House Echo — 2026-09-28

Desk Plan — House Echo — 2026-09-28

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-09-28-pm.md PM: ECHO

Regime Watch

• Read: data-infra/regime/current.json — status: ok, as_of: 2026-09-22, i.e. stale by three sessions (it lags 09-23, 09-24 and 09-25). Snapshot: spy_day flat, spy_open_to_close_pct −0.08% / spy_vs_sma50 above 50d / spy_drawdown_bucket <2% (spy_drawdown_pct −0.58%) / vix_regime LOW (<18), vix 14.21.
• Staleness handling: I take the bucket from the file and the level from the briefing. Live ^VIX is 16.40 (prior close 14.87) — the highest print of this window, and ~1.6 points below the LOW/MID boundary at 18. I am not treating 14.21 as current.
• Charter indicators moving: none of the three shared tags is in a stressed state. spy_vs_sma50 above 50d (SPY 771.35 vs SMA20 765.34; QQQ 744.44 vs SMA20 720.88); spy_drawdown_bucket <2% (the deepest close in the window is −4.03% off the trailing high per the regime picture); vix_regime LOW. The one live drift is VIX, +1.53 off the 09-25 close and now within ~1.6 pts of the boundary — a level drift inside the bucket, not a bucket change.
• Consistency with the desk's ## Regime Charter: the charter is DRAFT — not ratified (corrected per the 2026-09-25 peer review; the prior plan mislabelled it "ratified"), and I am held to it as drafted. The draft's claim is that this desk does not need the market to go up: ECHO trades contra the day's direction, SURGE with it, both intraday and direction-agnostic. Today's read — a gap-down, headline-driven, rate-sensitive-growth-led risk-off tape — feeds both stacks without requiring an SPY rally: ECHO's contra rule points long into the down morning, SURGE's with-trend rule points short on a down establishment. The plan is consistent with the draft on substance. It is not scored as charter-compliant — the charter is not ratified.
• Persistence rule: not armed. No spy_vs_sma50 = below 50d, no drawdown ≥ 2-5%, no vix_regime ≥ MID, and neither stack is near 10 consecutive no-fire sessions (ECHO and SURGE each last fired 09-23, so both are at 3 no-fire sessions). No proposal, no posture change, watch only. Item carried: if VIX closes ≥ 18 for 5 consecutive sessions the clause arms and I file through the Desk Review shapes — I do not move a gate on regime evidence alone.

Shared Market Read

• Event risk today: HIGH (briefing). No Tier-1 release today — next is PCE Wed Sep 30 (2 days), then NFP Fri Oct 2 (4 days). The rating rests on the live geopolitical clock: over the weekend President Trump rejected Iran's seven-day Hormuz proposal, and crude rebounded as the market marked down near-term de-escalation. Firm-wide status GO; the only scheduled US items are the 10:30 AM ET Dallas Fed manufacturing survey (prior 11.6) and Treasury bill auctions — both LOW. Hard skips all CLEAR for both members: no FOMC/CPI/NFP/PCE; prior VIX close 14.87 < 30; no QQQ mega-cap reporter after the prior close (CCL/KMX report Sep 29 BMO; JEF after today's close); regular Monday session.
• Session character expected: news-driven, gap-down open, and — per the briefing — with tier polarity and index direction explicitly unassigned. Rate-sensitive growth is the visible weak leg (Nasdaq futures weaker than S&P futures), oil and long-end yields are up, the dollar strengthened, gold weakened, Asian equities fell and European shares are mixed-to-firmer with defensives and energy outperforming. The briefing's own instruction is not to promote the initial tech-underperformance read into a directional base case — a policy/geo coin-flip can flip the tape on one headline. Base case: a wider-range, higher-participation session than last week's contained sub-ATR prints — but with genuine two-sided risk.
• VIX regime: 16.40 — LOW (<18), up 1.53 from the 14.87 prior close. LOW is the weakest cohort for both members — ECHO needs a completed, high-participation program to have something to exhaust (though a genuine shock day is the environment its edge is built for: "bigger days exhaust harder"), and SURGE's backtest PF is 1.36 in VIX < 20 vs 2.18 at 20–30. Neither stack has a VIX parameter (no VIX skip exists in SURGE's cleared code; ECHO's skip is a first-half-hour |return| proxy, 1.8%), so this is a forecast input, not a gate. Note the 16.40 print is the window's high and sits only ~1.6 pts under the boundary.
• Key levels: SPY prior close 771.35 (prior range 766.32–772.26; SMA20 765.34 — above; ATR14 6.83). QQQ prior close 744.44 (prior range 739.67–745.84; SMA20 720.88 — above; ATR14 9.44). Trailing ATR is normal, so an open beyond roughly ±0.5% (SPY ≈ 3.8 pts, QQQ ≈ 3.7 pts) is a genuine gap. Watchlist pre-market snapshots (NOT authoritative — Investing.com, asynchronous): SPY 768.15 (−0.41%, ≈ 0.47× ATR); QQQ 737.83 (−0.90%, ≈ 0.70× ATR). If those hold at the open, this is a real gap down, QQQ the weaker leg — confirm against the printed open. The levels that matter are Friday's session extremes (SPY 766.32 / 772.26; QQQ 739.67 / 745.84) and, for the fade, whether the gap extends or is bought back.
• Macro backdrop: a weekend geopolitical breakdown re-opens the oil-supply / inflation-shock channel while long-end yields are already elevated — crude and yields up, risk-off tone, and both channels pressure high-duration growth. PCE (Wed) is the near-term binary.
• Desk-level consequence. This is the opposite tape to last week: 09-24/09-25 were contained, sub-1.0 vol_ratio sessions with no program for either stack to work on; today is a macro-shock, higher-participation, gap-down session — the environment ECHO's mechanism is explicitly built for ("bigger days exhaust harder"), and the environment SURGE's with-trend rule can express short. The honest counterweights: (a) the briefing's own warning not to pre-assign polarity — a headline reversal can convert either call; (b) the gap-down-open-bought-back shape is exactly what denied SURGE its establishment on 09-24; and (c) for ECHO, energy is the leg with the strongest same-session catalyst and it is on the fade's refused side (see ECHO section — the driver-polarity label is a context tag, not an endorsement).

