The Nightcap White-Cap

Built 4:55 PM ET · 2026-06-25 · validates the 6/25 regular-session close (Massive /v3/snapshot, COALESCE(session_close,session_price); breadth from /v2/aggs/grouped 6/24→6/25) anchor: system-reminder 2026-06-25 == Bash date (agree) · static — regenerate to refresh

01Session Scorecard — 2026-06-25

6 setups in the open backlog; 6 finalized this run — 1 FIRE / 3 VOID / 2 MIXED / 0 NO_EVIDENCE. The day resolved as a third path neither morning fork called: May PCE printed in-line (core +0.32% MoM / +3.41% YoY), so the hot-PCE rejection short never armed and QQQ failed to hold the 720.85 reclaim (closed 716.38) — both Curd forks voided. The edge was the midday rotation pivot (SRB-ROT-L FIRE): a broadening tape (ADRN 1.56, 7/11 sectors green, RSP +0.65% > SPY +0.14%), not a failing one.

Setup	Outcome	Evidence	Δ-ATR	Actual
MS L SMH,QQQmm-260625-MS-SEMI-L	VOID	Opening-range-reclaim semis long required QQQ to hold 720.85; QQQ reclaimed at the open then lost it, closing 716.38 below (primary-level kill). Micron sold the news — SMH faded 650.03→613.93 intraday. PCE in-line / dollar soft (kill 1 not met); SPY held above 733 (kill 4 not met) — but the 720.85-hold thesis failed. Semis still closed green on the day.	—	QQQ 716.38
LRT S SPY,QQQmm-260625-LRT-SPY-S	VOID	Conditional gap-fade short that arms ONLY on a hot core PCE + rejection candle + A/D<1. PCE printed in-line (+0.32% core, benign), SPY showed no rejection (closed +0.14% at 734.30, high 739.37 never reached the 739.95 trigger), breadth positive (ADRN 1.56 > 1, the named kill). Never armed — the tape rotated risk-on, it did not reject.	0.47	SPY 734.30
SRB L RSP,IWMmd-260625-1209-SRB-ROT-L	FIRE	Equal-weight / broad-rotation long fired cleanly into the close: RSP +0.65% beat SPY +0.14%, IWM +0.75%, cyclical-value + healthcare led (XLI +2.17%, XLB +1.33%, XLV +1.49%; CAT +6.29%, DE +5.00%, UNH +2.40%), megacap-growth funded it (XLY -1.49%, XLC -0.90%). SPY tagged the 729.60 low and held above 731 (closed 734.30) — no kill. The day's one correct read.	—	RSP 211.75
MS S QQQ,SMHmd-260625-1209-MS-SEMI-S	MIXED	Megacap-growth/semis short (the funding leg to ROT-L). Right leg: QQQ stayed capped below 720.85 (closed 716.38) and megacap-growth closed red (XLY -1.49, XLC -0.90, NVDA -1.64, GOOGL -0.46). Wrong leg: SMH recovered +2.90% to 636.88, back above its 633.38 VWAP — the semis-stabilize kill. No sustained fade into the close; the 711 target tag (low 705.30) preceded the 12:09 signal.	0.22	QQQ 716.38
Backfilled — 2026-06-21 Sunday Sundae (weekly, event-gated; finalized this run)
LRT S SPYsun-260621-LRT-SPY-S	VOID	Weekly highs-rejection short, event-gated on a hot-PCE rejection at 750–757. SPY's weekly high (739.37) never reached the rejection zone and PCE printed in-line (the explicit “cool PCE” kill), breadth positive (ADRN 1.56). Conditional never armed; finalized VOID — the gating event has passed and cannot re-arm this week.	—	SPY 734.30
MS L SMHsun-260621-MS-AIInfra-L	MIXED	Weekly semis/AI-infra long. Catalyst confirmed (Micron blowout, MU +15.74%) and SMH closed green +2.90% (above the 6/24 618.92) — but the explicit gap-and-fade kill triggered (650.03→613.93 intraday) and SMH sits below the 6/21 weekly reference 659.88 (-3.5% WTD). Catalyst right, clean follow-through failed.	—	SMH 636.88

LensNeither morning fork fired — the semis-continuation long and the hot-PCE rejection short both voided when PCE came in-line. The session's signal was the midday rotation pivot: a broadening tape (RSP>SPY, 7/11 green, ADRN 1.56), not distribution. The semis legs split — megacap-growth stayed capped while the chips themselves recovered off the lows (SMH 614→637).

02Calibration

0.2435 Rolling Brier (last 50) ↓ from 0.2607 (6/24) · 0.25 = random

42% Hit rate (rolling 50) 40.9% over all 66 eligible

66 Calibration-eligible 6/6 scored this run · 0 NO_EVIDENCE gaps

Rolling Brier dipped below the 0.25 random line for the first sustained reading — the rotation read pulled the trailing-50 window to 0.2435. No data-feed gaps this run: all six setups scored from confirmed Massive closes, with breadth computed from the grouped daily tape (no NO_EVIDENCE fallback).

