Built 5:04 PM ET · 2026-06-22 · validates the 6/22 regular-session close (Massive /v3/snapshot, COALESCE(session_close,session_price))
anchor: system-reminder 2026-06-22 == Bash date (agree) · static — regenerate to refresh
01Session Scorecard — 2026-06-22
8 setups in the open backlog; 4 finalized this run (1 curd LRT short, 1 curd MS long, 1 midday LRT short, 1 midday SRB rotation long) — 3 FIRE / 1 MIXED. The other 4 are left interim (still open): the three 2026-06-21 Sundae weekly setups and the latent curd SBD reclaim all hinge on Thursday’s PCE, so their window has not closed — they carry forward for final scoring after the print.
| Setup | Outcome | Evidence | Δ-ATR | Actual |
|---|
| LRT S SPY,QQQmm-260622-LRT-SPY-S |
FIRE |
Rejection-fade FIRED on direction. SPY tagged 750.18 at the open (15c shy of the 750.33 trigger), rejected, and faded to close 744.39 (-0.32%), below VWAP 745.30; QQQ 737.95 (-0.36%) under its VWAP. No kill: only 6/11 sectors green and RSP (-0.17%) did not keep pace; PCE Thu pending. Caveat: the exact 750.33 entry never tagged — thesis right, live fill would not have armed. |
+0.53 |
744.39 |
| MS L NBIS,SMHmm-260622-MS-AIINFRA-L |
MIXED |
Split. NBIS gap-and-faded (O 289.96 → H 299.86 +4.6% intraday → close 283.61 -1.07%, below prior close) — the gap_and_fade_from_open kill fired, the same way it voided Thursday. But the broad-semis leg SMH held green +1.37% (668.91), so one exposed leg fired the long and the other voided. |
— |
NBIS 283.61 / SMH 668.91 |
| LRT S SPY,QQQmd-260622-1209-LRT-SPY-S |
FIRE |
Laggard-short FIRED on direction. The megacap-growth drag held all afternoon — GOOGL -4.97%, QQQ 737.95 never reclaimed its 740.16 VWAP (the stabilization kill); SPY drifted 744.91 midday → 744.39 close, pinned below VWAP. Caveat: the conditional bounce-into-746.77/750 entry never armed (tape stayed heavy) — no live fill. |
+0.52 |
744.39 |
| SRB L IWM,XLF,XLImd-260622-1209-SRB-ROT-L |
FIRE |
Rotation-long FIRED cleanly. IWM +0.88% (298.18) held above the 296 line (low 296.16); XLF +0.24 / XLI +0.49 / XLV +0.44 all green into the close; and RSP -0.17% outperformed SPY -0.32% — equal-weight beat cap-weight, the rotation signature. None of the three kills triggered. |
— |
298.18 |
Still open (interim — window not closed): sun-260621-LRT-SPY-S, sun-260621-MS-AIInfra-L, sun-260621-SBD-SPY-L (weekly Sundae setups) and mm-260622-SBD-SPY-L (latent reclaim). All arm/kill on Thursday’s PCE; carried forward, not scored. Day-1 read: SPY rejected the highs (constructive for the weekly LRT short); SMH held green (constructive for the semis long); the SBD reclaim did NOT arm (SPY 744.39 closed below the 745.34 trigger).
LensA high-dispersion rotation day, and both edges in it paid: the laggard/level-rejection shorts on the cap-weight indices AND the cyclical-rotation long both fired, while the AI-infra momentum chase split (SMH held, NBIS gap-faded a second time). Two clean directional reads — but the two LRT shorts each fired without ever tagging their entry (SPY high 750.18 vs 750.33/750.18 triggers). Right direction, unfilled mechanism: the recurring LRT tell, again.
02Calibration
0.284Rolling Briercumulative all-eligible · lower = better · 0.25 = random · last-50 window 0.271
41.5%Hit rateFIRE / calibration-eligible (22/53)
53Eligible sampleexcl. 2 NO_EVIDENCE / unevaluable
Pattern performance (Component 13) — calibration-eligible, by pattern × direction
| Pattern | n | Hit% | Mean prob | Mean Brier | Evidence class |
|---|
| LRT · S | 17 | 65% | 0.35 | 0.370 | lore_pending |
| MS · L | 13 | 31% | 0.49 | 0.263 | lore_pending |
| SRB · L | 6 | 67% | 0.39 | 0.315 | lore_pending |
| SBD · L | 5 | 0% | 0.39 | 0.153 | primary_source |
| SBD · S | 4 | 0% | 0.28 | 0.079 | primary_source |
| GFD · L | 2 | 50% | 0.45 | 0.205 | practitioner_backtest |
Singletons (GFU·S, LRB·L, MS·S, SRT·S, VAB·L, VSR·L) omitted for n<2.