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

Setup present: genuinely possible — a higher-participation shock day is the environment this mechanism is built for — but with an unassigned polarity and one leg on the wrong side of the fade. p_trade 0.35, direction long, conviction low.

p_trade 0.35 is anchored and lifted for a forward-looking, mechanism-specific reason — not for recency. Realized base rates from intelligence/plan-calibration.csv: 10 fires / 50 scored sessions = 0.20 all-time; 3 / 10 over the last ten = 0.30; 3 / 7 over the last seven eligible (non-event) sessions = 0.43. The lift above the recent ten-session rate is the one thing last week's tape could not give this mechanism and today's can: participation. 09-24/09-25 blocked on wide vol_ratio misses (09-25 all three instruments 0.998 / 0.980 / 0.787 against the 1.2 floor) on a contained, sub-ATR tape — a session with no program in it. A HIGH-event-risk, gap-down, oil-and-yields shock day is the opposite: wider ranges, elevated volume, and a same-day morning signal that clears ±0.25% readily. Not raised to 0.45 because (i) the briefing explicitly leaves polarity unassigned, (ii) the gap-down-open-bought-back shape (09-24) would leave every instrument sub-±0.25% and block at morning_signal, and (iii) even a qualifying morning signal still has to survive ADX ≥ 20, vwap_persistence and exhaustion_score ≥ 0.80 at 15:20.

direction long is the honest first-trade sign on the stated base case: my entry is contra the morning signal, and the briefing's base case is a gap-down, rate-sensitive-growth-led open — a down morning means my contra entry is long (SPY, and XBI if the biotech complex is sold with growth). If the tape instead opens and stays up (a headline-relief reversal), the first trade is a short and this sign is simply wrong — the mechanism resolves direction from the observed morning signal, not from this field. I am not pre-assigning tier polarity; the sign is a plan-layer forecast.

Grading the Scanner watchlist's Directional Catalyst Instruments against my gate stack (morning signal ≥ ±0.25%, VWAP persistence, ADX(14, 5-min) ≥ 20, vol_ratio ≥ 1.2, exhaustion_score ≥ 0.80):

• SPY — mild DOWN bias, CLEAR on skips — plan-evaluated (the default broad leg, not a distinguished catalyst). The watchlist rates SPY's catalyst weakest ("mild downside bias"). A down-morning SPY means my contra entry is LONG into a declining broad tape — the fade of a declining driver. Electronically the leg is plausible on a shock day (a −0.4%/0.5× ATR gap should clear the morning-signal floor and lift vol_ratio), but SPY's vol_ratio was the modal blocker all last month (1.166 / 1.179 / 0.792 / 0.998) and its signal has repeatedly sat sub-floor on contained days (09-22 +0.04%). Plan-evaluated, not endorsed.
• XLE — UP bias (oil rebounded on the rejected Iran proposal), CLEAR on skips — plan-evaluated; the driver-polarity label is a CONTEXT TAG, not an endorsement. Energy is the deliberate relative-focus sector on today's live oil-supply catalyst, and the watchlist frames it as a trending-day candidate. An up-morning XLE means my contra entry is SHORT XLE — fading a strengthening energy driver, which is the class my 08-10 loss template labels "refused." Per the 2026-09-25 peer review (House Mneme, disposition acted) that label is demoted to a context tag and is no longer a sufficient basis to endorse or veto a path on its own — the partition has one live instance per cell and no denominator, so I state it as a risk and open the denominator (allowed-side and refused-side XLE/XBI fades, wins and losses alike, alongside the pre-entry taxonomy live since 09-24). XLE also has the closest vol_ratio history to the floor (last two fired trades 1.3637 / 1.3604), so mechanically it is the leg most likely to clear on a headline energy session — which is exactly why an up-morning XLE short is the trade I most need to think about rather than wave through. Plan-evaluated, not endorsed.
• QQQ — DOWN bias (rate-sensitive growth is the visible weak leg), CLEAR on the gates — plan-evaluated, not endorsed. QQQ is not in my baseline universe and was removed from the dev universe in v4 on mega-cap-concentration grounds; today's catalyst is sector/rates-level, not a constituent print, so the v4 caveat does not arm. A down-morning QQQ is a long contra fade of the weak leg. Plan-evaluated, not endorsed.
• XBI — CONDITIONAL, and the condition is unmet at scan time — plan-evaluated on the mechanism, DECLINED unless the catalyst is confirmed. The watchlist flags a catalyst watch: Biofrontera's (BFRI) Ameluz supplemental-NDA PDUFA target date is today, and Egetis's Emcitate PDUFA is also Sep 28 — both decisions were still pending at scan time. My pre-committed rule for this situation is explicit: confirm the decision released before acting; if still pending, drop it — the watchlist itself says "treat as catalyst watch, not confirmed approval," and Egetis is foreign-listed with no US-listed equity identified. I am not repeating the 2026-08-20 error in the other direction: the 08-20 finding was that a calendar-anchored XBI exclusion was wrong while the engine traded the mechanism anyway. So the exclusion here is catalyst-status-anchored, not calendar-anchored, and if a decision does publish intraday the leg becomes live on the mechanism exactly as evaluated. xbi_leg sub-book n=3, WR 0.0%, PF 0.0; exclusion review triggers at 15+ live trades — not triggered.
• Single-stock event gaps (INTC −3.82% and other pre-market movers in the watchlist snapshot) — plan-declined as an INSTRUMENT CLASS. TEMPER's 2026-07-24 finding (the INTC 7/23-7/24 sequence): a single-stock event gap completes its move in the first ~90 minutes, not across a six-hour program cycle, so there is no exhausted all-day program at 15:20 to fade, and a large gap makes VWAP persistence degenerate. The engine may still take one; I cannot enforce a veto.

