House Echo — 2026-09-21

Desk Plan — House Echo — 2026-09-21

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-09-21-pm.md PM: ECHO

Shared Market Read

• Event risk today: LOW — domestic macro-silent, but the day's structural catalyst is mechanical, not a data print. No Tier-1 within 7 trading days (next PCE Sep 30, NFP Oct 2, CPI Oct 14 to follow). Only second-tier US items (7:30 Chicago Fed NAI, Goolsbee/Jefferson/Williams/Barkin speeches) — none a market-binary. Firm-wide status GO; both members' hard skips CLEAR (prior VIX close 14.87 < 30; no mega-cap reporter after prior close; FOMC/CPI/NFP out of window). The live dislocators are both S&P 500 AND Nasdaq-100 quarterly rebalances effective prior to today's open (passive flow cluster), the Houthi strike on Riyadh + Aramco Yanbu keeping crude two-way, and the Trump–Xi summit (Wed) / UNGA opening the diplomatic week.
• Session character expected: mixed / two-tier with a tech-led bid wrapped around a rebalance-rotation core. Pre-market QQQ +1.01% (SPY +0.62%), semis/AI leading (INTC +5.5%, NVDA modestly higher) on a macro-quiet Monday. The briefing is explicit that direction past the open is not cleanly pre-assignable — the shape (tech over value) is more reliable than a polarity. The rebalance effective date is the dominant intraday variable, and it is a flow event, not a price-discovery event. TEMPER's 2026-06-22 calendar-hygiene finding (A) is squarely on-point: quarterly rebalance effective dates are pre-disclosed structural events that create predictable, below-ATR, low-conviction, rotation-churn sessions — the effective date is the completion of a frontrun cycle, not the initiation of obligated flow, and passive program volume dominates intraday structure with "rotation noise rather than price discovery." That is the session modifier for BOTH Family 2 expressions.
• VIX regime: 14.87 — LOW (<18). LOW is the weakest cohort for both members — ECHO's mechanism needs a completed, high-participation directional program to have something to fade, and SURGE's PF is 1.36 in VIX < 20 vs 2.18 at 20–30.
• Key levels: SPY prior close 761.62 (prior range 757.98–762.00; SMA20 764.26 — above on the +0.62% open; ATR14 6.50 — trailing, possibly still FOMC-elevated). QQQ prior close 721.36 (prior range 715.19–721.73; SMA20 713.25 — above; ATR14 8.25). Same two-tier shape as recent weeks: QQQ above its 20-day, SPY crossing above on the open. Energy: XLE on crude's two-way geopolitical roll — WTI/Brent down ~2% prerecount on UNGA de-escalation hopes but spiking on any escalation headline.
• Macro backdrop: a macro-quiet week whose tone is set by the diplomatic calendar (UNGA + Trump–Xi Wed) and the AI growth narrative, with the oil/geopolitical vector under everything and a mechanical rebalance-flow frame on today's tape specifically.
• Desk-level consequence — the rebalance effective date is hostile to both Family 2 expressions, and it degrades each member the same way witching did Friday, namely by corrupting the inputs the mechanism reads: for ECHO, rebalance passive volume can mechanically lift vol_ratio and any ADX the rotation churn manufactures, producing a false-clearance (the 09-18 witching hazard, re-run on a structurally equivalent calendar event); for SURGE, the cumulative session VWAP its pullback/resumption geometry measures is computed from a rebalance-rotation open, so "did price pull back to VWAP and resume?" is a question about a distorted reference, not institutional commitment. TEMPER finding 06-22(A) says to rate setup likelihood LOWER on these effective dates. Honest forecasts, not a hedge: echo p_trade 0.22 / short / low, surge p_trade 0.10 / long / low.

ECHO — dynamic universe (SPY/XLE/XBI baseline; Scanner watchlist supplements), close-leg exhaustion fade at 15:30 ET

Setup present: unlikely-to-possible — the rebalance is a rotation-noise session, not a completed-program session, and a clean-looking clearance is likelier to be passive flow than an exhaustible directional program. p_trade 0.22, direction short, conviction low.

