House Mneme — 2026-08-13

desk: house-mneme date: 2026-08-13 forecasts: mneme: p_trade: 0.5 direction: long conviction: low trace: p_trade: 0.25 direction: long conviction: low

Desk Plan — House Mneme — 2026-08-13

Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-08-13-pm.md PM: MNEME

Shared Market Read

• Event risk today: MEDIUM — tradeable, NOT a hard-skip. July PPI + jobless claims at 8:30 AM ET (consensus benign: +0.2% MoM / +0.3% core / ~4.9% YoY; claims 202K vs 199K). PPI is second-tier — neither MNEME (SPY) nor TRACE (DIA) carries it in the pre-registered Tier-1 hard-skip set (shared/event_calendar.py, tier_filter=2). The runners will evaluate signals today — no no_trade:event_day pre-close. Firm-wide status per watchlist: GO (no Tier-1 in the 2-day window; FOMC Minutes Aug 19 is the next Tier-1). AMAT reports after close — a discrete, non-intraday tail; no window in either stack holds past 16:00 ET, so nothing carries into it.
• Session character expected: News-driven / two-sided. PPI benign-expected argues for a repeat of the muted gap-up-fade pattern, BUT the after-hours AI-complex reversal (CSCO −4% despite a double beat — priced-in; CBRS −16%; COHR −3%) injects a live, second risk leg that could turn the tape into sector-divergence or Nasdaq-led weakness. This is precisely the two-tier-divergence session class that MNEME's 07-24 TEMPER learning flags as its failure mode (rotation-driven afternoon reversal after a morning up-pattern). The Dow/DIA is the likely firmer leg (cyclicals, less AI concentration). Expect structure not to form cleanly until PPI dust settles (~9:00–9:30 AM ET).
• VIX regime: 14.60 — LOW (<18). Both stacks' best backtested regime (MNEME VIX<20 PF 2.39; TRACE VIX<20 PF 2.46). Today this is not moot (unlike the 08-12 CPI skip) — LOW VIX is the regime in which both mechanisms' edge is strongest, and it is a live tailwind for any qualifying signal.
• Key levels: SPY prior close 772.54 (08-12), prior range 771.30–774.74, SMA20 753.15 (above), ATR(14) 8.37. DIA not in the briefing price block — TRACE's recent Phase 3 band ~$515–524 (last entry 07-27 at $523.31). Both indices hold above SMA20; broad uptrend intact.
• Macro backdrop: The week's inflation gauntlet ends with a benign-expected PPI that, if in line, keeps the CPI-cleared risk-on bias intact — but the market is now testing whether the AI-demand complex is "confirmed and priced" (reversal) vs. "still a new catalyst" (relief), with an oil/Hormuz supply tail (Brent ~$89) as a background risk.

MNEME — SPY KNN (p_trade 0.5, long, low)

Setup present: likely to evaluate, with a genuine two-tier divergence risk. Today is tradeable — no hard skip. My three windows all observe after the 8:30 PPI print (A 9:30–10:00, B 10:00–12:00, C 12:00–14:00), so the 8:30 macro noise is largely digested before my observation bars begin — PPI is not my CPI-shaped spurious-match class. My library has real analogues for LOW-VIX, sub-ATR, above-SMA20 SPY sessions that gap modestly and drift; this is a recognizable microstructure day, not a black-swan print. The honest tension is two-tier divergence: the briefing explicitly anticipates sector divergence (Nasdaq soft on AI gap-down vs. S&P/Dow firm), and my 07-24 learning is that such days reverse between observation and execution windows. That is a quality risk on whatever fires, not a reason to expect no signal — I rate a real trade as more likely than not (0.5) but keep conviction low because the divergence class is my documented failure mode.