Sizing vs. event risk: event risk HIGH, but the 15:58 ET hard flat means nothing survives a headline, the PCE print Wednesday, or the session — and the near-term binary is 2 days out, not today. 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(14, 5-min), hard flat 15:58 ET. With no distinguished endorsed leg today this is most plausibly a one-position session, occasionally two.

Pre-entry flags (peer-review commitment, applied as written): from 09-24 every engine fill is logged as plan-endorsed / plan-evaluated / plan-declined, wins and losses alike. There is still no denominator and therefore still no enforcement proposal — the plan/engine enforcement gap remains an unvalidated hypothesis (reclassified 09-24) and the plan-veto stays out of the gates. No gate logic is touched by anything in this plan.

Key levels (context, not triggers): SPY 771.35 prior close / range 766.32–772.26 / SMA20 765.34 / ATR14 6.83 — a down morning ≥ 0.25% is the fade I evaluate on SPY; a down morning on XLE is my long fade (allowed class), an up morning on XLE is a short of a strengthening energy driver (context tag). QQQ 744.44 / 739.67–745.84 / SMA20 720.88 / ATR14 9.44. XLE — no authoritative briefing level; the driver line is crude, and today crude is UP (rebounded after the Iran proposal was rejected), so the energy leg's bias is up, not down as it was 09-25. XBI — no authoritative level; the operative event is the PDUFA status.

Invalidation: stand down entirely if (1) every candidate is sub-±0.25% in the first half hour or direction flips ≥2 times — the modal chop case; (2) vol_ratio < 1.2, ADX < 20, or exhaustion_score < 0.80 at 15:20 — the standard wide block; (3) the open confirms a one-way up program (a headline-relief reversal) and the only clearing legs are up-morning longs-turned-shorts — the refused class; (4) crude reverses back below the pre-weekend level and energy is the weak leg — XLE flips fade direction and I re-evaluate rather than assume; (5) any single-stock event gap presents as a fade — declined as a class; (6) XBI first-half-hour ≤ −5% (my hard skip). Most likely logged reason: vol_ratio or morning_signal. No overnight exposure — 15:58 ET flat — so the Iran clock cannot be traded today, only gapped through.

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

Setup present: possible but marginal — a directional, gap-down open is where an establishment comes from, but the gap-down-open-bought-back shape is exactly what denied it three sessions ago. p_trade 0.20, direction short, conviction low.

SURGE is live (no FOMC/CPI/NFP/PCE; no QQQ mega-cap earnings day; no VIX skip exists in the cleared code, so VIX 16.40 does not gate it).

p_trade 0.20 is anchored, not asserted. Realized base rates from intelligence/plan-calibration.csv: 6 fires / 50 scored sessions = 0.12 all-time; 1 / 10 over the last ten = 0.10; 2 / 15 over the last fifteen eligible sessions = 0.13. The lift above that is the same forward-looking input as ECHO's: today carries a genuine directional gap (SPY ≈ −0.4%, QQQ ≈ −0.9% on the pre-market snapshots — a real gap, not the sub-ATR noise of last week) against a live macro shock, which is the environment in which a first-hour establishment ≥ ±0.30% is more likely to form. It is held at, not above, 09-24's 0.20 — because 09-24 was the same shape (a gap-down, tech-led open on a rates-driven risk-off tape) and produced no establishment at all: SPY blocked first_hour_signal at −0.22% and QQQ sat on the wrong side of VWAP. A larger gap does not by itself manufacture a measured pullback, and my chronic binder is pullback_resumption. No recency lift for the 09-23 win (N=1).

Direction: short — a plan-layer sign only. My mechanism resolves direction from the observed first-hour impulse at evaluation time, not from a pre-committed sign (on 09-23 it scored a direction miss on a winning short; TEMPER has the field flagged as a scoring artifact for this stack). I sign short because the observable today is a gap-down open with rate-sensitive growth the weaker leg — the shape that, if it establishes, establishes down, and a with-trend resumption on a down day is a short SPY or QQQ. This is the leg my charter flags as not yet shown to pay (live long +$468 vs short +$15, n=8 total), so I state it as the honest sign, not a confident one. If the tape opens up or the gap is bought back without a sustained establishment, the mechanism takes no side and this sign is simply wrong, as designed.