The watchlist qualifies four candidate families. I grade each against the gate stack (morning signal ≥ ±0.25%, VWAP persistence, ADX ≥ 20, vol_ratio ≥ 1.2, exhaustion_score ≥ 0.80) at 10:00 and again at 15:20:

• SPY — Long bias (weak-to-moderate program), hard skip CLEAR, and the +0.62% pre-market now clears the ±0.5% eligibility bar. This is the only index-level candidate above the desk's eligibility threshold, and its qualification is a short (fade an up morning). But an S&P rebalance effective day puts passive index buying behind the tape into the close — fading that is the 08-10 loss template on the exact session the passive flow is obligated to be there. A vol_ratio clearance lifted by rebalance participation is a false-clearance, the 09-18 witching hazard re-run. Low-quality candidate; not pre-committed, and I expect to refuse the SPY short if it fires on rebalance-flow vol_ratio.
• Semis/AI complex (SOXX/SMH, NVDA, MU, AMD, INTC, MRVL) — Long bias, CLEAR, the day's genuine directional bid. Refused as a class. A sector in a live extending bid is a strengthening driver; fading the driver's own strengthening sector is my pre-committed 08-10 loss template (09-15 XLE short: −$64.98, stopped in 25 minutes with the complex still rising). exhaustion_score did not discriminate that loss (1.6913) from my 09-08 win (1.7405), so the composite gives me no cover here. INTC +5.5% premarket is the driver itself, not a fade candidate.
• XLE — Two-way (down on de-escalation hopes, spikes on escalation), CLEAR. This is today's most mechanism-interesting candidate and the only one consistent with the 08-10 rule's spirit: crude fell ~2% overnight on UNGA/US-Iran diplomacy hopes, so the energy driver is weakening — the correct side of the rule is a long fade IF XLE's morning signal is down AND the stack clears. But today is not an EIA Wednesday — the watchlist is explicit that XLE's scheduled-catalyst status does not apply; this is pure headline-driven two-way, not a completed institutional program. A supply-restoration headline is not program completion; an escalation spiking crude intraday would reverse the fade mid-program (and the Trump–Xi / UNGA diplomacy is live in the same window). Evaluated, not pre-committed. This is the single mechanism-faithful clearance path available — but on rebalance-rotation participation it is more likely to be blocked wide than to fire.
• XBI — not a candidate. No FDA decision published this morning (Ultragenyx FAYUVI / Moderna mFLUSIVA were last week; Sept PDUFAs remain pending). No scheduled catalyst; the xbi_leg sub-book (n=2, WR 0.0%) carried, not touched. Note the carried N=2 XBI vol_ratio calibration flag (09-17 vol 1.120 / 09-18 vol 1.187, exhaustion clearing both) — it does not clear N≥3 and no new XBI data point comes today.
• Single-name event names — refused as a class per TEMPER's 2026-07-24 finding (single-stock post-earnings gap-ups reverse within the first ~90 minutes, not across the full program cycle) and because on a rebalance session a single-name spike is mechanical rotation, not an exhaustible program.

The rebalance hazard, named explicitly. vol_ratio is my historically binding gate (0.953 on 09-09; 1.179 on 09-14; 1.187 XBI on 09-18, 0.013 graze). On a day with both index rebalances effective, passive reconstitution volume can mechanically lift it — and the reflection's calibration lesson from 09-18 is directly relevant: last Friday I forecast an engine-fire false-clearance (p_trade 0.35) premised on mechanical volume inflating vol_ratio, and the opposite happened — SPY 0.992 and XBI 1.187 both stayed sub-floor. The 09-18 reflection explicitly warned me to hold engine-fire estimates to the evidence that a contained LOW-VIX sub-ATR day rarely inflates vol_ratio, regardless of the calendar. I am applying that lesson today by NOT inflating p_trade on the bare presence of a rebalance. A clearance at 15:20 that rests on rebalance-mechanical participation rather than a completed directional program → I stand down; but I am not assuming the false-clearance fires just because it's a rebalance day.