Pattern performance — Component 13 (calibration-eligible; n≥3)

Pattern · Dir	n	Hit%	Mean prob	Mean Brier	Evidence class
LRT · S	18	61%	0.35	0.354	lore_pending
MS · L	14	29%	0.49	0.261	lore_pending
SRB · L	7	71%	0.41	0.299	lore_pending
SBD · L	6	0%	0.37	0.138	primary_source
GFD · L	4	75%	0.41	0.298	practitioner_backtest
MS · S	4	75%	0.44	0.318	lore_pending
SBD · S	4	0%	0.28	0.079	primary_source
LRB · L	3	33%	0.43	0.216	lore_pending

Plus six n=1 singletons (FBR-S, GFU-S, SRT-S, VAB-L, VBR-L, VSR-L) — too thin to read. SRB-L (rotation) is the north-star edge at 71% over 7, reinforced by tonight's FIRE; MS-L continuation is the weakest at 29% over 14.

Calibration trend — Component 14 (rolling Brier by validation date)

06-100.311

06-120.301

06-150.289

06-160.283

06-180.278

06-220.271

06-230.272

06-240.261

06-250.2435

Mechanism leak · C15 8/21 right-direction / wrong-mechanism (rate 0.38) — dominated by LRT-S (4). Both of tonight's voids were leaks: mm-260625-LRT-SPY-S (highs capped as predicted, but the hot-PCE mechanism never fired) and md-260625-1209-MS-SEMI-S (megacap-growth capped, but semis stabilized).

LensThe model is improving and now marginally better than a coin flip (Brier 0.2435 < 0.25). The persistent failure is mechanism, not direction: the LRT short keeps being right that the highs cap (61% hit) yet keeps getting voided by its arming kill — a setup-design leak worth re-specifying. Continuation baskets (MS-L) remain over-confident through event risk.

03Tape & Rate Backdrop

Index	Close	Day	Note
SPY	734.30	+0.14%	o738.91 / hi 739.37 / lo 729.60 — held the 731 50-day, never tested 740
QQQ	716.38	+0.81%	reclaimed 720.85 at the open, lost it, closed below; lo 705.30
IWM	298.91	+0.75%	small-caps green, tagged 301.50 (just shy of the 299.49 record on a closing basis)
RSP	211.75	+0.65%	equal-weight beat cap-weight SPY — the broadening tell

Realized regime: in-line-PCE risk-on ROTATION — a broad-but-rotational, two-sided chop. The morning's binary (semis-continuation up vs hot-PCE rejection down) resolved as a third path: benign May PCE (core +0.32% MoM / +3.41% YoY; headline +0.45% / +4.07%) cleared the event gate, the dollar softened, yields stayed contained, and the median stock had a strong day while the cap-weight index barely moved. Dominant drivers: macro (PCE) + catalyst (Micron).

VIX: refresh-required — Massive does not carry index symbols (I: = 403) and FRED VIXCLS posts T+1; last confirmed 18.63 (FRED VIXCLS 6/24 close). A calm, broadening tape with flat HYG is consistent with VIX holding the high-17s/18s — unconfirmed for 6/25.
10Y yield: ~4.50% (FRED DGS10 6/23; refresh-required for 6/25) — contained, no rate-path repricing on the in-line print.

LensThe cap-weight index did almost nothing (SPY +0.14%) while breadth was strongly positive (ADRN 1.56) — the textbook signature of a healthy rotation, not distribution. The PCE gate was the whole game: an in-line print took both tail scenarios off the table.

04Cross-Asset

Asset	Proxy	Close	Day	Read
US Dollar	UUP	28.48	-0.18%	soft — confirms risk-on / disinflation
Crude oil	USO	109.31	+2.84%	firm, but no inflation scare on the in-line PCE
Gold	GLD	369.46	+0.97%	up with a soft dollar (GDX +1.45%)
Long Treasuries	TLT	87.35	-0.03%	flat — yields contained
HY credit	HYG	79.88	+0.04%	flat / calm — zero credit stress under the rotation
Bitcoin	IBIT	33.52	-1.03%	the one risk-off divergence

LensSoft dollar + calm credit (HYG flat) corroborated the risk-on rotation. The only non-confirms were crypto (IBIT -1.0%) and megacap-growth — i.e. the funding sources of the rotation, not a risk signal.