Why not higher — the shape risk is concrete and chronic, and it is precisely today's most likely shape. A gap-down open that is bought back (the 09-24 tape: gap down, low in the first quarter-hour, then a grind back toward unchanged) produces no first-hour impulse and no retracement structure — exactly the 09-24 double-block (first_hour_signal −0.22% on SPY; vwap_side on QQQ). And even when a leg trend-qualifies, my chronic binder is pullback_resumption: 09-21 (SPY ADX 54.9 / QQQ 71.0, neither produced a VWAP-touch resumption), 09-10, 09-22 — a trend-qualified first move that never retraces is structural absence, not an adjustable threshold. A headline-driven session is more prone to a one-way impulse and to mid-session reversals than a slow trend day — both of which are denial shapes for me.

Sizing vs. event risk: event risk HIGH, but I never carry it — the 14:00 ET hard flat leaves no exposure into the afternoon of a live geopolitical clock or the PCE build-up. Standard sizing applies only if a full three-phase structure completes: 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 merely tilt toward. No VIX skip is added; the session-character read is a forecast, not a parameter.

Key levels (context, not triggers): SPY 771.35 prior close / range 766.32–772.26 / SMA20 765.34 / ATR14 6.83; QQQ 744.44 / 739.67–745.84 / SMA20 720.88 / ATR14 9.44 — QQQ is the weaker leg today (the inverse of 09-25), so a completed down-structure calls short QQQ or SPY. The operative level is each leg's session VWAP through the window, and today it is a clean reference (no rebalance, no witching): a bar-over-bar pullback that touches VWAP on contracting volume and then a resumption close back beyond it in the trend direction, with ADX still ≥ 22 at the resumption bar. Note the 22.0 ADX floor (09-18 graze: 21.1) and the moderate 22–30 band where the 09-23 winner (26.05) landed — the ADX re-tune watch, not a change.

Invalidation (the expected outcome, not a contingency): (1) the gap-down is bought back in the first hour and the tape is directionless chop — the contained-day modal case, and 09-24's exact shape; (2) a leg trend-qualifies but no VWAP-touch pullback resumes inside 10:00–12:00 — the chronic pullback_resumption block, and the single most likely logged reason (09-10, 09-21, 09-22); (3) the gap extends one-way down without retracing — the same failure mode on this tape's directional case; (4) neither leg reaches a ±0.30% first-hour impulse. Most likely reason: first_hour_signal or pullback_resumption — structural absence, not a threshold graze, and neither re-opens the first-hour-floor question (closed by the 09-05 deliverable). No overnight exposure — 14:00 ET flat.

Plan Filed

• Filed: 2026-09-28 04:50 PDT
• Frontmatter forecasts complete for every active member: yes
• Firm-wide status: GO (event risk HIGH, VIX 16.40 LOW, regular Monday session). No hard skip fires for either member — both live on paper.
• Desk net forecast: ECHO p_trade 0.35 / long / low, SURGE p_trade 0.20 / short / low. The two are not symmetric and not averaged: ECHO's 0.35 is a forward-looking lift for participation above its recent ten-session rate (all-time 0.20, 10/50), capped by the unassigned polarity and the 09-24 bought-back-gap shape; SURGE's 0.20 is held at 09-24's level (all-time 0.12, 6/50) on the same participation logic, offset by the identical 09-24 no-establishment outcome and the chronic pullback_resumption block. Both directions are plan-layer signs — ECHO's long is contra the expected down morning, SURGE's short is with-trend on the expected down establishment.
• Carry: neither stack carries overnight risk — ECHO flat 15:58 ET, SURGE flat 14:00 ET — so the Iran/Hormuz headline clock cannot be traded today, only gapped through; the near-term binary, PCE Sep 30, is 2 days out. Next Tier-1 after that: NFP Oct 2.
• Peer-review corrections carried into this plan (2026-09-25 session, both dispositions acted): (1) the 08-10 driver-polarity label is demoted from endorsement-grade to a context tag — no path is endorsed on it alone, and the allowed/refused denominator is opened this session; (2) the "already-tested floor=1.00 cell" claim is retracted — the sub-1.0 region is essentially unmeasured below floor 0.9 (its only unlocked trade, XBI 2026-08-04 at vol_ratio 0.9785, reads n=3, WR 33.3%, PF 0.055; the 0.787 depth was never tested at any rung), and the 09-05 deliverable's own verdict word is restored: inconclusive; (3) the desk ## Regime Charter is DRAFT — not ratified, and both the desk and member files say so.
• Carried TEMPER items (unchanged, untouched — no gate logic modified this session): ECHO plan/engine enforcement gap (unvalidated hypothesis, pre-entry flags logged each session, no denominator yet, no enforcement proposal); ECHO xbi_leg (n=3, WR 0.0%, trigger 15+); ECHO vol_ratio sub-1.0 region (inconclusive, not "tested"); ECHO exhaustion_score live-median watch; SURGE ADX 22.0 floor / re-tune watch; SURGE event-day skip diagnostic (resolved MIXED 2026-08-01); SURGE 2023 PF 0.95 watch.
• Regime: consistent with the draft charter; Regime watch line above; no persistence trigger armed. current.json is stale by three sessions — I state the bucket from the file and the VIX level (16.40) from the briefing.

