The Nightcap White-Cap

Monday, 07-20-2026

Evening market read · post-close validation & recap

The Milkman
OuroTaurus
Built Mon 2026-07-20 · ~4:35 PM ET · validates the 07-20 regular-session close (Massive consolidated) · automated Nightcap run (4:31 PM slot, on time) static — regenerate to refresh
The 30-second skim

01Session Scorecard

Five setups scored against tonight’s regular-session close — four asset-forecast leans from The Early Bird Curd (07-20) and one intraday fade from The Midday Frappé (07-20, 1:21 PM). 2 FIRE / 2 VOID / 1 MIXED / 0 NO_EVIDENCE. Two weekly setups from The Sunday Sundae (07-19) remain open on an unclosed holding window. Backlog after this run: 2 open (both weekly).

SetupOutcomeEvidenceΔ-ATR*Actual
AF LONG CPERmm-260720-AF-CPER-L FIRE Copper closed 38.42 (+1.32%, +0.78 ATR in-direction) — the second-strongest asset in the basket and the cleanest expression of the day’s reflation/growth bid. No growth-scare or dollar-spike kill (UUP only +0.21%). +0.78 38.42
LRT SHORT QQQmd-260720-1321-LRT-QQQ-S FIRE The failed-breakout / VWAP-rejection fade played out: after tagging 705.80 and failing the 702.30 gate, QQQ closed 696.06 — right at the 695.33 prior-close target shelf, below the 699.96 VWAP and the 698.08 trigger low. Neither kill fired: no decisive reclaim/hold above 702.30, and sector breadth stayed 8/11 red into the close. 696.06
AF SHORT XLPmm-260720-AF-XLP-S MIXED Staples closed 84.86 (−0.39%) — the right direction for the short, but only +0.22 ATR, inside the 0.25-ATR decisiveness bar. Defensives were sold (thesis intact) but not decisively enough to score. outcome_binary 0. +0.22 84.86
AF LONG XLFmm-260720-AF-XLF-L VOID Financials closed 56.04 (−0.39%, −0.29 ATR against the long). The premarket green faded — banks sold with the broad tape (SPY −0.16%, breadth negative) rather than leading. Just over the 0.25-ATR bar → VOID. −0.29 56.04
AF LONG XLImm-260720-AF-XLI-L VOID Industrials closed 178.12 (−0.72%, −0.42 ATR against the long) — the cyclical bid that copper confirmed at the commodity level did not carry the equity sector; XLI rolled with the soft breadth. VOID. −0.42 178.12
OMR LONG SMHsun-260719-OMR-SMH-L OPEN Weekly oversold-reversal, arms Wed evening only (after Alphabet reports) on an SMH reclaim of 567.41 + QQQ >702.30 — not yet triggerable. Interim: not armed, not killed. SMH closed 558.83 (>536.81 kill) and QQQ 696.06 (>686.76 kill); the constructive Monday bounce (SMH +0.41%) keeps it alive into its window. Stays open. 558.83
EXF SHORT XLEsun-260719-EXF-XLE-S OPEN Weekly exhaustion-fade, zone-only / not armed (needs a genuine de-escalation headline + a failed high; the Sunday note flagged it unlikely to arm this week while the strait stays shut). Interim: XLE closed 57.94 (+0.45%), a marginal intraday high (58.385) but off the high — no decisive new-high breakout to kill it, no arm trigger. Stays open, leaning toward the fade never expressing. 57.94

*Δ-ATR shown for asset-forecast rows is the close-vs-previous-close move in daily-ATR(14) units (the AF scoring metric); — where the mechanical validator did not populate it (the midday row carries a descriptive level, not a numeric one).

LensThe book split exactly along the day’s fault line: the two calls that rode the commodity/reflation impulse (copper long, and the QQQ fade that read the soft breadth) fired, while the two that bet the cyclical bid would lift the equity sectors (financials, industrials) voided when breadth stayed red. The morning correctly saw a risk-on-growth lean in the tape; it over-extended that lean into sectors the average stock dragged down.

02Calibration

0.2210 Rolling Brier (last 50) lower = better; 0.25 = random · best of the record · from 0.2228 (07-16), −0.0059 vs 07-15
60.0% Hit rate (last 50) all-time 48.6% over 138 eligible
138 Calibration-eligible 141 validated − 3 NO_EVIDENCE/unevaluable · +3 scored tonight

Pattern performance (Component 13) — edge by pattern × direction (n ≥ 5)

PatternDirnHit%Mean probMean Brier
AF (asset-forecast)Long1974%0.550.224
AF (asset-forecast)Short1457%0.550.235
SRT (sector-rotation)Short667%0.470.317
SRB (sector-rotation)Long862%0.430.299
MS (mean-reversion)Long3145%0.520.249
LRT (late-reversal fade)Short3244%0.390.278
SBD (small-body doji)Long60%0.370.138

Calibration trend (Component 14) — rolling-Brier (last 50) by validation date

07-08: 0.2483 / 46% 07-09: 0.2352 / 54% 07-10: 0.2323 / 54% 07-15: 0.2269 / 60% 07-16: 0.2228 / 60% 07-20: 0.2210 / 60%

Mechanism-leak (Component 15) deferred this run. The canonical leak count and the dashboard xlsx refresh both crashed on a numeric comparison against the midday setup’s descriptive-string level (lvl); filed ISS-20260720-10. The scored outcomes were written cleanly by score_setup before the augmentation step, so calibration above is unaffected; only the leak tally and the .xlsx are stale (last built 07-18).

