Reversal Desk — 2026-08-06

Desk Plan — Reversal — 2026-08-06

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-08-06-pm.md PM: Wicker

Shared Market Read

• Event risk today: HIGH (briefing) — no Tier-1 today (firm-wide GO), but NFP is TOMORROW (Fri Aug 7 — both members' hard-skip fires then, not today), the memory complex gaps down from last night's after-hours cluster (WDC -11%, SNDK -5%, APP -16–20% — the third leg of the week's earnings-reversal cycle), SPCX's first lockup (~$123B eligible, ~20% of shares) hits the tape today, and the Hormuz deal window is open with Trump's "tomorrow or the next day" deadline landing today. Pre-open data: Challenger 7:30, Initial Jobless Claims 8:30 (205K est / 197K prior — the labor pre-read the day before NFP), Q2 Productivity 8:30. COP (the oil/Hormuz read-through) reported ~7 AM.
• Session character expected: Mixed / event-driven, two-tier — Dow/S&P modestly higher on deal hopes + healthcare/energy + super-Thursday European earnings; Nasdaq under pressure from the memory/tech gap-downs. Name-level structure dominates over index direction into tomorrow's NFP. The briefing's calibration flag applies in full: when the AM's own marquee risk names gap down hard, continuation for the AI/memory complex is a conditional to stress, not the base case — base case is divergence/event-driven tape.
• VIX regime: 15.85 (LOW, <18) — both members' strongest historical cohort (Null's low-VIX backtest WR 80.1%; three of Null's four live wins logged low_vix; Meridian's lone win was low_vix). Low-VIX gaps are precisely the noise-gap population Null's thesis targets.
• Key levels:
• QQQ: prior close 717.10, prior high 728.42, prior low 716.95, SMA20 700.85 — QQQ above, ATR(14) 15.56. Prior-day range 11.47. Null gap thresholds: 0.1× ≈ ±1.56 pts (±0.22%), 0.3× ≈ ±4.67 pts (±0.65%), 0.7× ≈ ±10.89 pts (±1.52%). NDX futures -87.50 (-0.30%) → implied QQQ gap ≈ -2.1 pts ≈ 0.13× — in-band, tiny-to-small.
• SPY: prior close 769.79, prior high 776.81 (sweep level, +0.7% over the implied open ~771.4), prior low 769.53, SMA20 748.37, ATR(14) 9.65. Prior-day range 7.28. S&P futures +16.25 (+0.21%).
• The swing change of the day: Meridian's volatility gate passes for the first time in five sessions (QQQ ~0.96×, SPY ~0.98× vs the 1.25× ceiling after Wednesday's quiet session) — the ADX band, not the vol gate, is now Meridian's binding constraint.
• Macro backdrop: Two-tier — deal-hope + earnings support the broad tape while the memory/AI complex's beat-and-dump reversals (third consecutive session of the pattern) keep tech under pressure; Fed-hike narrative live again (Governor Cook, first FOMC-member hike talk since the July 29 hold), oil coiled at ~$75–76 on the Hormuz window — any announcement breaks oil lower and drags the energy-heavy S&P leg.

Null — MNQ/MES (gap fade / FVG)

Setup present: likely. The implied cash gap is in-band and trend-aligned for the first time in three sessions — and this is the aligned branch, not a relaxation:

• Branch A (base case, aligned): gap-down open → LONG fade. NDX futures -0.30% implies QQQ opens ~-2.1 pts (~0.13× ATR), and QQQ (717.10) is above SMA20 (700.85) → gap-down fades (long) toward prior close 717.10 are the permitted, aligned direction. The gap-down is name-driven churn (memory complex — none of it top-10 QQQ weight), not index repricing — precisely the noise-gap population the thesis targets. The two-tier tape (Dow bid, Nasdaq weak) is a divergence/rotation day, the friendliest character for gap-fill mechanics: no strong index trend to fight the reversion.
• Branch B (relaxation): tiny gap-up < 0.3× (under ≈ +4.67 pts / +0.65%) → SHORT fade via the tiny-gap relaxation — the exact shape of the live winners (07-20, 07-31, 08-05). Less likely than Branch A given futures are down, but the 08-05 lesson stands: the cash auction decides, not the futures. Direction resolves at entry by structure, not pre-market.
• Blocked: gap-up in the 0.3–0.7× band → trend filter (08-04's exact block). No-trade: |gap| < 0.1× — a flat open is a live branch today (NDX -0.30% is close to the band floor; the 08-03 bust was a 0.027× flat open, and the 08-05 reflection's dispersion lesson applies: on news-driven overnights the cash gap size, not just direction, is resolved at the auction).

Timing: Claims + Productivity land 8:30 — pre-open, so they set the gap rather than interrupt the window. No mid-window prints today (ISM was yesterday) — the 9:30–10:30 window is clean except headline risk: a Hormuz announcement or SPCX-lockup headline can hit any time. No FVG by 10:30 = no trade.

