The Nightcap White-Cap · 2026-06-15

Built 2026-06-15 22:32 ET · anchor: system-reminder = Bash = 2026-06-15 (Mon EDT), agree · subject_date 2026-06-15 (validated session) Static after build — regenerate to refresh.

01. Session Scorecard

Validated this run — Monday 06-15

Setup	Outcome	Evidence	Δ-ATR	Actual
MS L SMH/NVDA/AVGOmm-260615-MS-SMH-L	FIRE	Beaten-down semis reclaimed the US–Iran-peace / oil-relief risk-on gap: SMH +4.14% closed near the high (649.24); the extended-at-open dip to 639.25 (below VWAP 645.50) was reclaimed; NVDA +3.29 / AVGO +3.33 never rolled red. No kill — semis held bid, SPY held 745.34, yields flat (TLT −0.19).	—	SMH 647.10

Backfilled — Friday 06-12 (Friday Nightcap did not run)

MS L SPY/QQQ/IWMmm-260612-MS-SPY-L	FIRE	Primary risk-on continuation long held the 737.76 reclaim and extended (SPY 741.75) on broad participation; none of the three kills (lose 737.76 / Iran-oil-reversal / breadth-narrows-to-megacap) triggered. Monday's gap to 754.83 ratified it.	—	SPY 741.75
LRT S SPY/QQQmm-260612-LRT-SPY-S	VOID	Conditional late-day exhaustion fade never armed: its precondition (breadth narrows into the weekend) resolved opposite — broad risk-on — and SPY's 744.44 high never tagged the 745.34 entry. The clean-break-above-745.34 kill also stayed untriggered. Correct stand-aside.	0.39	741.75

Still open — FOMC-gated (windows active through Wed 06-17 FOMC / Thu 06-18 quad-witch)

SBD L SPY/QQQ/SMHmm-260615-SBD-SPY-L	OPEN	Interim — wall-of-worry reclaim long. The reclaim > 745.34 is already in hand (SPY 754.83), but it fires only on a non-hawkish Warsh (Wed 2pm). No kill (y10 ~4.42 < 4.60; SPY >> 737.76). Scored at the post-FOMC Nightcap.	—	↻ pending
LRT S SPY/QQQmm-260615-LRT-SPY-S	OPEN	Interim — level-rejection short into 755–760 supply. SPY tagged 756.68 and closed 754.83 above the 745.34 cap on 7-of-11 breadth, so the fade-from-755–760 path is live into the FOMC; not yet triggered, not yet killed.	—	↻ pending
SBD / LRT L Ssun-260615-SBD-SPY-L · sun-260615-LRT-SPY-S	OPEN	Weekly-frame (Sundae) twins of the two FOMC-gated theses above — the same long-on-non-hawkish-Warsh / short-on-breadth-fail fork. Scored together at the post-FOMC Nightcap.	—	↻ pending

Lens Risk-on broke the read wide open. The momentum-scalp semis long and Friday's continuation long both fired on a broadening, oil-relief tape; the Friday level-short correctly stood aside (never armed) — the short that didn't trigger is a win, not a miss. The week's two directional bets are deliberately parked on Warsh: the tape has already reclaimed 745.34, so Wednesday's FOMC is the single binary that resolves both the wall-of-worry long and the level-rejection short.

02. Calibration

0.2889Rolling Brierlower = better · 0.25 = random

39.4%Hit rate13 FIRE / 33 eligible

33Calibration-eligible35 validated · 2 NO_EVIDENCE excluded

Pattern performance — which setups carry an edge

Pattern	Dir	n	Hit rate	Mean prob	Mean Brier	Evidence class
MS	L	9	44%	0.55	0.319	lore_pending
LRT	S	8	50%	0.40	0.352	lore_pending
SRB	L	5	60%	0.40	0.293	lore_pending
SBD	S	4	0%	0.28	0.079	primary_source
GFD	L	2	50%	0.45	0.205	practitioner_backtest

Five single-instance patterns (n=1) omitted: GFU-S, SRT-S, VAB-L, VSR-L (all VOID/NE) and LRB-L (FIRE). Hit rate = FIRE / eligible; VOIDs are correct stand-asides, not losses.

Calibration trend — cumulative Brier by validation date

05-270.490

05-280.410

05-290.339

06-080.333

06-090.326

06-100.311

06-120.301

06-150.289

Mechanism leak 4 right-direction / wrong-mechanism — the SBD breadth-divergence short is 4-of-4 MIXED: directionally right every time, but its defining breadth/sentiment-divergence trigger never armed (the bearish payoff kept arriving via unmodeled catalysts — NFP, CPI — not the modeled leading divergence). It is well-calibrated (mean Brier 0.079, mean prob 0.28), so this is a setup-design leak, not overconfidence. Two isolated instances also seen (MS-SMH 06-11, LRT 06-09). Diagnostic, not a trade plan.

Lens Calibration is on a clean glide path — cumulative Brier has fallen every validation date from 0.49 to 0.289, now just above the 0.25 random line. The cleanest edge is SRB energy-reactivation longs (60% hit, best Brier); the subtler edge is discipline — LRT and SBD VOIDs are correct stand-asides, and SBD's low Brier shows the model already prices it as unlikely to cleanly fire. The one place to keep tightening is momentum-continuation longs (MS-L: mean prob 0.55 vs 44% hit) running slightly hot through event risk — the AI-infra-dispersion lesson, restated.