05Sector & Breadth (realized)

XLKTech+0.83

XLCComm-0.90

XLYDisc-1.49

XLVHealth+1.49

XLIIndu+2.17

XLFFin-0.50

XLBMatl+1.33

XLEEnrgy+0.97

XLPStaples-0.59

XLUUtil+0.68

XLRERE+0.18

7 green / 4 red. Leaders: Industrials (XLI +2.17), Healthcare (XLV +1.49), Materials (XLB +1.33), Energy (XLE +0.97). Laggards: Consumer-Disc (XLY -1.49), Comm-Svcs (XLC -0.90, a GOOGL drag), Financials (XLF -0.50, faded from midday green).
Equal-weight > cap-weight: RSP +0.65% beat SPY +0.14% — broadening, not narrowing.
Breadth internals: ADRN proxy 1.56 (2,399 adv / 1,541 dec across 4,001 liquid names ≥$5 & ≥200k vol; computed from Massive grouped 6/24→6/25 — confirmed). $S5FI / $S5TH / $TICK / $TRIN refresh-required (Massive 403 indices; BarChart not reached this headless run).

LensBroadening confirmed three independent ways — RSP>SPY, 7/11 sectors green, ADRN 1.56 — with leadership rotating into cyclical-value (industrials/materials/energy) + healthcare, funded by megacap-growth (XLY/XLC). The classic relief-rotation signature.

06Single-Name Movers (illustrative)

Name	Close	Day	What it says
MU	1213.56	+15.74%	the catalyst — FQ3 blowout, but sold from its +17% premarket high
CAT	1057.01	+6.29%	machinery led the rotation
DE	630.76	+5.00%	industrials/ag-equipment, the broadening leadership
SMH	636.88	+2.90%	semis recovered off the 613.93 low (sell-the-news, then bid)
UNH	415.53	+2.40%	healthcare participation
NVDA	195.74	-1.64%	megacap-growth funded the move (GOOGL -0.46%)

LensThe tape in six names: Micron's blowout (+15.7%) was sold from its premarket high while real money rotated into machinery (CAT/DE) and healthcare — exactly the equal-weight leadership the midday flagged. Illustrative of market state only; close-bell newsletters were not pulled this headless run, so moves are drawn from confirmed Massive closes — not trade recommendations.

07Morning & Midday Lens vs Reality

The morning Early Bird Curd framed a clean binary around the 8:30 PCE print: a semis-continuation long if QQQ held its 720.85 reclaim, or a hot-PCE rejection short if SPY rejected 739.95 with breadth rolling under 1. Reality took neither. PCE printed in-line, so the rejection short never armed; and QQQ failed to hold 720.85 (closed 716.38), so the continuation long voided too. The Curd correctly identified the two pivots that mattered — the PCE gate and the 720.85 level — but over-weighted the two tail paths and under-weighted the in-line / rotation middle.

The midday Frappé (12:09) caught exactly that: it pivoted to a broad-rotation / equal-weight long (RSP>SPY, cyclical-value + healthcare leadership) and a megacap-growth funding short. The rotation long was the only setup to fire, and it fired cleanly into the close. Net: the lens improved intraday, and the day rewarded waiting for the PCE print over pre-positioning either morning fork.

LensProcess takeaway: on a binary event day, the edge was in the update, not the open. The morning's job was to mark the gate (PCE) and the level (720.85) — which it did — and the midday's job was to read the resolution — which it did. The two voided forks weren't bad calls; they were correctly-gated bets whose arming conditions never triggered.

08Execution Debrief — realized execution on closed setups, NOT a forward trade plan

Setup	Profile	Entry	Stop	Target	Realized R
md-260625-1209-MS-SEMI-S	day	720.85	722.37	711.00	+2.94
mm-260625-LRT-SPY-S	swing	739.95	757.90	704.05	+0.31
mm-260625-MS-SEMI-L	swing	720.85	690.27	782.01	n/a

md-260625-1209-MS-SEMI-S +2.94R — the QQQ leg marked to the 716.38 close (below the 720.85 entry); a real day-trade win even as the setup scored MIXED (SMH recovered).
mm-260625-LRT-SPY-S +0.31R (hypothetical) — the short never armed (SPY high 739.37 never reached the 739.95 trigger on an in-line PCE); R is a mark-to-close artifact, not a taken trade.
mm-260625-MS-SEMI-L: excluded — the tool's −2.75R is a cross-instrument artifact (SMH's 636.88 close marked against a QQQ-scale 720.85 entry, because inst lists SMH first). Logged ISS-20260625-01; omitted from the honest tally.

Dashboard Execution Debrief (all-time, Component 16): n=22, mean R +0.56, win 68%, stop 9%.

LensOn closed setups only: the QQQ semis-short marked +2.9R into the close while the hot-PCE rejection short never triggered — the discipline of waiting for the arming print rather than pre-shorting the gap is exactly what the gate enforced. Realized execution on already-closed, scored setups — not a forward trade plan.