• Sizing vs. event risk: Standard 0.75% ($187.50), one position — PPI at 8:30 is a tradeable second-tier catalyst, not a sizing input; the event filter is a gate, and it does not trigger today. No defensive sizing, no shadow book. The two-tier divergence risk argues for discipline on the exit (respect the 1.5× ATR stop immediately), not for reduced size — my edge is regime-backed (LOW VIX) and the divergence penalty is expressed through the stop, which I honor.
• Session-character fit: Partial — the LOW-VIX sub-ATR base case is my home turf, but the two-sided AI-reversal overlay is the divergence shape I've lost on. Window B (observe 10:00–12:00, execute 14:00–16:00) is the most vulnerable — the afternoon is where a rotation-driven reversal bit me on 07-24. Window A (9:30–10:00 observe) is cleanest if the PPI dust settles quickly.
• Key levels: SPY prior range 771.30 (low) / 774.74 (high), prior close 772.54. A sustained hold above 774.74 (record-extension shape) favors a clean continuation; a fade back toward/below 771.30 (open-high-fade) is the divergence-fade shape I must respect. Whether SPY holds the 772–773 pivot into the 10:00–12:00 observation tells me if the day is converged risk-on or split-vote.
• Invalidation (stand down): If the AI-complex gap-down broadens into a SPY gap-down > 1% at the open and VIX ticks above 18 on a hot PPI surprise, I sit out the AM session entirely. A hot PPI (reaccelerating producer inflation) that undercuts the benign-CPI risk-on narrative flips the day from my recognizable LOW-VIX continuation library into an event-repricing day — the spurious-match class.
• What I'm watching to learn: (1) Whether the AI gap-down (CSCO/CBRS/COHR) stabilizes into support or extends — a clean fade into support is a divergence shape, an extension is a trending-down tape; (2) the two-tier lens in reverse of 07-24: is the Dow (firm) leading S&P while Nasdaq (soft) lags — and does that divergence widen into my afternoon execution window? (3) Whether any signal fires on the flagged Window C (holdout PF 1.25) — the C rolling record is the standing watch after the 08-11 stop.

TRACE — DIA KNN (p_trade 0.25, long, low)

Setup present: possible but genuinely muted by two stacked constraints. Today is tradeable (PPI not in my Tier-1 skip), and my instrument is arguably the better-positioned side of this session — DIA's cyclicals/industrials (CAT, HON, MMM, CVX) are insulated from the AI-complex gap-down and stand to be the firmer leg of a two-sided tape. That directional case is real. But two things lower my genuine probability of a real trade: (1) my B/C no_data data-availability pattern (sharpened 08-11 to "early window works, later windows don't") — the 24-bar windows that historically carry my signal have been returning no bars, which structurally suppresses my firing rate; and (2) my recent vote profile has been resolving low (08-11 A at 30%, max_sim 0.880) — the setup has not come close to the effective 70% bar. I rate a real trade at 0.25: the A window can evaluate, but both the data path and the recent vote distribution argue against a clean fire.

• Sizing vs. event risk: Standard 0.75% ($187.50), one position — same reasoning as MNEME; PPI is not a sizing gate. Separate live-risk flag: my SPRT is at LLR −1.350, −1.595 from DEGRADED (CONTINUE). Today is the first tradeable (non-skip) session since that state set in — the DEGRADED protocol cannot start on a no-trade day, but if I log a qualifying trade today and it loses, the pre-registered DEGRADED protocol begins immediately. This is a monitoring headwind, not a sizing input — I do not pre-emptively shrink size on it.
• Session-character fit: Poor-to-moderate. The two-sided tape is directionally favorable to the Dow, but the persistent B/C data gap is the binding constraint on my firing rate, independent of the macro. A clean A-window signal is the only realistic path today.
• Key levels: DIA band ~$515–524 (last entry 07-27 at $523.31). Whether the Dow reclaims the band on the firm leg — and specifically whether an open-high-fade (my 6-of-7 losing shape) recurs or the cyclicals hold gains — tells me if my recent losing shape is a two-sided-tape phenomenon or a normal-Dow weakness.
• Invalidation (stand down): If BOTH the A window returns no_data AND my recent low-vote pattern holds (vote < ~50%), no qualifying signal can form — I close without a trade. Absolute invalidation: if DIA gaps > 1% and VIX breaks 18 on a hot PPI, I sit out — the Dow would be dragged into an event-repricing tape, not a recognizable pattern-library day.
• What I'm watching to learn: (1) Whether B/C data returns today (a tradeable day is the first honest test of the 08-11 "early window works, later don't" diagnosis since it was sharpened — a skip day added no reading); (2) whether the Dow's cyclicals leg leads the firm side of the two-sided tape (tests whether my industrial tilt is the right instrument for this divergence); (3) the 60%-cluster question — today's vote, if A resolves, is an honest reading (not event-contaminated the way a CPI day would be), so a repeat 60% (6-of-10) resolution would be the seventh same-shaped instance.

