Jul 14, 2026

Regime & Verdict#

• Regime: CHOP. SPY 10d +1.46% — just under the +1.5% UPTREND cutoff, consolidating off the 07-10 high 755.42. 07-13 was a real pullback day (SPY -0.77%, QQQ -1.90%). SPY 751.83.
• directional_tradable = TRUE. s1_standdown = FALSE (SPY ret5 only +0.55%, no qualifying ≥2% dip — crash guard OFF, MOM_SHORT/OI_FADE shorts stay live).
• Breadth divergence: pct_green only 40.4% on a green index day; 64.4% of tickers bearish-flow — narrow, mega-cap-led tape. IWM 10d -1.49%, near its own PULLBACK threshold.
• Sector rotation: INTO Financials/Energy/Real Estate, OUT OF Tech/Healthcare/Consumer Cyclical. Read into per-name regime_fit below (rotation-OUT sectors are tailwinds for shorts; rotation-IN sectors are headwinds for shorts / tailwinds for longs).
• Vol-state (2nd-moment only): VIX 16.50. SPY VRP FAIR (-1.7%), SPY dealer GEX NEGATIVE/short-gamma (spot 752 below the 759.96 flip) — vol-amplifying. QQQ VRP PREMIUM_BUYING (-6.6%, cheap vs realized), QQQ GEX positive but pinned at the flip — fragile.
• Event calendar (h10 window 07-14 → ~07-28): PPI 07-15 (Tier-1, tomorrow) stacks directly on the CTAS/ASML/JNJ/CAG earnings-IV-crush cluster. FOMC 07-28 lands essentially at the window edge; Core PCE 07-30 just past it.
• Bottom line: 3 open carry-forward positions (unchanged), 1 new sized call (LNG, starter), and a wide watch list. LNG (OI_FADE short) is the only fresh name that clears every gate at a real size tonight. Nothing sizes above starter — half-cap-no-auto-full, and CHOP is not a clean enough regime to push any of tonight's CAUTION/marginal names past that anyway.

Directional Book (excess-scored)#

ORCL hold-note (explicit, as requested): the position has realized +3.10% excess in one trading day against a lane mean of +0.81% over the full h10 hold — roughly 3.8× the lane's entire expected edge, already. That +0.81% is a full-holding-period average across the sample, not a per-day accrual rate; a name front-loading several multiples of the mean on day 1 doesn't thereby "owe" another +0.81% through day 10, but it also doesn't mean the trade is over — we simply don't have the conditional distribution (excess-at-h10 | excess-at-h1 > 3%) to size that quantitatively tonight. The framework has no profit-taking or trim rule, and none is being invented here. The correct action, and the one taken: hold at unchanged starter size, do not upsize on the strength of the move (would violate the hard never-upsize rule), do not invent a discretionary trim. The stated invalidation (RSI reclaims >45 + range reclaim, or the downgrade narrative reverses) is the only defined exit trigger, and it has not fired. The size of the move does raise give-back/mean-reversion risk over the remaining window — flagged for closer next-run monitoring of the invalidation trigger, not a size change.

Every deviation from a lane's implied ceiling tonight was a downgrade (LNG half→starter on CAUTION-1; WSC, TDG, XENE, FMC, XE, OLED, BSX, CPRT, all 8 S2 names, and DKS all downgraded to watch), consistent with half-cap-no-auto-full and the hard "never up-size" rule. No name was upsized on the strength of any move, catalyst, or fundamentals CONFIRM.