Earnings filter (tier=2): CLEAR — no top-10 QQQ name (AAPL/NVDA/MSFT/AMZN/GOOGL/META/TSLA/AVGO/COST/NFLX) reported after Wednesday's close (WDC/SNDK/APP/FIGMA/DUOL/Z/DASH/ELF — none top-10) or pre-market today (COP/BRK.b/KVUE/SRE/AKAM/FOX/RL/TFX — none top-10). No T+1 residual tag. NVDA is the sector tone, not a filter event — and the read-through line for the whole desk: NVDA +3.4% Wednesday on the SpaceX chip commitment; if it holds, the memory rout stays contained to memory names; if it breaks, semis cascade and Branch A's fill thesis is at risk.

Sizing: standard — 0.5% risk, 1 MNQ (paper). No event-risk reduction: NFP is tomorrow's hard-skip, not today's, and the exposure window ends 11:00 — no NFP-into-position risk. The risk control is the gates, not the size: pre-market NQ volume vs 2× the 5-day average (a memory-cascade morning can spike it — if triggered, stand down: institutional commitment to the gap direction) and the cascade invalidation below.

Key levels: entry = 1-min FVG equilibrium after the open drive; target = prior close 717.10 (long branch) or gap-fill level (short branch); stop beyond FVG extreme + 0.25× ATR, max 1.5×; hard exit 11:00; daily loss limit 1.5%.

Invalidation: gap > 0.70× either way (committed move); gap-up 0.3–0.7× (trend filter); |gap| < 0.1× (flat open); pre-market NQ volume > 2× the 5-day average; no clean 1-min FVG by 10:30; memory-cascade continuation: gap-down long where the open drive extends beyond 1.0× ATR without a reversal impulse — stand down, the gap is trending, not filling (this is the stressed conditional the briefing's calibration flag names); post-claims one-way breakaway with no reversal impulse.

Meridian — QQQ/SPY (kill-zone sweep + RSI(2), bearish only)

Setup present: possible — the first live session in five. The volatility gate finally passes, and the binding constraint shifts to the ADX band. This is the swing change the 08-05 reflection called: Wednesday's quiet session (QQQ range 11.47, SPY 7.28) put both legs under the ceiling for the first time since 07-27:

• Volatility gate: PASS both legs. QQQ prior range 11.47 vs 20-day avg (~11.9) ≈ 0.96× vs the 1.25× ceiling; SPY prior range 7.28 vs avg (~7.4) ≈ 0.98×. The 08-03/08-04/08-05 streak (1.33×/1.57×, 1.38×/1.38×, 1.52×/1.73×) is broken. The gate is no longer the pre-closed door.
• ADX band (18–32, 15-min) — now the swing gate:
• QQQ: treat as blocked. 38.1 on 07-31, then a ~+3.9% two-day surge (Mon/Tue) — per the 07-24 TEMPER arithmetic, a 14-period smoothed indicator that far over the 32 ceiling cannot resolve into the band in one session, especially after only one quiet day.
• SPY: the swing factor. 22.2 on 07-31, spiked hard through Mon/Tue's rally, one quiet day (Wed) decaying it — the honest read is uncertain at the 32 ceiling (30s, could be either side). This single number at the open decides whether today is a session or a fifth straight stand-down.
• Sweep structure: SPY prior high 776.81, implied open ~771.4 → +5.4 pts (+0.7%) overhead — reachable, and for the first time in the streak the direction points toward the sweep: S&P futures +0.21% on the deal-hope bid (06-18 lesson in reverse — direction context now supports the push-up). The chain: push up, spike through 776.81, close back below in the same bar (a clean close above = breakout, not a sweep), RSI(2) > 88 at the sweep candle, ADX re-check 18–32, bearish FVG within 15 minutes, entry, before the 11:30 hard close.
• Session-character fit: moderate. Mixed/event-driven two-tier is not the clean trend-up tape where sweeps are structurally rare, and not a down-tape with no fuel. The one hostile vector: Hormuz headline risk — any deal announcement breaks oil lower, XOM/CVX (heavy SPY weights) drag the index down mid-window, and a down-move is not the push-up the sweep needs; the push must come in the open drive before any headline. COP (reported ~7 AM, watchlist reads in-line) is the morning's oil anchor.

Sizing: standard if a leg clears — $250 risk per trade, MES cap 5, hard close 11:30, max 2 losses/day. No event-risk reduction: NFP is tomorrow's hard-skip, not today's, and the 11:30 close keeps the book flat well before any NFP positioning. The discipline today is mental, not mechanical: after five pre-closed sessions, the ADX read and the sweep chain decide — not the dry spell, and not the expansion-backtest pressure (see ledger below).