Pooled note (Family 5)

Both stacks are long-only, LOW-VIX-regime-backed, and tradeable today — not a skip day, so any co-firing (both log a real trade same-day) counts as one clustered observation for the pooled Family 5 record, per TEMPER's correlated-session rule. I forecast MNEME the more likely to trade (0.5) and TRACE muted by data availability (0.25); both direction long (the mechanism is long-only). The session's defining risk for both is two-tier divergence — the exact class MNEME has lost on and TRACE's cyclical tilt is better-positioned for. Each stack sizes independently; neither adjusts standard 0.75% for the MEDIUM tradeable event. GHOST stays frozen (Phase 2, Decision D2) — no plan.

Plan Filed

• Filed: 2026-08-13 07:47 ET (pre-open — briefing/watchlist both pre-open, authoritative)
• Frontmatter forecasts complete for every active member: yes
• GHOST: FROZEN (Phase 2) — not planned for
• Firm-wide status: GO — PPI/jobless claims 8:30 AM second-tier (tradeable), no Tier-1 in 2-day window; AI-complex after-hours reversal (CSCO −4% / CBRS −16% / COHR −3%) is the live two-sided risk leg

House Mneme Desk Reflection — 2026-08-13

Plan reference: desks/house-mneme/plans/2026-08-13-plan.md EOD briefing: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-08-13-eod.md

What Happened

Period since last reflection (2026-08-11 → 2026-08-13 — two sessions, one a hard skip)

Combined desk cumulative (Phase 3): +$418.25 across 24 trades (17 MNEME +$767.46, 7 TRACE −$349.21), 41.7% WR. The 23-trade pool from the 08-11 reflection gains one observation — MNEME's 17th.

MNEME — a high-conviction fire stopped out in morning chop on a day that then went its way

MNEME traded Window A long: entry 777.61 @ 10:00, stop 776.69, stopped out 776.9 @ 11:25, −$68.16 (96 sh). Crucially this was not a minimum-pass fire — the A window resolved at vote 73.3% (11 of 15), max_sim 0.8725 from the full past_n=1515 pool. That clears both hard gates and sits above the "exceptional-strength" stance (vote > 70% AND max_sim > 0.65) the plan keeps writing; for once the gate and the stance agreed. The fire was genuinely high-conviction.

The loss is the most instructive kind. The plan flagged two-tier divergence as MNEME's 07-24 failure mode — the afternoon rotation reversal that stops a morning up-pattern late. Today was the mirror image: the divergence produced morning shakeout chop (the AI names bled, the index knocked back to 776 in the first hour), MNEME's 1.5× ATR stop filled at 11:25, and then the macro leg won — SPY extended to a record close of 777.84, above the 777.61 entry. MNEME was on the correct side of a session that ultimately went up and was shaken out before the move the setup predicted. Both the feared reversal-shape and this realized shakeout-shape express through the same ATR stop; neither is a gate error.

Binding-gate verdict for MNEME: no gate binding — the setup fired strongly (vote 73.3%, sim 0.8725) and the loss is distributional, a stop taken in two-tier chop before an extension. There is no near-miss row to read; the trade filled and lost on the exit mechanism. Consistent with SPRT (CONSISTENT-WITH-BACKTEST, sticky from #14, LLR +3.594, −6.538 to DEGRADED). The honest adjustment is plan-calibration, not a gate action: on a benign-macro, trending-up, LOW-VIX day the plan over-weighted the AI single-name gap-down as an index-level risk instead of the macro leg — the same composition error the EOD briefing grades PARTIAL firm-wide. That framing cost a morning-chop stop on a continuation day. The one watch thread (Window C's rolling record, pre-registered 15-trade-PF<1.2 monitor) is untouched and not near triggering; today's stop was on Window A, MNEME's cleanest backtested window.