03. Tape & Rate Backdrop

Index closes: SPY 754.83 (+1.76%) · QQQ 744.00 (+2.99%) · IWM 294.64 (+0.58%) · RSP 212.88 (+0.58%). The Dow notched records (Daily Rip). Cap-weight led decisively — QQQ/SPY ran while equal-weight and small-caps lagged. VIX ~16.3 est. (AM anchor · cash-close refresh required), compressed in contango.

Dominant driver: catalyst, not rates — the weekend US–Iran peace framework + Strait of Hormuz reopening (a 60-day pause) collapsed oil and gapped equities above the 745.34 June-8 cap. Rates backdrop: 10y ~4.42% est. (AM anchor · close refresh required), TLT −0.19% (realized, ~flat) — yields were a non-event today. The week funnels into Warsh's inaugural FOMC Wed 06-17 (a hold is ~97% priced; the dot-plot and tone are the binary, with a December hike still live on the hot CPI/PPI).

Realized regime — RISK-ON RELIEF breakout (confirmed, not a forward call): the morning's "range-breakout attempt above 745.34" resolved up and held — SPY's 751.76 low stayed well above the cap, which flipped to intraday support. A soft Empire State print (5.7 vs 13.0 cons) did not derail it.

Lens The breakout is real, but read the leadership: a cap-weight, semis-and-megacap, oil-relief tape (QQQ +2.99 vs RSP +0.58) two days ahead of a Fed that hasn't spoken. Real breakout; durability is Warsh's call Wednesday.

04. Cross-Asset

UUP +0.07% (dollar flat) · USO −3.36% (crude −4.2% per Daily Rip) · GLD +2.59% · TLT −0.19% · HYG +0.13%.

Cross-asset ratified the equity move: credit firm (HYG green), dollar flat, and crude down hard on the Hormuz de-escalation — a disinflationary tailwind that quietly helps the bull into the Fed. The one diverger is gold: +2.59% on a risk-on, lower-oil day is gold trading its own structural bid, not a fear hedge.

Lens Oil down + credit firm + dollar flat = clean risk-on with a disinflation tailwind. Gold is the lone cross-current — worth watching, not yet worrying.

05. Sector & Breadth (realized)

XLKTechnology+3.78%

XLVHealthcare−0.60%

XLFFinancials+0.41%

XLYCons. Cyclical+1.69%

XLPCons. Defensive−0.40%

XLEEnergy−3.48%

XLIIndustrials+1.42%

XLUUtilities+0.47%

XLBBasic Materials+0.61%

XLREReal Estate−0.82%

XLCComm. Services+0.48%

7 green / 4 red. Tech led (+3.78), energy was worst (−3.48, with crude); defensives (XLV/XLP) and real estate red — textbook oil-relief risk-on rotation. Equal-weight vs cap-weight: RSP +0.58% badly lagged SPY +1.76% — the index did its lifting through semis and megacap. S5FI ~61.2 / S5TH ~61.0 est. (AM anchor · intraday TICK/TRIN/ADRN refresh required) keep the breadth-thrust regime intact above 60.

Lens Seven-of-eleven green is broad enough, but the RSP/SPY gap says it isn't broadening — this was a cap-weight engine, not a participation surge. The S5FI ~61 thrust is the bull's structural support if the Fed cooperates.

06. Single-Name Movers illustrative only — never trade recommendations (Daily Rip close)

Semis the standout: MU +10.8%, WDC +16.1%, NVDA +3.5% ($212.45, on a reported $20–25B+ senior-notes sale — its first bond-market visit in five years), SMH +4.14%. AI-infra capex bid widened beyond chips: AMZN +3.1% on a multibillion Missouri data-center buildout / $200B 2026 capex. Supply wave: SPCX +20% (raised $85.7B; +40% since Friday's IPO). Also: ROKU/Fox media-M&A chatter; FISV sold off on its CEO's departure to Truist.

Lens The tell was the AI-infrastructure bid widening from chips to power, land, and balance sheets — Amazon's $200B capex and Nvidia's $20B+ debt sale on the same day. Constructive for the theme; the SpaceX +20%/$85.7B raise is the same coin's other face — the late-cycle supply wave the Sundae flagged. Illustrative only.

07. Morning Lens vs Reality

The Early Bird Curd framed today as a RISK-ON RELIEF range-breakout attempt above 745.34, conviction medium-low, event-gated — and the lens held on all three counts. SPY gapped above the cap and closed 754.83 (breakout confirmed, not faded); the MS-SMH-L semis-continuation long fired (+4.14%); and the week's directional bets correctly stayed parked on Warsh rather than chasing the gap. The honest caveat the Curd flagged — "extended at open, favor a pullback/AVWAP-reclaim entry" — was vindicated when SMH dipped to 639.25 (below VWAP) before reclaiming and closing near its high.

Lens A clean lens day: the Curd called risk-on-above-cap, the semis continuation, and the Warsh-gated patience — all three landed. The real test is whether Wednesday's Fed lets the breakout keep its gains.

08. Execution Debrief realized execution on closed setups — not a forward trade plan

Setup	Profile	Entry	Stop	Target	Realized R
LRT S mm-260612-LRT-SPY-S	swing	745.34	759.23	717.56	+0.26*

*Notional — this conditional short never armed (SPY high 744.44 < 745.34 entry), so no capital was committed; the figure is the counterfactual entry→close R, not a realized trade. One closed scored setup carried an exec block this window; the three FIRE/VOID outcomes otherwise scored were zone/continuation theses with no entry/stop/target. Realized execution on closed setups — not a forward trade plan.

Lens The discipline read, not a P&L read: the level-short the gate would have armed at 745.34 never triggered because price stalled at 744.44 — the no-entry was correct, and the notional +0.26R confirms even a hit would have been a small, low-conviction scratch. Nothing to chase here.