Sizing vs. event risk: event risk is LOW, no event tail, and my 15:58 hard flat means no overnight exposure regardless. No event-driven size reduction. Mechanically standard: 0.75% dollar risk per instrument, 2.0% session budget ($500), proportional scaling if combined risk exceeds budget, stop 1.5× ATR (trailing ATR: SPY 6.50, QQQ 8.25 — flag if still FOMC-inflated), hard flat 15:58 ET. Candid disclosure: my refusal of the SPY-on-rebalance-flow and single-name classes is a plan note, not a defense — the plan/engine enforcement gap is open at N=3 same-shaped (08-19 XBI, 08-20 XBI, 09-15 XLE) and escalated as calibration-shaped; nothing here prevents the engine from trading a clearance. p_trade 0.22 is the honest engine-fire estimate.

Key levels (context, not triggers): SPY 761.62 prior close, range 757.98–762.00, SMA20 764.26 (open crosses above), ATR 6.50. QQQ 721.36, range 715.19–721.73, SMA20 713.25 (above), ATR 8.25. XLE on two-way crude. 10-yr eased to ~4.95%; a re-top toward 5.00% would invert the tech relief mid-session. The operative read: an up-morning short fade into a rebalance-supplied close is the strengthening side, not exhaustion.

Invalidation: stand down entirely if (1) morning signal sub-±0.25% or direction-flips ≥2 times in the first half hour — the rebalance open's modal rotation noise; (2) vol_ratio sub-1.2 at 15:20 (modal branch on a contained rebalance day — the lesson of 09-18), or any clearance resting on mechanical rebalance participation rather than a completed directional program; (3) the semis/AI driver or the energy leg is still strengthening into the close window — the 09-15 loss template / 08-10 rule, refused; (4) a single-name rebalance mover (BE/ILMN/Everpure adds, TTD/BLDR/TAP deletes) presents as a fade — index add/delete forced flow is obligated purchasing/selling, the opposite of an exhaustible program. The most likely logged reason is vol_ratio or morning_signal / vwap_persistence on a two-way rebalance-rotation tape.

SURGE — SPY/QQQ, VWAP pullback resumption (10:00–12:00 ET)

No trade expected — structurally incompatible session. p_trade 0.10, direction long, conviction low.

SURGE is live today (FOMC was 09-16, so no hard skip fires) — the low forecast is a session-character call, not a mechanical skip. Three reasons, in order:

1. Today's first hour and its VWAP are both rebalance-rotated inputs, and my mechanism reads both. Both the S&P 500 and Nasdaq-100 quarterly rebalances are effective prior to the open. My three-phase structure needs a clean trend establishment in the first hour; today's first hour is dominated by passive reconstitution rotation (subtle rotation churn, index rebalancing in the open and near the closing auction), not a clean institutional commitment. Worse, my pullback/resumption geometry is measured against the cumulative session VWAP, which is computed from that rebalance-rotation open — "did price pull back to VWAP and resume?" is a question about a distorted reference. This is the input-validity problem, not a threshold problem — identical class to the 09-18 witching argument — and TEMPER finding 06-22(A) backs exactly this call: rebalance effective dates rate setup likelihood lower for directional impulse and clean pullback structure.
2. This is my documented loss class, structurally. A firm tech-led gap/relief shape (0/2 at the 2026-07-24 catalyst-gap learning; the 08-14 finding) and the 09-17 relief-gap-and-hold session both say the first hour on a gapping, news/rotation-driven tape is a price-discovery process, not a clean trend establishment. Today is that shape with a second mechanical distortion (the rebalance) layered on top.
3. LOW VIX (14.87) is my weakest cohort — PF 1.36 in VIX < 20 vs 2.18 at 20–30. My thesis lives in elevated vol.
4. The chronic binding gate remains pullback_resumption, not the first-hour floor. The ±0.30% floor plausibly clears today (a semis-led up-open). That is precisely why the floor is not the question: even on the 09-10 session where the floor genuinely cleared, the resumption gate denied anyway — now 14th consecutive no-trade session on this stretch every session through 09-18 (ADX 21.1 graze on 09-18 never even reached the resumption gate). On a rebalance day with passive flow concentrated at the open and closed auction while midday goes quiet, the measured-pullback-then-resume window is the least likely part of the session to complete.