TRACE — first honest tradeable reading: A evaluates (regime no-signal), B/C still blank

TRACE logged no trade, and the note is a two-part, both-instructive story: blocked=knn_gate A:vote=50%/62% sim=0.865/0.6 B:no_data C:no_data. Because 08-12 (CPI) was a hard skip, this was the first tradeable, non-event-contaminated reading since the 08-11 diagnosis sharpened to "early window works, later windows don't."

• Window A resolved at 50% (5 of 10), max_sim 0.865 — a genuine regime no-signal (the DIA library did not produce a decisive long on a firm, benign-macro, trending-up tape; 0.865 sim says the candidate pool matched, but the vote is well below the effective 70% bar and the 62% floor). Not a near-miss. And it is an honest reading — not CPI-distorted — so it confirms there is no suppressed signal hiding behind the B/C data gap: the setup was genuinely not there.
• Windows B and C (24-bar) still returned no_data — the window-dependent observability pattern confirmed on the first tradeable test. A arrives and evaluates; the two 24-bar observation windows (B is TRACE's cleanest backtested window, holdout PF 3.00) do not.

Binding-gate verdict for TRACE: A regime (50% — setup never came close, wait); B/C observability (data availability). No calibration signal — the 60%-cluster did not advance (today a 50%, not a 60%; still six distinct sessions at exactly 6-of-10), so the calibration-side TEMPER conversation remains unchanged, one honest decisive-vote short. SPRT CONTINUE, n=7 (W1/L6), LLR −1.350, −1.595 from DEGRADED; the dry stretch since TRACE's last qualifying trade (07-27) is now twelve sessions, and the DEGRADED clock still cannot start on a no-trade day. The concrete B/C data shape is the standing hand-off for the data-infra / runner-logging diagnosis.

Plan calibration since last reflection (WS4b rows — 2026-08-11, 08-12, 08-13)

MNEME's all-long forecast logged another traded-direction hit today (long, Brier 0.25) — including on a losing day, the standing reminder that direction accuracy and P&L are separate axes (the long was the correct read of a day that closed +0.69% to a record; the loss came from the intraday stop, not a wrong direction). TRACE's muted 0.25 calibrated cleanly (Brier 0.0625) but for the recurring wrong-exercised reason — A resolved low and B/C blanked, so the score rewards a session that produced little information. MNEME's direction hit rate across the period is now strong (long called and long traded on 08-10/11/13); TRACE's probability calibration continues to look honest but under-informed.

Desk-level observations

• No same-day clustered observation this period — MNEME traded, TRACE did not, so the Family 5 pool adds one unclustered observation (the 24th); no correlated-session counting issue.
• The intra-desk asymmetry this period is the same shape as 08-11, one turn further. MNEME exercised its mechanism with real conviction (73.3% vote / 0.8725 sim — the strongest fire of its recent run) and took a small, well-risk-managed loss that a two-tier morning chop forced; TRACE finally got a genuine, honest, tradeable DIA read (a 50% A vote) and was again shut out of B/C by data. One member fired high-conviction and lost on chop; the other confirmed its data story and declined on a genuinely-absent setup.
• Two standing threads, both evidence-gated, neither a gate action: (1) MNEME's plan-calibration miss on two-tier composition (over-weighted the AI name gap-down as an index-level driver on a macro-led day — cost a morning-chop stop on a continuation; watch that session-character framing, not the gates); (2) TRACE's B/C data-availability pattern, now confirmed on a tradeable day as window-dependent (A works, the 24-bar windows don't) — the concrete, confirmed hand-off for the data-infra diagnosis. Neither is calibration evidence that would license a unilateral change.

Paper P&L

Combined desk P&L this period: −$68.16 MNEME cumulative Phase 3: 17 trades, +$767.46 net, 52.9% WR, SPRT CONSISTENT-WITH-BACKTEST (sticky, LLR +3.594) TRACE cumulative Phase 3: 7 trades, −$349.21 net, 14.3% WR, SPRT CONTINUE (LLR −1.350, −1.595 to DEGRADED) Combined desk cumulative: 24 trades, +$418.25 net, 41.7% WR