Calibration trend (Component 14) — cumulative Brier by validation date
05-28.410
05-29.339
06-08.333
06-09.326
06-10.311
06-12.301
06-15.289
06-16.283
06-18.278
06-22.284
Mechanism leak · C15Both of today’s LRT shorts are in it (mm-260622-LRT-SPY-S, md-260622-1209-LRT-SPY-S): SPY topped at 750.18, below both the 750.33 and 750.18 entries — direction right (faded to a red close), but neither short ever armed/paid. LRT·S now n=17 at Brier 0.370, the worst of any multi-n pattern. Canonical C15 fraction deferred — dashboard build failed (see footer).
LensToday’s four scored a run-Brier of 0.356 and nudged the cumulative reading up 0.278 → 0.284, snapping a six-print improvement streak. That is not a bad-call signal — it is an under-confidence signal: three setups FIRED while priced at only 0.30–0.35. The LRT-short and SRB-rotation theses keep being better than their assigned probabilities (LRT·S 65% hit at 0.35 mean prob; SRB·L 67% at 0.39). The lever is two-fold: raise prior probability on these reads, and fix LRT entry location so the right-direction calls actually fill. MS·L (mean prob 0.49, 31% hit) remains the overconfident momentum-chase basket — NBIS gap-fading twice in three sessions is the tell.
03Tape & Rate Backdrop
HIGH-DISPERSION ROTATION
Realized regime (lookback, not a forward call): index-flat masked a wide rotation — small-caps + cyclicals up, megacap-growth down hard. SPY rejected 750.18 and closed mid-low range below VWAP. The morning CHOPPY read held; the midday rotation/laggard-short reframe nailed the afternoon.
| Index | Close | Day % | Range | VWAP |
|---|
| SPY | 744.39 | -0.32% | 743.13–750.18 | 745.30 |
| QQQ | 737.95 | -0.36% | 734.39–745.45 | 738.46 |
| IWM | 298.18 | +0.88% | 296.16–299.49 | 297.84 |
| RSP (equal-wt) | 209.61 | -0.17% | 209.10–210.34 | 209.76 |
- Driver: a sharp GOOGL -4.97% selloff anchored megacap-growth weakness (XLC -2.37%, XLY -1.89% the two worst sectors), while money rotated down the cap stack into small-caps (IWM +0.88%) and cyclicals (financials/industrials/healthcare/energy). Equal-weight beat cap-weight — a rotation, not a de-risking.
- NBIS popped on its Nasdaq-100 inclusion (effective today) then gap-and-faded from +4.6% to a red close — the inclusion-pop chase failed for the second time in three sessions.
- VIX not entitled. Sunday anchor ~18.4 — close refresh required
- Rates: TLT -0.76% (86.09) → long yields firmed modestly; 10Y est. ~4.54–4.56 from the midday 4.50. est. — FRED DGS10 daily close is T+1
LensThe index tape (SPY/QQQ -0.3%) was a decoy — the real action was the rotation underneath: IWM and the equal-weight outperformed while a single megacap (GOOGL) dragged the cap-weighted complex. A laggard-short / rotation-long pairing was the day’s edge, exactly the midday reframe.
04Cross-Asset
| Asset | Close | Day % | Read |
|---|
| UUP (USD) | 28.36 | +0.21% | Dollar firm — no easing bet, no fear bid |
| USO (WTI) | 112.69 | -1.90% | Crude soft — continued post-Iran-deal disinflation tailwind |
| GLD (gold) | 384.59 | -0.65% | Soft — no haven demand despite the growth selloff |
| TLT (20y+) | 86.09 | -0.76% | Bonds eased, long yields firmed — no flight-to-safety |
| HYG (HY credit) | 79.94 | -0.09% | Credit essentially flat — no systemic stress |
LensThe cross-asset tape confirms rotation, not de-risking: soft gold, firm dollar, and flat high-yield credit all say the megacap-growth selloff was a positioning rotation, not a risk-off event. Bonds slipping (yields up) is the small headwind that kept a lid on the broad bid.