Key levels: sweep SPY 776.81 / QQQ 728.42; targets 1.5R or prior lows (SPY 769.53 / QQQ 716.95) if closer; stop 2.0× 5-min ATR beyond the swept wick.

Invalidation: SPY ADX outside 18–32 at the open (QQQ treated as blocked); no push up into 776.81 (down-tape has no fuel); clean close above 776.81 (breakout, not sweep); sweep without RSI(2) ≥ 88 (structure without exhaustion — 07-27); no FVG within 15 minutes of the sweep candle; hard close 11:30 unchanged.

⚠️ Reconfirmation ledger: Meridian's expansion backtest is due 2026-08-09 — 3 days out, binding (TEMPER: "binding, not advisory"; failure to materially raise the firing rate → park candidate, escalates to ZEUS). No progress last period. This is the desk's open obligation — it outranks any single session.

Plan Filed

• Filed: 2026-08-06 07:45 ET
• Frontmatter forecasts complete for every active member: yes
• Wicker (PM): strategy retired 2026-07-14 (WS2c Tier-2 NO-GO) — no plan, no forecast; persona continues desk oversight only.

Reversal Desk Daily Reflection — 2026-08-06

PM: Wicker — for desk eyes only.

What Happened

One trade in the window since the 08-05 reflection (same-day reflection — Null traded): Null long QQQ, +$284.50 net, 5/5 in Phase 3. Meridian stood down at the ADX band (QQQ) and sweep mechanics (SPY) after the volatility gate passed both legs for the first time in five sessions — the swing change the plan called, with the binding constraint shifting exactly as projected. Desk window P&L: +$284.50. Desk cumulative (post-July 14): +$1,420.85 (Null +$1,222.60, Meridian +$198.25).

Session character was the two-tier reversal the plan framed, with the tier direction flipped: the deal-hope trade unwound — the Dow led the downside (-0.85%, worst of the three indices) as oil rebounded on the Iran vessel-barring report (Brent +4%+) and the Warsh-hike narrative rekindled, while the Nasdaq recovered from a ~0.9% gap-down open to close roughly flat (-0.06%) as SPCX +6.1% (lockup supply absorbed) and MSFT +2.57% offset the memory rout (WDC -13%, SNDK -6.8%) and the software pain trade (APP -19%, DDOG -16-18%, HUBS -21%). Ranges were quiet (SPY 0.47× ATR — the week's quietest — QQQ 0.71×), VIX fell to 15.15 (LOW), both indices closed above SMA20. NFP is tomorrow (08-07) — both members hard-skip. The data block (claims 199K beat, Challenger ~33K three-year low, productivity +1.4% beat) was quiet and drove nothing.

Null — traded, +$284.50 (fifth Phase 3 trade, fifth win)

Null went long QQQ at 709.975 at 09:35 (1 MNQ, stop 705.4961), target = prior close 717.10, hit at 10:10. Net +$284.50 (gross +$285.00, commission $0.50), ~1.6R. The plan's base case — Branch A, gap-down long fade — fired: the cash gap printed -6.60 pts (~0.42× ATR, above SMA20), larger than the futures-implied -2.1 pts (~0.13×) but in-band and trend-aligned; the open drive was modest (low 708.59, ~0.12× ATR of extension), the 1-min bullish FVG formed against it, entry at equilibrium, and the reversion ran the full distance to prior close in 35 minutes. The memory cascade the briefing stressed stayed name-level (none of WDC/SNDK/APP is top-10 QQQ weight), so QQQ's gap-down was precisely the noise-gap population the thesis targets — the cascade invalidation never came close to triggering. VIX 15.15 (LOW): four of Null's five live wins now logged LOW, the 80.1%-WR backtest cohort. This is the mirror of Wednesday's short-QQQ fade — both worked the same two-tier axis from opposite sides.

Binding gate question: N/A — full pass, traded. The one forecast finding is gap-size dispersion, not a gate: estimated -0.13×, printed ~0.42× — the third data point on cash-auction gap resolution (07-24 direction, 08-03 size, 08-06 size again). p_trade 0.5 / long / med → traded long, hit (Brier 0.25).

Meridian — no trade, $0 (ADX band, QQQ; no_sweep, SPY)

The desk forecast the swing change and it landed: the volatility gate passed both legs for the first time in five sessions (the 08-05 reflection's call), and the binding constraint shifted down the chain exactly as projected — QQQ blocked at the ADX precheck (35.6 vs band [18,32]; 38.1 on 07-31 cannot resolve into the band in one session per the 07-24 arithmetic, and the live read is decaying but still 11% over the 32 ceiling), SPY blocked at sweep mechanics (session high 771.80 vs sweep level 776.81 — ~5 pts / 0.65% short; the deal-hope leg the market expected to supply the push reversed into the session's biggest loser, so the push-up never got fuel). The chain never started. Sixth straight stand-down — and this one resolves the 08-05 tension in the opposite direction: the gate that had blocked five sessions finally passed, but the shape did not appear. The gate stack is working; the binding constraint just moved, as planned.