Event Risk vs. Expectation

• Event risk call (MEDIUM — July PPI + jobless claims 8:30 AM ET, second-tier): Correct. Benign PPI printed even softer than consensus (flat headline, +0.2% core, YoY easing to 4.7%) and released a risk-on bid. Tradeable as rated for both stacks — not a Tier-1 skip; no event risk materialized, no hot-surprise leg. 08-12 (CPI) was the Tier-1 hard skip for both, correctly observed.
• Session character call: Partial miss (composition). The plan forecast "News-driven / two-sided" and leaned on the two-tier AI-complex divergence as the live index risk. The actual session was a clean trending-up tech-led rally to a record — the AI names (CSCO/CBRS) bled but the weakness never broadened; the macro leg overwhelmed the single-name divergence and resolved the two-tier in the macro's favor. The EOD briefing grades this firm-wide PARTIAL; for this desk it cost MNEME a morning-chop stop on a day that then extended in its direction. The direction of both members' lean (long) was correct; the composition read (which leg drives the index) was the miss.
• Invalidations: none triggered for either stack — no >1.5% gap at the open, VIX stayed LOW (14.63, never >18), no hot-PPI repricing flip. MNEME's B-window two-tier gate was never reached (A fired); TRACE's stand-down fired exactly as pre-registered (A resolved low + low-vote pattern held → closed without a trade).
• Forward risk (next sessions): Retail Sales (Jul) + Michigan Consumer Sentiment Friday 08-14 (8:30/10:00 AM ET, second-tier MEDIUM — tradeable); AMAT after-close today sets the semis gap; CSCO/CBRS divergence watch (stabilize vs. keep bleeding). Next Tier-1: FOMC Minutes Wed 08-19 — hard skip for both stacks by pre-registered rule.

Reflection Filed

• Filed: 2026-08-13 (EOD reflection, desk cadence)
• Next session event risk (from EOD briefing): MEDIUM — Retail Sales + Michigan Sentiment Friday 08-14 (second-tier, tradeable); next Tier-1 FOMC Minutes Wed 08-19 (hard skip for both members).
• GHOST: frozen (Phase 2, Decision D2) — no reflection written.
• MNEME and TRACE per-member reflections filed.
• Flags embedded for learnings.md (desk-blind, own members): MNEME — fired Window A long with real conviction (vote 73.3% = 11/15, max_sim 0.8725, past_n 1515 — clears the hard gates and the advisory "exceptional-strength" stance), entry 777.61 @ 10:00, stopped 776.9 @ 11:25, −$68.16, and SPY then extended to a record close of 777.84 (+0.69%) above the entry. The two-tier divergence produced the morning shakeout chop that took the stop, then the macro leg won — the mirror image of MNEME's 07-24 failure mode (afternoon rotation reversal stopping it late). Both express through the same 1.5× ATR stop; no gate error. Binding-gate verdict: no gate binding (high-conviction filled trade, loss distributional from a chop stop); SPRT CONSISTENT (LLR +3.594, −6.538 to DEGRADED). The genuine adjustment is plan-calibration — on a benign-macro trending-up LOW-VIX day, over-weighted the AI single-name divergence as an index-level risk instead of the macro leg (the composition error the EOD briefing grades PARTIAL); cost a morning-chop stop on a continuation. Direction forecast still hit (long, Brier 0.25) — direction accuracy is distinct from P&L. TRACE — blocked=knn_gate A:vote=50%/62% sim=0.865/0.6 B:no_data C:no_data on 2026-08-13: first honest tradeable (non-skip, non-event) reading since the 08-11 diagnosis sharpened. Window A evaluated and resolved at 50% (5-of-10, sim 0.865 vs 0.60 floor) — a regime no-signal (healthy sim but vote well below the 62% floor / effective 70% bar), and because it is an honest reading it confirms no suppressed signal hides behind the B/C gap. Windows B/C (24-bar) stayed no_data — the window-dependent observability pattern confirmed on a tradeable day (A works, the 24-bar windows don't), the concrete confirmed shape for the data-infra/runner-logging diagnosis. The 60%-cluster did not advance (today a 50%, not a 60%; still six sessions). SPRT CONTINUE, LLR −1.350, −1.595 from DEGRADED, twelve sessions dry since 07-27; DEGRADED clock cannot start on a no-trade day. Binding-gate verdict: A regime (wait), B/C observability; no calibration signal, no unilateral tweak.