House Echo Desk Reflection — 2026-09-28 (Monday)

Cadence rule header: Weekly/Friday or same-day-on-trade (WS5). Today is not the Friday floor — it is a same-day-on-trade trigger: SURGE logged a real fill, the desk's first print in three sessions. Period covered: 2026-09-28 — the single session since the 09-25 reflection (09-26/09-27 weekend). Desk trade_count 1, desk net P&L −$172.50.

Plan reference: desks/house-echo/plans/2026-09-28-plan.md EOD briefing: dadbrain/Analysis/briefings/2026-09-28-eod.md

What Happened

The desk's first fill in three sessions, and it lost. SURGE took a QQQ short — 25 shares, entry 734.37 at 11:40 ET, stopped at 735.75 at 12:10 ET, net −$172.50 (strategy=v9, signal_return −0.92%, ADX 34.53 at entry, retrace 0.414). ECHO logged no_qualifying_instruments for the third consecutive session. Desk cumulative is now +$50.48 (ECHO −$290.45, SURGE +$340.93).

The tape was the session both plans sketched, and it is worth naming before the critique: news-driven risk-off, the US–Iran standoff renewing oil-supply and inflation concerns with rising Treasury yields pressing duration-sensitive shares; QQQ closed 736.55 (−1.06%) on a true range of 12.79 vs ATR 9.91 (1.29×), SPY 765.49 (−0.76%) at 1.09× ATR; VIX 16.07, +1.20 on the prior close, LOW but the highest print of the window. Both indices finished above their 20-day averages (SPY 765.49 vs SMA20 765.15). Selling held into the last hour in both proxies — no late reversal. Energy was the supported sector on the oil bid; NVDA was an idiosyncratic countertrend on its buyback increase.

• ECHO — 0 trades, and the plan's central thesis was falsified. All three baseline instruments cleared the morning signal — SPY −0.39%, XLE +0.84%, XBI −0.44%, all beyond ±0.25% — and all three blocked on vol_ratio: SPY 1.102, XLE 1.011, XBI 0.832 against the 1.2 floor. The plan had lifted p_trade to 0.35 — above its own 3/10 recent-ten rate — for one explicitly forward-looking, mechanism-specific reason: that a HIGH-event-risk, gap-down, oil-and- yields shock day is "the opposite of a session with no program in it," and would therefore lift participation over the floor. It did not. On a day whose QQQ range ran 1.29× ATR, vol_ratio came in at 1.102 or below on every instrument. That is the honest headline of my half of the desk today: I was right about the shape of the session and wrong about the participation it would carry — and participation is the one input my mechanism cannot work without. The plan's modal call was correct (vol_ratio or morning_signal was named as the most likely logged reason), but the direction of the lift was wrong.
• SURGE — 1 trade, and the plan's modal call was wrong in the opposite direction. The 09-28 plan wrote p_trade 0.20 and named its most likely outcome explicitly: "first_hour_signal or pullback_resumption — structural absence, not a threshold graze," with the gap-down-bought-back shape (09-24) called out as the concrete chronic risk. Instead the three-phase structure completed — a first-hour establishment, a VWAP-touch pullback, a resumption — and the trade was stopped 30 minutes later. This is the first time since 09-23 that SURGE's precondition chain got all the way through, and it is the first thing this reflection has to say plainly: structure-completion is not edge. The mechanism found the setup it was built to find and the setup did not pay. Brier on the row is 0.64 — the desk's worst scored forecast since 09-23.
• The trade was a short — the leg my own charter flags as not yet shown to pay. Both the plan and the desk charter say so in as many words (live long +$468.28 across 2 trades vs short −$127.35 across 8). Today added the 8th short and it lost. That is not evidence of a gate problem; it is an incremental, honest data point on an open question the charter already records, and it is now 8 shorts, 4W/4L, net −$127.35 against 2 longs, 1W/1L, net +$468.28. The trade count remains far too small to call short-capability dead — but it is no longer accurate to describe the short book as merely "unproven at +$15." It is now negative.

Self-critique, both halves. ECHO's plan was honest about its own uncertainty (conviction low, polarity explicitly unassigned, the 09-24 bought-back-gap shape named as a live denial path) and it anchored its probability to realized base rates rather than lifting on the 09-23 clear. What it got wrong is the one thing it was most confident about: that a macro-shock day manufactures participation. SURGE's plan was, in retrospect, too confident in the opposite direction — it spent most of its setup section explaining why the structure would not form, and the structure formed. Neither stack touched a gate, and neither should have.

Desk Performance Summary

Desk net P&L today: −$172.50. Desk cumulative: +$50.48.

• ECHO sits +3.026 from the CONSISTENT-WITH-BACKTEST boundary, −2.863 from DEGRADED — neither in play, and unchanged, because there were no new observations. xbi_leg unchanged at n=3, WR 0.0%, PF 0.0 (exclusion review triggers at 15+ live trades — not triggered).
• SURGE moved away from the boundary for the first time in two weeks: LLR +1.414, now +1.531 from CONSISTENT (it was +1.384 on 09-25, the closest it had been). One losing observation is one observation; the book remains comfortably clear of DEGRADED (−4.358).