Sizing vs. event risk: event risk is LOW, no event tail, and my 14:00 ET hard flat means I never carry a print regardless. Standard sizing applies only if a full three-phase structure completes despite the rotation frame: 0.75% ($187.50), single position, co-fire limiter (stronger |signal_return| between SPY/QQQ wins). No adjustment — and I am explicitly not sizing for a trade I do not expect. There is no VIX skip in my cleared code and I am not adding one; the session-character read is a forecast, not a parameter.

Key levels (context, not triggers): SPY 761.62 prior close, range 757.98–762.00, SMA20 764.26 (open crosses above on +0.62%), ATR 6.50. QQQ 721.36, range 715.19–721.73, SMA20 713.25 (above), ATR 8.25; pre-market QQQ +1.01% — the resilient, leading leg, and the one a with-trend resumption would call long today (an up bid, unlike 09-18 where the qualifying leg was down). The operative level is each leg's session VWAP through the 10:00–12:00 window, with the rebalance-rotation caveat above.

Invalidation (the expected outcome, not a contingency): (1) neither leg prints a ±0.30% first-hour impulse — the modal case on a rebalance-rotation open; (2) a leg trend-qualifies but no VWAP-touch pullback resumes inside the window — the chronic block, and today the VWAP it must touch is partly a rebalance artifact; (3) the rebalance-driven open's first-hour drive reverses after 10:00 (the rotation was mechanical, not an establishment). The most likely logged reason is no_signal at first_hour_signal or pullback_resumption; both are structural absence, not threshold grazes, and neither re-opens the floor question (the 09-05 deliverable closed it — loosening the floor converts none of those sessions).

Plan Filed

• Filed: 2026-09-21 04:34 PDT
• Frontmatter forecasts complete for every active member: yes
• Firm-wide status: GO (event risk LOW, VIX 14.87 LOW regime, both index rebalances effective today — the session's structural catalyst). No hard skip fires for either member — both are live on paper, one is expected to be refused in prose but not in code.
• Desk net forecast: ECHO p_trade 0.22 / short / low, SURGE p_trade 0.10 / long / low.
• Calibration disclosure: ECHO's 0.22 is deliberately restrained below the fire-overshoot it logged on 09-18 (0.35, Brier 0.1225). That overshoot was premised on mechanical volume inflating vol_ratio into a false-clearance — and the reflection found a contained LOW-VIX day rarely does that, regardless of the calendar. The same reasoning applies to today's rebalance, so I am not re-running the same forecast error on a structurally-equivalent calendar event. SURGE's 0.10 reflects the input-validity argument (rebalance-rotated VWAP/open), a genuine "no trade expected," not a fence.
• Carry: neither stack carries overnight risk — ECHO flat 15:58 ET, SURGE flat 14:00 ET. No Tier-1 event today, so no event tail.
• Carried TEMPER items unchanged and untouched this session: ECHO plan/engine enforcement gap (N=3, escalated); ECHO XBI vol_ratio floor flag (N=2, incomplete data — flagged, not escalated); XBI exclusion review (n=2, not triggered); SURGE session-character / catalyst-gap conversation; SURGE ADX-band watch; SURGE first-hour floor question (closed). No gate logic modified — nothing in this plan touches a baseline. The rebalance-effective session modifier is a forecast, not a parameter change.

House Echo Desk Reflection — 2026-09-21 (Monday)

Cadence rule header: Weekly/Friday or same-day-on-trade (WS5). Today is same-day-on-trade — ECHO logged a real trade, so the desk reflects today. Period covered: 2026-09-21 — the single live session since the 09-18 weekly reflection (09-19/09-20 were weekend). ECHO traded 2 (SPY short −$31.43, XLE long +$24.03 → net −$7.40); SURGE no-trade.

Nominal desk net P&L this period: −$7.40 (ECHO −$7.40; SURGE $0.00).

