desk: house-mneme date: 2026-07-22 forecasts: mneme: p_trade: 0.35 direction: long conviction: low trace: p_trade: 0.25 direction: long conviction: low
Briefing reference: /Users/dadbot/Desktop/ClaudeBod/projects/dadbrain/Analysis/briefings/2026-07-22-pm.md PM: MNEME
Setup present: possible, not likely. The consolidation/pre-positioning character is structurally neutral for KNN — the session is unlikely to produce a strong directional pattern either way. MNEME's recent 5-trade record shows a session-character bifurcation (wins on clean post-data sessions, losses on narrative-driver sessions). Today is a pre-earnings holding pattern, which is closer to the "clean" category than the "narrative-driver" category, but the absence of a directional catalyst during the RTH session reduces the probability of a distinctive pattern for the KNN to match.
Key levels for SPY: If the open stays flat to slightly negative (~747-748), the SMA20 at 744.94 is the nearest support. A dip below 745 would be a structural concern given SPY's constructive posture above SMA20. Resistance at prior-day high 749.03.
Sizing vs. event risk: Standard 0.75% per trade. The after-hours earnings are outside MNEME's session window (all three windows close by 16:00 ET). No position ever carries into the after-hours catalyst — the hard flat exit ensures this. Standard sizing is appropriate.
Session-character fit: LOW — consolidation days are the least structurally favorable for KNN pattern matching. The market's "rhyme" is harder to detect when the session has no clear directional impulse. Combined with the LOW VIX regime (where MNEME's PF is weakest at 2.39), the setup likelihood is below average.
Invalidation: If SPY opens below 744 (below SMA20) after the prior day's constructive close, the session character changes from consolidation to a gap-down test of the recovery — reduce position sizing or skip. A gap-up above 750 (exceeding prior-day high by >1 full ATR) would also be structurally unusual for a pre-earnings consolidation day and may indicate pre-positioning ahead of GOOGL/TSLA that is not reliable.
Setup present: unlikely. Two structural headwinds today. First, IBM's 22% pre-market gap-down as a Dow component creates a mechanical drag on DIA at the open that is outside the KNN's pattern library — the KNN has never seen a Dow component crater 22% in a single pre-market session, so any similarity match to a "normal" DIA open is based on a fundamentally different structural context. Second, TRACE's recent Phase 3 record is weak (1 win / 3 losses, WR 25%, SPRT LLR -0.289 — within 2.655 of the DEGRADED boundary at -2.944). The single win came at max_sim 0.8795 (far above the 0.60 gate); the losses came at 0.8574, 0.7707, and 0.6013. This suggests trade quality correlates with how far above gate the similarity score sits — a finding consistent with TRACE's backtest showing near-zero similarity spread within the qualified cohort.
Key levels for DIA: No specific DIA levels in the briefing, but the prior-day close for DIA is approximately correlated with SPY's close. The IBM gap-down implies a 0.3-0.5% mechanical drag on DIA at the open (IBM is ~3-4% of DJIA by price weighting). If DIA opens below prior-day support, the KNN's historical patterns from non-IBM-crisis days may not be applicable.
Sizing vs. event risk: Standard 0.75% per trade if a signal fires. The after-hours earnings are outside TRACE's session window. However, given the IBM headwind and the weak recent live record, a discretionary sizing reduction to 0.50% is worth considering if a marginal signal fires.
Session-character fit: VERY LOW — the IBM gap-down is a structural event that DIA's KNN pattern library has never encountered. Any similarity match to a historical DIA session with a flat-to-negative open on a pre-earnings day would be based on a fundamentally different structural context (no 22% single-stock gap-down). The KNN contamination diagnostic (2026-07-16) confirmed that even the presence of blackout days in the candidate pool did not degrade the edge — but that was a general finding about event-day contamination, not a specific test of single-stock catastrophic gap-downs.
Invalidation: If DIA opens more than 0.8% below prior close (accounting for the mechanical IBM drag plus any broader market weakness), the session character is no longer "pre-earnings consolidation" but "crisis-gap" — skip entirely. The KNN's pattern library contains no meaningful analogue for a Dow component losing 22% overnight.
No trades taken.