05Sector & Breadth (realized)
XLEEngy+0.54
XLIIndu+0.49
XLVHlth+0.44
XLKTech+0.37
XLRERE+0.36
XLFFin+0.24
XLUUtil-0.09
XLBMatl-0.37
XLPStpl-1.34
XLYDisc-1.89
XLCComm-2.37
- 6 green / 5 red — below the 8+ green threshold the LRT-short kill required, so the short kill never armed (one of the two reasons that short FIRED).
- Leadership rotated OUT of megacap-growth (XLC -2.37%, XLY -1.89% the two worst) and INTO energy / industrials / healthcare / financials — a defensive-cyclical tilt, not a growth tape.
- RSP -0.17% beat SPY -0.32%: equal-weight over cap-weight directly confirmed the SRB rotation-long thesis from the SPDR close — no NO_EVIDENCE needed.
LensTextbook dispersion: the breadth-gated legs of both the LRT short (only 6 green) and the SRB long (equal-weight leading, cyclicals green) were measurable straight off the SPDR + RSP closes. The regime read and both trade theses were confirmable from primary data alone.
06Single-Name Movers (illustrative)
Confirmed from the Massive close (market_status snapshot). The Stocktwits Daily Rip close had not yet labeled at build time (5:04 PM ET) — catalyst color is limited to what primary data confirms; no catalyst is asserted without a source. Illustrative of the tape only — never trade plans.
- GOOGL -4.97% (close 349.68, low 341.72) — the day’s megacap anchor and the engine of the growth-sector drag (XLC/XLY). catalyst refresh-required — not asserted
- NBIS -1.07% (O 289.96 → H 299.86 → close 283.61) — Nasdaq-100 inclusion-pop gap-and-faded; the illustrative MS-long’s lead name voided the same way it did Thursday.
- SMH +1.37% (668.91, held its open) — broad semis followed through green even as the inclusion name faded; the AI-infra complex held while the single chase name did not.
- IWM +0.88% small-caps and the cyclical SPDRs (XLE/XLI/XLF/XLV green) were the relative winners — the rotation destination.
LensOne megacap (GOOGL) did most of the index damage while the broad market quietly rotated. “The tape isn’t weak — megacap-growth is.” The illustrative split (SMH held / NBIS faded) is the same chase-caveat lesson as Thursday: own the complex, not the single inclusion-pop.
07Morning & Midday Lens vs Reality
The Early Bird Curd called CHOPPY, resistance-capped at 745.34 / 750.33 with breadth failing to broaden — and the tape obliged: SPY topped at 750.18 (below 750.33), closed red below VWAP, breadth stayed narrow (6 green / 5 red). The midday Frappe then made the sharper call, reframing the day as a high-dispersion rotation and pairing a laggard/megacap-drag short with a small-cap/cyclical rotation long. Both fired.
Where the trade ideas diverged from execution: the two LRT shorts were directionally correct but never armed (price topped 15c shy of entry, then stayed heavy rather than offering a bounce to short). The MS AI-infra long split — the complex (SMH) held, the inclusion-pop (NBIS) faded. The honest scorecard: the reads were excellent (3 FIRE / 1 MIXED), the fills on the shorts were not.
LensThe day’s value was the midday rotation reframe and the cyclical-long, which paid cleanly. The carryover lesson is unchanged: the LRT short keeps being right on direction and wrong on trigger — an entry-location problem to solve pre-trade through the gate, not a regime miss.
08Execution Debrief — realized execution on closed setups, NOT a forward trade plan
| Setup | Profile | Entry | Stop | Target | Realized R |
|---|
| LRT·S mm-260622-LRT-SPY-S | swing (1d ATR) | 750.33 | 766.47 | 718.05 | +0.37R* |
| LRT·S md-260622-1209-LRT-SPY-S | day (5m ATR) | 750.18 | 751.31 | 748.49 | +5.12R* |
* Mechanical / notional, not booked. Both R-values are computed entry→close as if filled at the trigger. Neither short actually armed: SPY’s day high was 750.18, so the swing entry (750.33) was never tagged and the midday entry (750.18) printed only once, at the open — before the 12:09 call — never on the afternoon bounce the setup required. Read the R-column as the value of the directional thesis, not a realized win.
LensSame execution lesson as the 6/18 pin: the LRT shorts were “right but unfilled.” The harvestable fix is entry location / trigger tolerance on conditional resistance shorts, decided pre-trade through the gate — never chased intrabar. The day’s genuinely fillable edge was the zone-based SRB rotation long (no single trigger to miss).