Binding gate: ADX band (QQQ) + sweep mechanics (SPY) — verdict: regime, not calibration. No threshold grazed: 35.6 is 11% over the ceiling (decaying on schedule from 38.1), and the sweep level sat 0.65% above the session high — a structural miss. No TEMPER conversation warranted on the gate. What the sixth straight stand-down sharpens is the firing-rate question: the expansion backtest (due 08-09 — 3 days, binding, no progress) is the desk's highest-priority obligation; it must answer whether expansion materially raises firing rate, or Meridian becomes a park candidate (that decision escalates to ZEUS). p_trade 0.2 / short / low → no trade (Brier 0.04) — the stand-down calibration stays accurate at the floor.

Scoring and process notes

• Frontmatter forecasts scored for both active members: Null 0.5 / long / med → traded long, hit (Brier 0.25); Meridian 0.2 / short / low → no trade (Brier 0.04). Mean desk Brier 0.145.
• No signal_detail_2026-08-06.md files exist for either member. Desk plan filed 2026-08-06 07:45 ET covering both members.
• Desk-blind learnings.md flag candidates carried in each member's reflection for TEMPER's weekly audit (not appended here — TEMPER's file): Null's third cash-auction gap-size dispersion data point + fifth live win; Meridian's vol-gate-pass → ADX/sweep binding-constraint shift with the live ADX decay path (38.1 → 35.6).

Paper P&L

Desk window P&L: +$284.50. Desk cumulative (Null + Meridian, post-July 14): +$1,420.85. Both SPRT monitors remain inside their continuation bands; Null's LLR (+1.060) keeps climbing toward the CONSISTENT-WITH-BACKTEST boundary (+2.944).

Event Risk vs. Expectation

• Plan rated HIGH; correct in substance. No Tier-1 today (firm GO), NFP tomorrow, and the named risk legs behaved as framed: the memory complex cascaded exactly as named (WDC -13%, SNDK -6.8%, APP -19% — all among the S&P's biggest drags), the Hormuz window produced the oil rebound on the negative side (Iran vessel-barring report), and the Warsh-hike narrative extended. All of it stayed name-level and sector-level — the session was two-tier and quiet, not an index event — which is exactly the structure that let Null's gap-fill run and denied Meridian's sweep its fuel.
• The tier-direction flip is the desk-relevant lesson. For the third straight session the two-tier structure dominated, but the direction of the tiers flipped: the deal-hope/Dow leg (the side the tape expected supported) was the session's biggest loser, while the Nasdaq leg (expected under supply) held roughly flat. For this desk the flip was immaterial to Null (gap mechanics don't care which tier wins) and decisive for Meridian (no push-up fuel into 776.81). Desk-plan framing for this regime: name the hostile vector per member — Meridian's sweep needs the "supported" leg to actually be supported; when the deal-hope leg is the fragile one, the sweep chain should be rated structurally absent.
• VIX 15.15 (LOW, falling) — the regime cohort for both members stays favorable; Null's low-VIX live wins now 4/5.
• Data block immaterial to the desk — claims 199K beat / Challenger ~33K / productivity +1.4% beat set the gap pre-open and drove nothing after; they matter as the labor pre-read into tomorrow's NFP, which is a hard-skip for both members.
• Next session (Fri 08-07): Tier-1 NFP 8:30 ET — both members hard-skip; no trade expected; next live window Monday 08-10. Meridian's expansion backtest due 08-09 (2 sessions out, binding) remains the desk's open obligation.

Reflection Filed

• Filed: 2026-08-06 18:45 ET
• Frontmatter forecasts scored for both active members: Null p_trade 0.5 / long / med → traded long, hit (Brier 0.25); Meridian p_trade 0.2 / short / low → no trade (Brier 0.04). Mean desk Brier 0.145.
• Load-bearing gate question answered per member: Null — no gate question (traded); the gap-size estimate missed by ~3× (est -0.13×, printed ~0.42×) — the third cash-auction dispersion data point, a plan-calibration observation, not a gate issue. Meridian — regime, not calibration: the ADX precheck blocked QQQ at 35.6 (11% over the 32 ceiling, decaying from 38.1 on the 07-24 arithmetic) and SPY never reached its sweep level (high 771.80 vs 776.81) — no grazed threshold on either leg. No TEMPER conversation warranted on either stack.
• Reconfirmation ledger: Meridian expansion backtest due 2026-08-09 — 3 days out, binding, no progress. Sixth straight stand-down, now across two different binding gates (vol ×5, then ADX/sweep), makes this the desk's highest-priority obligation. Wicker feed/session re-run — still outstanding, unchanged.