Plan reference: desks/house-echo/plans/2026-09-21-plan.md EOD briefing: dadbrain/Analysis/briefings/2026-09-21-eod.md

What Happened

The gate stack finally opened on a genuinely big day — and the desk's single open calibration exposure, the plan/engine enforcement gap, re-fired and cost money while the mechanism-faithful trade won. This is the first same-day-on-trade reflection since 09-15, and it breaks a notable pattern: after weeks of structurally-absent setups, ECHO's gates cleared on a 2.3×‑ATR tech-led day and the engine traded two instruments. One was the plan's own best trade; the other was the plan-refused trade that has been the desk's documented loss template since 08-10.

• Session character (named correctly by the AM, held to the close). Both S&P 500 and Nasdaq-100 quarterly rebalances effective prior to the open; a macro-quiet Monday opening firm on the AI-chip complex; AMD's $1T milestone capped a clean tech-led up day — QQQ +2.78%, SPY +1.56% (+0.94% vs open), both above SMA20, true ranges 2.30× / 1.86× ATR. VIX 14.87 LOW. Last hour flat-to-marginal-negative — the bid held into the close. The desk plan forecast exactly this: low event risk (no Tier-1 through 09-28), a two-tier tech-led shape, and — the key frame — that the rebalance effective date is hostile to both Family 2 expressions by corrupting the inputs the mechanisms read.
• ECHO traded 2 — and the split is the whole story.
• XLE LONG (+$24.03, time_exit 15:55) — the plan's mechanism-faithful trade, and it worked. XLE morning signal −2.05% (short bias) → the correct long fade (the energy driver weakening), vol_ratio 1.3604, ADX 31.47, exhaustion 2.7924, all gates cleared. This is the exact path the plan pre-committed to (the 08-10 rule's correct side — fading a weakening driver, not a strengthening one).
• SPY SHORT (−$31.43, stop 15:35) — the plan-refused trade, and it lost. This is the load-bearing result of the day. The plan was explicit: expect to refuse the SPY short "if it fires on rebalance-flow vol_ratio" — an S&P rebalance-effective day puts passive index buying behind the tape into the close, and fading that is the 08-10 loss template, the exact session passive flow is obligated to be there. The engine traded it anyway: SPY morning signal +0.79% (long → fade short), vol_ratio 1.2013 — cleared the 1.2 floor by 0.0013x, the marginal false-clearance the plan predicted on rebalance participation — ADX 66.74, exhaustion 0.951. Stopped in 5 minutes at 774.5962, −$31.43. The plan identified the hazard precisely; the engine still took it.
• SURGE no-trade — 14th consecutive session since the 08-31 stop-out. Both legs trend-qualified hard today (SPY first-hour +0.77%, ADX 54.9; QQQ +1.47%, ADX 71.0) and neither produced a VWAP-touch pullback that resumed inside the 10:00–12:00 window — blocked at pullback_resumption on both, no no_signal. A +2.78% tech-led one-way day is structurally the least likely tape on which the measured-pullback-then-resume precondition completes. The session-character call (p_trade 0.10, "no trade expected") was correct; the chronic binding gate remains pullback_resumption.

Self-critique — the forecast side was the best-calibrated in many sessions; the enforcement side was the day's failure. ECHO's p_trade 0.22 / short / low was honest and held (traded=1, direction short→short = hit, Brier 0.6084), and the low-fire estimate correctly did NOT inflate on the bare presence of the rebalance — the 09-18 discipline held. But the plan's own pre-committed stand-down (SPY-on-rebalance, the 08-10 loss template) was not enforced — this is the 4th same-shaped instance of the plan/engine enforcement gap, and the second that cost money (09-15 XLE −$64.98, 09-21 SPY −$31.43). The uncomfortable corollary ends: last period said the enforcement gap survives by inactivity when the gate stays shut; today the gate did not stay shut and the gap was exposed with a price tag. SURGE's forecast was cleanly calibrated (p_trade 0.10 → no trade, Brier 0.01) and its direction: long was correct this time (both qualifying legs were up).

Desk Performance Summary

Desk net P&L this period: −$7.40. ECHO cumulative falls to −$155.93 (now 6W/6L, WR 50.0%, LLR +0.275, CONTINUE — no boundary risk; XBI sub-book still n=2 WR 0.0%, not triggered). SURGE unchanged at +$482.98 (LLR +1.131), its 14th straight no-print.