LensSix straight sessions of Brier improvement have pulled the trailing-50 to its best mark (0.2210) — the record is now meaningfully better than a coin-flip (0.25). The engine of that gain is the asset-forecast long bucket (74% hit at n=19); the standing drag is the LRT short-fade (44% at n=32), the pattern most worth interrogating next.

02bMacro-Prediction Calibration

A distinct calibration source — the macro / cross-asset prediction track carried in the thinktank-v2 DB (index, FX, commodity, crypto targets), read-only and never combined with the setup Brier/hit-rate above. Rendered from the reader label verbatim: est. (thinktank-v2 derived, asOf 2026-07-20).

33% Macro hit-rate 2 HIT / 4 MISS over 6 resolved
6 / 11 Resolved / total 5 unresolved — no price series (score N/A, never guessed)
InstrumentDirTargetEntryCloseRealizedResult
SPX (S&P 500)up+0.5%7369.07537.4+2.29%HIT
BTC (Bitcoin)up+4.0%6036563086+4.51%HIT
ES (S&P Sep)down−1.5%7385.37537.4+2.06%MISS
NQ (Nasdaq Sep)down−2.5%2934729698+1.20%MISS
NQ (Nasdaq Sep)short−3.0%2920029698+1.71%MISS
GC (Gold, spot)up+2.0%4046.04104.1+1.44%MISS

Pending / unresolved (no price series — never scored by guess): XLK/XLU pair, IWM/QQQ pair, Brent (BZ), USD/JPY, Wheat (ZW). Read-only — the DB was not written.

LensThe macro book’s tell this window: the outright up calls (SPX, BTC) hit, while every down/short index bet missed as equities kept grinding higher — a directional-bias miss, not a sizing one. Small sample (6 resolved); flagged distinct from the setup record on purpose.

03Tape & Rate Backdrop

IndexCloseDay %
SPY (S&P 500)742.09−0.16%
QQQ (Nasdaq-100)696.06+0.11%
RSP (S&P equal-wt)212.42−0.45%
IWM (Russell 2000)292.31−0.59%
SMH (semis)558.83+0.41%
Realized regimeNarrow risk-on-at-the-top, soft-breadth underneath — a megacap/semis-led index with a commodity-reflation overlay (copper, oil up; yields up) but negative internals (equal-weight, small-caps, 8/11 sectors red). Not a broad risk-on; a two-speed tape. (Realized, not a forward call.)

04Cross-Asset

ProxyCloseDay %Read
CPER (copper)38.42+1.32%Reflation/growth bid — the day’s strongest cyclical signal
USO (WTI oil)125.51+1.25%Energy firm — supply premium persists
UUP (US dollar)28.39+0.21%Firm, near 13-month highs — no dollar spike
HYG (HY credit)79.68+0.04%Spreads steady — no credit stress
GLD (gold)367.60−0.22%Haven soft — consistent with risk-on-at-the-top
TLT (20y+ USTs)83.89−0.75%Duration sold — yields up on the reflation lean
LensCopper and oil up, dollar firm but not spiking, gold and duration down, credit steady: the cross-asset board voted reflation / risk-on — growth-cyclical demand with rising yields and no haven bid. That macro vote is exactly what the copper long expressed and fired on; the equities just didn’t fully follow. All proxies confirmed (Massive).

05Sector & Breadth (realized)

XLEEngy+0.45
XLCComm+0.14
XLKTech+0.07
XLPStpl−0.39
XLFFins−0.39
XLRERE−0.42
XLUUtil−0.51
XLIIndu−0.72
XLYDisc−0.72
XLBMatl−0.99
XLVHlth−1.14
LensThis is where the two VOIDs came from: the morning bought the cyclical sectors (financials, industrials) off a real commodity-reflation signal, but the sector tape was red almost across the board — only megacap-tech and energy held green. The reflation bet belonged in copper, not XLF/XLI, and that is exactly the split the scorecard recorded.

06Single-Name Movers

This automated Nightcap run did not fetch the closing-bell newsletters (Stocktwits Daily Rip close / Axios Closer) — single-name narrative color is refresh-required. The confirmed close tape stands in as the illustrative read.

Illustrative only — never trade recommendations. Individual-name detail is refresh-required (no newsletter pull this run).

07Morning Lens vs Reality

Closing the loop on The Early Bird Curd (07-20) and The Midday Frappé (07-20). The morning read a risk-on-growth day and leaned into cyclicals; reality delivered the growth impulse where it belonged — in commodities — but not in the equity sectors the brief bought. Right: the reflation lean itself (copper long fired +0.78 ATR; oil firm; yields up), and the midday adjustment — when the intraday brief saw QQQ fail the 702.30 gate on 8/11-red breadth, the LRT short read the two-speed tape correctly and faded it clean to the 695.33 shelf. Wrong: the two cyclical-sector longs — financials and industrials VOIDed because the commodity bid never carried the equity sectors on a day the average stock fell. Mixed: the staples short was directionally right but sub-threshold. Net: the macro call (growth/reflation) was right and the midday pivot sharpened it; the equity-sector expression of that macro call was the miss — and the two open Sunday weekly setups sit untouched, neither armed nor killed, as designed.