Period Outcome Table (per member)

Per-Member Binding Gate Analysis

ECHO — binding gate: vol_ratio (3 of 3 instrument-sessions); reading REGIME, but the plan's participation hypothesis is now 0-for-2 and that is worth stating

vol_ratio was the binding gate on every instrument-session today, on a session where the signal layer worked perfectly (all three cleared ±0.25%). The shape of the misses, against the floor at 1.2:

• SPY 1.102 — a 0.098 miss, and inside the band the Desk Review flagged. This lands in the 1.08–1.20 range the 2026-09 Desk Review counted 8 times in 8 weeks (concentrated on XLE) — a 9th instance. It is also double-blocked on the strategy's own composite: exhaustion_score = vol_ratio × |signal_return| ≈ 1.102 × 0.39 = 0.43 against the 0.80 floor, more than half under. Same structure the 2026-09-05 vol_ratio deliverable found on the flagged XLE grazes — loosening vol_ratio alone converts nothing because exhaustion_score binds independently.
• XLE 1.011 — a 0.189 miss, and it lands in the already-tested floor=1.00 cell. The 09-05 diagnostic ran floors 1.2/1.1/1.05/1.0/0.9 across every logged day since 2026-06-22: floor 1.00 produced 2 new trades, 50% WR, PF 0.128 — worse than the strategy's own benchmarks. (XLE's exhaustion_score today would have been ≈1.011 × 0.84 = 0.85, which would clear — but only at a floor whose counterfactual is already measured and negative.)
• XBI 0.832 — a 0.368 miss. No reading of that as a graze.

The genuinely new item is not a threshold — it is a falsified plan hypothesis. The 09-28 plan lifted p_trade to 0.35 on the stated reasoning that a HIGH-event-risk shock day is "the opposite of a session with no program in it" and would lift participation over the floor. The tape delivered a 1.29× ATR QQQ range on a live macro shock and vol_ratio still came in at 1.102 / 1.011 / 0.832. This is the second instance of the same shape: 09-24 was also a gap-down, rates-driven macro-shock open, and SPY blocked vol_ratio at 1.166 with XLE and XBI blocked downstream on vwap_persistence. So the desk now has N=2 for "a gap-down, headline-driven open does not lift ECHO's vol_ratio over the floor," against a plan layer that has twice cited shock-day participation as the reason to lift its forecast. I state the N because it is 2, not 3 — this does not clear the Desk Review's N ≥ 3 bar and I am not escalating it. What it does mean is that the plan's own participation reasoning has been wrong twice in three sessions and should stop being used as a probability lift until it is supported.

Regime or calibration? REGIME. No threshold was repeatedly grazed in the desk-calibration sense: one near-band miss that is double-blocked downstream (SPY 1.102 → exhaustion ≈0.43), one miss in an already-tested and already-negative floor cell (XLE 1.011 → floor=1.00 is PF 0.128), and one wide miss (XBI 0.832). The 1.2 floor and T=0.80 are not mine to move, and nothing here is evidence that moving them would help. No gate touched, no floor conversation.

SURGE — no gate bound today: the precondition chain completed and the trade was stopped. This is not a gate question

For the first time in this desk's recent record, the thing that stood SURGE down is not the thing that happened. The plan named first_hour_signal or pullback_resumption — structural absence — as its single most likely outcome, with the 09-24 bought-back-gap shape as the concrete chronic risk. Instead:

• The first-hour impulse cleared. Both legs were in play on a directional, gap-down open, and QQQ produced a qualifying establishment (entry at 11:40, inside the 10:00–12:00 window).
• The pullback/resumption structure completed — the chronic binder that blocked 09-10, 09-21 and 09-22 did not block today. retrace_pct 0.414, ADX 34.53 at the resumption bar, comfortably over the 22.0 floor.
• And it lost anyway, stopped at 12:10 for −$172.50.

So the binding constraint today was not a gate at all — it was the edge itself. The mechanism found exactly the three-phase structure it was cleared to trade, on a session whose character (directional, news-driven, QQQ the weaker leg) is the one its backtest says it does best in, and the resumption failed. One trade at n=10 proves nothing on its own, and I am not going to dress a single stop-out up as a finding about the mechanism. But I will say the honest thing the plan did not anticipate: this desk's story for six weeks has been "the setups were structurally absent, and that is fine." Today the setup was present, and the desk still lost money. That is a different kind of session and it belongs in the record as such.

Two side observations, neither an action:

• The entry was in the strong-trend ADX band (34.53). TEMPER's own supplemental analysis found moderate-trend (ADX 22–30) outperforms strong-trend (ADX ≥ 30) — PF 2.14 vs 1.65 — which is why the ADX re-tune watch exists. Today's loss is a point for that hypothesis, but the live ledger still leans opposite it: strong-band live trades are now 3W/3L, net +$459.85 (07-08, 07-13, 07-21, 07-23, 08-24, 09-28) against moderate-band 2W/2L, net −$118.92 (07-09, 07-24, 08-31, 09-23). I record the data point; I do not move the watch.
• The short side is now net negative. 8 live shorts, 4W/4L, −$127.35; 2 longs, 1W/1L, +$468.28. The plan flagged this trade as "the leg my charter flags as not yet shown to pay," which was exactly right. The charter already says the trade counts are too small to call the desk short-capable or short-incapable; today's job is only to keep the ledger honest, and the ledger now reads negative on the short book.

Regime or calibration? Neither — no gate was involved. The first-hour floor question stays closed (the 2026-09-05 diagnostic: zero new trades unlock at floor 0.0025); the pullback_resumption binder was not the binder today; the ADX 22.0 floor was not grazed (34.53). No gate touched.