Period Outcome Table (per member)

Per-Member Binding Gate Analysis

ECHO — binding gate: none (the gate stack cleared); the binding failure was plan enforcement, and the near-miss data is calibration, not regime

Today the binding question is inverted from the last six weeks: ECHO's mechanism did not need to be blocked — it fired. On a 2.3×‑ATR tech-led day the morning signals (−2.05% XLE, +0.79% SPY) were present and the gates cleared (vol_ratio 1.3604 / 1.2013, ADX 31.47 / 66.74, exhaustion 2.7924 / 0.951). The day's bound was the plan/engine enforcement gap, re-fired — and it is the strongest calibration-shaped evidence the desk has produced since it froze dormant. The plan pre-committed to refuse the SPY short on rebalance-flow vol_ratio; the trade that fired was exactly that — SPY vol_ratio 1.2013, a 0.0013x marginal clearance the plan itself predicted would be rebalance- lifted (the 09-18 false-clearance hypothesis, now validated live as a real, if rare, mechanism: a big index-rebalance day CAN lift vol_ratio to a marginal clear, and the fade it admits is the 08-10 loss template). The engine took it and lost −$31.43.

Regime or calibration? Calibration. This is the 4th same-shaped instance (08-19 XBI short −$118.47, 08-20 XBI long −$116.84, 09-15 XLE short −$64.98, 09-21 SPY short −$31.43 — a plan-no or plan-refused stack that the engine traded, 0W/4L, two of them since the Desk Review sized it sizing-(b)). The 09-15 learnings entry already crossed the N≥3 calibration bar and flagged it as a gate-level-enforcement conversation for TEMPER; today adds a 4th live instance and the first one where the plan's specific predicted hazard (rebalance mechanical clearance) fired within it. No gate logic modified — this remains a TEMPER conversation, never a unilateral tweak. The plan/engine gap is the desk's single open calibration-class exposure and it just cost $31.43 on the day the gate opened. Separately, the mechanism-faithful XLE long winning (+$24.03) is a healthy signal: when ECHO takes its own best setup, the edge is intact (2 pinch-loss only on the plan-violation).

SURGE — binding gate: pullback_resumption (both SPY and QQQ trend-qualified, neither pulled back); regime — structural absence on a one-way day

14th straight no-trade. Today both legs trend-qualified simultaneously at strong-trend ADX (54.9 SPY / 71.0 QQQ) — the strongest dual qualification of the stretch — and neither produced a VWAP-touch pullback that resumed within the window. A +2.78% tech-led one-way day is structurally the least likely tape for the three-phase precondition (a clean pullback then resumption needs a retrace that a relentless rally never gives). This is the 08-13 finding on repeat, and the ADX 54.9/71.0 — far above the moderate-trend band (22–30) that TEMPER's own analysis found outperforms — supports the ADX re-tune watch, not an intervention. Regime, no fix, no TEMPER conversation on a threshold: pullback_resumption is a structural precondition, not a grazed floor, and no gate was touched. Forecast side clean: p_trade 0.10 honest (Brier 0.01), direction: long correct.

Cross-Member Synthesis

Narrative finding: last period closed the thesis that "a closed gate, not a good plan, is what protects this desk" — a week the gate stayed shut. Today is the mirror test that answers it: the gate opened, the good plan was the only defense, and it was not enforced. ECHO took both its best setup (which won) and its refused setup (which lost, on exactly the predicted rebalance false-clearance). The enforcement gap is no longer an abstract exposure — it has a live cost and a clean, repeatable signature (plan-refused stack → engine fire → loss). SURGE remains the desk's well-calibrated, zero-cost side; its absence is structural, not broken.

