pulled September 25. An 11% gap with no reason is a question, not a ticket.
• TEVA cleared $40.79 on 2.73x. Its news pull failed, so I have no read, not "no news".

TEAM fails the chase rule at three-month +132.71%. That's the DELL lesson.

Now the shadows, and this is the week the gate earned its keep. Last Friday I logged AVT and ARW as the paired gate-only refusals, the best setup on the board.

• AVT is $99.36, −7.09% on the week and 4.99% back under its $104.58 pivot.
• ARW is $227.90, −5.83% and 2.58% under $233.94.
• AIS rolled into pullback, 16.89% off its high.

Three of last week's five refusals failed inside five sessions. DAC (+2.39% on the week, 1.17% off its high) and CIBR (+5.06% on the week) kept working. Further back, ASX is $46.43, 12.23% over the $41.37 pivot it cleared September 18, still uptrend. BIIB is $225.71, 2.07% under its recalculated $230.49 pivot.

The group-gated fork's running ledger is now three worked (ASX, DAC, CIBR) and four failed (BIIB, AVT, ARW, AIS). That is not a case for loosening the gate. If anything it's the gate's best week on record. The base that fails most often is the one built while the index is still deciding.

DECIDE:

• Entries: PASS. Rule 6 is a hard veto: SPY is basing and $3.05 under its $781.62 pivot. SKYY, BUG, PLTR, HUM and TEVA are exhibits, not tickets. I never reached the thesis-receipt step, so I wrote no proposal.
• Trims: none on this book.
• Exits: none. The book is flat.
• Trail: nothing to trail.
• Forks: no new one. The 09-02 group-gated variant stays filed with its ledger at 3 worked / 4 failed. Tonight's evidence argues against it.

REFLECT: A flat book again, and this time the receipts say the flat book was right. The paired breakout I'd have taken first if the gate had opened last week gave back its whole pivot inside five sessions. Same night, same group, same slide: the August lesson, run in reverse, without costing me a dollar.

The one structural point stands. When the gate opens, the software/cloud/cyber cluster is one bet, not three. I take one ticket from the lane, sized at one unit, with a thesis receipt. A fund (SKYY/BUG) can cite the sector thesis; a single name needs its own receipt.

Watch next session:

• SPY's regime flipping to uptrend/breakout. The base pivot is $781.62. Nothing else unlocks this book.
• Whether the software cluster holds its pivots with a second close or a held retest. The 5% chase ceilings are PLTR $214.66, SKYY $181.81 and BUG $52.10.
• HUM's cause, once there's a current news pull.
• CPI on October 14, with ASML (October 14) and TSMC (October 15) for the chip group.
• The shadows: AVT, ARW and BIIB under their lines, and ASX still running.

Phase 5a outcome: exits checked 0 positions for this book (flat): 0 SELL, 0 TRIM, 0 WATCH. Breaker clear. No TRIM decisions.

Trades Executed: none. PASS.

Evolved Rules: No change. The ## Evolved Rules header is still missing from this playbook; flagged to the lead again.