Cross-Member Synthesis

Narrative finding. These two stacks failed in opposite directions today, and the contrast is the period's most useful output. ECHO's plan was right and its premise was wrong: the session was the shock day it wanted, and participation still did not arrive. SURGE's premise was right and its plan was wrong: the session was the denial shape it feared, and the setup arrived anyway — then lost. Both halves point at the same uncomfortable place, which is that this desk's forecasts have been getting better at predicting the branch while getting worse at predicting the outcome. ECHO has now correctly named its own blocking gate three sessions running; SURGE has now had its first real trade in three sessions and it was a loser. Neither is a gate problem, and neither is something a threshold move would fix — but the desk should stop describing its no-print streaks as "the setups were absent and that is fine" without also carrying that the first time a setup did arrive, it cost $172.50.

Regime Watch

Regime watch: data-infra/regime/current.json — status: ok, as_of: 2026-09-22, i.e. stale by four sessions (it lags 09-23, 09-24, 09-25 and today). Snapshot: spy_day flat, spy_open_to_close_pct −0.08%, spy_vs_sma50 above 50d, spy_drawdown_bucket <2% (spy_drawdown_pct −0.58%), vix_regime LOW (<18), vix 14.21. Live levels from the briefing, not the stale file: VIX 16.07 (close, +1.20 vs 14.87) — still LOW, still the window's high, and now ~1.9 points under the LOW/MID boundary at 18; SPY 765.49 above SMA20 765.15; QQQ 736.55 above SMA20 721.89. None of the three shared tags is in a stressed state. Against the desk's ## Regime Charter — still DRAFT, not ratified, and I am held to it as drafted — today is charter-consistent on substance: the desk does not need the market to go up (ECHO fades the day's direction, SURGE rides it, both intraday and direction-agnostic), and on a down day the desk's two expressions pointed opposite ways exactly as the charter says they should — ECHO's contra rule pointed long, SURGE's with-trend rule pointed short and traded. Persistence rule: not armed. No 5-session stretch in a stressed tag state; ECHO is 3 no-fire sessions into a 10-session trigger (fired 09-23), and SURGE's fire today resets its own count. If VIX closes ≥ 18 for 5 consecutive sessions the clause arms and I file through the Desk Review shapes — I do not move a gate on regime evidence alone. No proposal, no posture change, watch only.

Open TEMPER Items (Carried Forward)

• ECHO — NEW, desk-scoped (not a gate item): the plan's "shock day lifts participation" lift is 0-for-2. 09-24 (gap-down macro-shock open) → SPY vol_ratio 1.166, sub-floor; 09-28 (HIGH-event- risk, oil-and-yields shock, QQQ 1.29× ATR) → vol_ratio 1.102 / 1.011 / 0.832, all sub-floor. The plan layer has twice used this reasoning to lift p_trade above realized base rates, and both times the tape refused. N=2 — below the Desk Review's N ≥ 3 bar; not escalated. Filed as a plan-layer observation so the next reflection does not re-derive it.
• ECHO — vol_ratio floor (1.2): unchanged, and today's instances are all covered. SPY 1.102 is a 9th instance in the Desk Review's flagged 1.08–1.20 band but is double-blocked (exhaustion_score ≈0.43 vs 0.80); XLE 1.011 is in the already-tested floor=1.00 cell (2 new trades, 50% WR, PF 0.128); XBI 0.832 is wide. No floor conversation.
• ECHO — xbi_leg sub-book: n=3, WR 0.0%, PF 0.0. Exclusion review triggers at 15+ live trades — not triggered. The open 09-18 XBI-specific vol_ratio question stays N=2 (09-17 1.120, 09-18 1.187 with exhaustion clearing); today's XBI blocked at 0.832, a wide different-shaped miss, so the question is neither advanced nor closed.
• ECHO — plan/engine enforcement gap: still an UNVALIDATED HYPOTHESIS. No fill today, so the plan-endorsed / plan-evaluated / plan-declined reference class gained zero rows again. Plan-veto stays out of the gates; no enforcement proposal.
• ECHO — exhaustion_score live-median watch: unchanged, still far from the 40-trade validation point. vol_ratio ≥1.5-tier discrimination watch (N=1): unchanged.
• SURGE — ADX 22.0 floor / ADX re-tune watch. Today's entry cleared at 34.53, so no graze. One new data point in the strong band, on the wrong side of the re-tune hypothesis; the live split still leans opposite it (strong 3W/3L net +$459.85 vs moderate 2W/2L net −$118.92). Watch, no action.
• SURGE — first-hour-signal floor (±0.30%): CLOSED (2026-09-05 deliverable). Not in play today — the floor cleared. No re-opening.
• SURGE — session-character / catalyst-gap filter conversation. Carried unchanged. Today is a counter-instance worth attaching: a gap-down, catalyst-driven session did produce a completed three-phase structure (the class the 2026-07-24 learnings entry called fundamentally invalid on catalyst-gap days) — and the trade still lost. That cuts against the "catalyst-gap days are structurally untradeable" reading and for "they are tradeable but not necessarily profitable." N=1; no action.
• SURGE — short-side question (charter item, restated): live shorts now 8 trades, 4W/4L, net −$127.35; longs 2 trades, 1W/1L, +$468.28. The charter already records short-capability as unproven; the ledger now reads negative. N far below any review trigger — no action, ledger updated.
• SURGE — event-day skip diagnostic (resolved MIXED 2026-08-01, TEMPER-ratified) and 2023 PF 0.95 watch: carried, no new data.
• Operational: intelligence/runner-health-2026-09-28.md reads CLEAR — both echo and surge completed and wrote their daily artifacts. This is the first clean runner-health check after two consecutive REVIEW-level sessions (09-24, 09-25); the 09-25 SURGE watchdog backfill is closed out.
• Plan-scoring note (desk-level, for score_plans.py): unchanged — SURGE's direction frontmatter is not an input to its signal (direction resolves at evaluation time), so the direction-hit field remains a scoring artifact for that stack. Proposed for TEMPER's consideration, never applied unilaterally. (Today it happened to score a hit on a losing short, which is the artifact in action.)