Open TEMPER Items (Carried Forward)

• ECHO — plan/engine enforcement gap: ESCALATED already, now N=4 and costful. Same-shaped instances: 08-19 XBI (−$118.47), 08-20 XBI (−$116.84), 09-15 XLE (−$64.98), 09-21 SPY (−$31.43). The 09-15 learnings bar (N≥3, calibration) was already crossed and the Desk Review sized it sizing-(b): gate-level enforcement of plan stand-downs. Today re-demonstrated it on the day the gate finally opened and added the first live validation of the 09-18 false-clearance hazard (SPY vol_ratio 1.2013 on a rebalance-effective day). No live gate modified.
• ECHO — XBI vol_ratio floor: carried, N=2, data-incomplete — flag, not escalate. No new XBI data point today (XBI not a candidate; no FDA catalyst). Unchanged.
• ECHO — exhaustion_score binding-gate watch: today's two live clears (0.951 SPY, 2.7924 XLE) continue the live-median tracking; sample still small. The 09-08/09-15 non-discrimination caveat (the composite doesn't separate the 09-15 loss from the wins) is unchanged.
• ECHO — XBI exclusion review (WR < 50% over 15+): sub-book n=2, WR 0.0% — NOT triggered. Carried.
• SURGE — session-character / catalyst-gap filter conversation: open, carried. Today was not a catalyst-gap day but a one-way strong-trend rally — the same structural absence class (no retracement), consistent with the 08-13 finding.
• SURGE — ADX 22.0 floor / ADX re-tune watch: reinforced, not an intervention. Today's ADX 54.9/71.0 sit far above the moderate band (22–30) TEMPER found outperforms strong-trend; the strong trend was all day a no-pullback. Watch stays open; N now includes a 14th no-trade with the trend leg securely qualified.
• SURGE — first-hour floor question: closed; unchanged.
• SURGE — event-day skip diagnostic (resolved MIXED); 2023 PF 0.95 watch: carried, no new data.
• Plan-scoring note (desk-level, for score_plans.py): ECHO p_trade 0.22 forecast a low fire-rate and the stack fired — worth noting the forecast was right on calibration while the plan text refused the specific SPY leg. The p_trade/Brier only captures "traded", not "traded the refused leg"; the plan/engine gap is invisible to Brier and must be tracked by the N-count, not the score.

Learnings Flags (desk-blind)

• Flag for learnings.md — ECHO / HOUSE ECHO: the plan-level stand-down gap (already N≥3) now has a 4th live instance (09-21 SPY short −$31.43) where a plan-refused SPY fade fired on a rebalance-lifted vol_ratio 1.2013 vs 1.2 (0.0013x marginal clear) and stopped out — the first live validation that a big index-rebalance effective day CAN mechanically lift vol_ratio to a marginal clear and admit the plan-flagged 08-10 loss-template fade. Spend, not avoided. This is the evidence a gate-level enforcement of plan stand-downs conversation should cite.
• Flag for learnings.md — SURGE / HOUSE ECHO: 14th straight no-trade; strong-trend simultaneous qualification (ADX 54.9/71.0) with no pullback-resumption on either leg reinforces the session-character/ADX-re-tune watch: one-way strong-trend rallies structurally deny the measured- pullback precondition.

Desk Summary for Orchestrator

Desk-cadence reflection for the single same-day-on-trade session 09-21 — desk net −$7.40. ECHO traded 2: XLE long +$24.03 (plan-consistent, gate-clean) and SPY short −$31.43 (the plan-refused rebalance-fade). Cumulative −$155.93 (6W/6L, WR 50%, SPRT CONTINUE). Load-bearing verdict: calibration — the plan/engine enforcement gap re-fired (N=4, 0W/4L, two cost money) on the day the gate finally opened; SPY vol_ratio 1.2013 (0.0013x) validated the 09-18 false-clearance hazard live, and the plan's read was correct but unenforced. No gate modified; the already-N≥3 TEMPER conversation (sizing-(b): gate-level enforcement of plan stand-downs) is reinforced. SURGE no-trade, 14th straight — pullback_resumption on both legs (SPY +0.77%/ADX 54.9, QQQ +1.47%/ADX 71.0) after trend-qualification; regime (one-way strong-trend day, no retracement), cleanly forecast (p_trade 0.10, Brier 0.01), cumulative +$482.98, SPRT CONTINUE. Both stacks flat at their hard-flat times; no event tail into 09-28; next Tier-1 PCE Sep 30. One flag for learnings.md (ECHO enforcement N=4 + vol_ratio validation; SURGE strong-trend no-pullback).