Learnings Flags (desk-scoped)

Flag for learnings.md: house-echo / ECHO — vol_ratio was the binding gate on 3 of 3 instrument-sessions on 2026-09-28, on a session where every instrument cleared the morning signal (SPY −0.39%, XLE +0.84%, XBI −0.44%) and blocked at vol_ratio 1.102 / 1.011 / 0.832 vs the 1.2 floor. The SPY miss (1.102) falls inside the 1.08–1.20 band the 2026-09 Desk Review already flagged (8 instances in 8 weeks) and is double-blocked on the strategy's own composite (exhaustion_score = 1.102 × 0.39 ≈ 0.43 vs the 0.80 floor); XLE's 1.011 lands in the already-tested floor=1.00 cell (2026-09-05 deliverable: 2 new trades, 50% WR, PF 0.128); XBI 0.832 is wide. Reading: regime, no floor conversation, no gate touched. The separate, desk-scoped plan-layer finding is that the plan's "a HIGH-event-risk shock day lifts participation" hypothesis is now 0-for-2 (09-24: SPY vol_ratio 1.166 sub-floor on a gap-down macro-shock open; 09-28: all three sub-floor on a live oil/yields shock with QQQ true range 1.29× ATR) after twice being used to lift p_trade above realized base rates. N=2, below the Desk Review's N ≥ 3 bar — filed as a plan-layer observation, not an escalation, and no gate or threshold is implicated.

Flag for learnings.md: house-echo / SURGE — 2026-09-28 is the desk's first session where no gate bound and the trade lost anyway: SURGE's precondition chain completed (first-hour establishment cleared; QQQ VWAP-touch pullback with retrace_pct 0.414; resumption at ADX 34.53 vs the 22.0 floor) inside the 10:00–12:00 window, and the position was stopped 30 minutes later (entry 734.37 @11:40 → exit 735.75 @12:10, net −$172.50). The plan had named first_hour_signal/pullback_resumption structural absence as its single most likely outcome; the structure formed instead, so this is a plan-accuracy miss in the opposite direction (Brier 0.64) rather than a gate near-miss — the first time in the desk's recent record that structure-completion was observed and did not pay. Two carried observations attach, neither an action: (1) the entry was in the strong-trend ADX band (34.53) — one point for TEMPER's moderate-trend-outperforms finding, though the live split still leans opposite it (strong 3W/3L net +$459.85 vs moderate 2W/2L net −$118.92); (2) the trade was a short, and the live short book is now 8 trades, 4W/4L, net −$127.35 against longs 2 trades, 1W/1L, +$468.28 — the charter's "short side has not yet shown it pays" is now "short side is net negative," still on a sample far too small to review. No gate touched.

Desk Summary for Orchestrator

Same-day-on-trade reflection for 2026-09-28 — desk net −$172.50, the desk's first fill in three sessions and it was a loss. SURGE 1 trade: QQQ short 25 sh, entry 734.37 @11:40, STOP 735.75 @12:10, −$172.50 (strategy=v9, signal −0.92%, ADX 34.53). No gate bound — the three-phase structure completed (the plan had called structural absence as its most likely outcome) and the resumption failed; forecast 0.20/short, Brier 0.64. Cumulative +$340.93 (n=10, 50.0%, SPRT CONTINUE, LLR +1.414, now +1.531 from CONSISTENT — the loss pushed it off the boundary for the first time in two weeks). ECHO 0 trades: all three baseline instruments cleared the morning signal (SPY −0.39%, XLE +0.84%, XBI −0.44%) and all three blocked on vol_ratio — 1.102 / 1.011 / 0.832 vs 1.2 — on a 1.29× ATR macro-shock day that the plan had explicitly expected to lift participation. SPY's 1.102 is a 9th instance in the Desk Review's flagged 1.08–1.20 band but is double-blocked (exhaustion_score ≈0.43 vs 0.80); XLE's 1.011 is in the already-tested floor=1.00 cell (PF 0.128). Cumulative −$290.45 (n=13, 46.2%, SPRT CONTINUE, LLR −0.082). Verdict: regime — no floor conversation, no gate touched. Desk cumulative +$50.48. Regime watch: current.json as_of 2026-09-22, stale by four sessions; no shared tag stressed (spy_vs_sma50 above 50d, spy_drawdown_bucket <2%, vix_regime LOW with live VIX 16.07); charter-consistent on substance, persistence rule not armed, no posture change. Two flags filed for learnings.md (one per member); runner health CLEAR. Two plan-layer items carried: ECHO's shock-day participation lift is 0-for-2 (N=2, not escalated); SURGE's short book is now net negative (−$127.35 over 8 trades, sample far too small to review).