Watch / Stood-down (gate-by-gate detail)#

OI_FADE cohort#

• WSC ($26.84, Industrials) — signal live (+7,500 07-13, +1,198 07-14, ratio 1.225) and fundamentals CONFIRM the short (4/4 EPS misses, rev -4.12% YoY, unprofitable). Nominal CAUTION/CONFIRM-adjusted tier is starter, but downgraded to watch on a stack of independent tail risks the fundamentals confirmation does not offset: (a) 20d $-ADV $53.2M, only ~6% above the $50M fail-closed floor — thin practical exit liquidity; (b) earnings 07-30 is only +2 trading days past the h10 window close — thinnest calendar margin tonight; (c) short float 10.73%, 7.6 days to cover = MODERATE-HIGH squeeze fuel, exactly the momentum-crash/squeeze tail this gate stack exists to cap; (d) extreme call:put skew (1,443:12) today is ambiguous — could be thesis-consistent hedging or the leading edge of a squeeze, cannot disambiguate. Fundamentals confirmation is a veto channel, never a promotion mechanism (research/80) — it does not override this risk stack. Also flagged: raw debt/equity 4.19× — the "low leverage" classifier label is miscalibrated (data-quality note, not sized on).
• TDG — fundamentals VETO, 2nd consecutive scan (3/3 beats, Zacks+SeekingAlpha upgrades 07-06, raised FY guidance). Watch-only.
• XENE — fundamentals CONFIRM, but the orchestrator's signal-quality demotion stands: 90% of the 5-day OI build (10,030/11,179) is a single 07-09 print — the same artifact family the lane excluded BHVN for. The validated mechanism is persistent multi-day accretion, not a one-day print. A CONFIRM cannot promote it back. Watch-only.
• Data-quality flag (not sized, no action tonight): tonight's raw OI_FADE ranking was dominated by single-day-print artifacts (FIG 95% one-day, AGEN +82%/-17% round-trip on binary news, BHVN 99% one stale print) plus ETP-mislabel (DXYZ, wearing a "Common Stock" issue_type — the JPST trap again) and a too-new listing (CBRS, IPO 05-14-2026, ~43 trading days old). The pinned floors did not catch the single-day-print family. This is a proposed rule-set gap, not a change shipped tonight — any lane/threshold fix must clear retro_harness.py --all first.

MOM_SHORT cohort — sized nothing tonight#

• BSX — fundamentals VETO. Earnings 07-29 (+1 trading day past window close), 4/4 beats, rev +17.44%, EPS +74.72%, net margin 17.29%, price only 0.90% off its 52w low — fundamentals flatly fight the short.
• CPRT — fundamentals VETO, reaffirmed and strengthened (fresh 06-30 beat +3.79%, net margin 33.48%, incoming-CEO catalyst).
• FMC, XE, OLED — all CAUTION -1. Independently confirming the orchestrator's read, and not merely deferring to it: MOM_SHORT's own +0.81% lane mean maps to starter, not HIGH, under tonight's explicit sizing formula (+0.3–1% band → starter). A CAUTION -1 on a starter-tier name mechanically lands at watch — every survivor falls out this way without any additional discretionary cut needed. On top of that mechanical result: FMC carries HIGH squeeze risk (21.06% short float, 6.96 days to cover); OLED carries accelerating insider-buying-to-ceiling (+79.7→+94.5→+100 MSPR) plus MODERATE-HIGH squeeze (11.12%/5.49 days); XE has an independent, sufficient, fail-closed blocker on its own — next_earnings_date is confirmed unverifiable (NaT) in tonight's screener export, and an unverified earnings date inside a 10-day horizon is a hard blocker regardless of the tier math. Every sizable MOM_SHORT survivor tonight is simultaneously an already-crowded short (near-52w-low names are heavily shorted by construction) and event-fragile (every earnings date sits 1–2 trading days past the window edge) — that combination is precisely where the momentum-crash/squeeze tail lives. Verdict: size nothing from MOM_SHORT tonight.

MOM_LONG basket (184 names, +2.35% mean but median ≈ 0, tail-driven — never per-name HIGH)#

• Sector concentration: Financials 29, Real Estate 27, Healthcare 26, Utilities 20, Industrials 20, Tech 16, Consumer Cyclical 15.
• Top-proximity names (watch/basket-only): NUVL, QLYS, GTLS, CRNX, CM — cluster_id: mom_long_basket_20260714.

S2 liquidity-reversion (+0.47%, h3-5 — IC ≈ 0, no rankable "best pick")#

• 8 names passed the lane's own filters: PJT, OVV, BJ, HGV, SSB, HLI, THO, NNN. All clear the liquidity floor. No fundamentals-gate verdict was obtained for any of them tonight — flagged back to the /market-scan orchestrator for a fundamentals-gate spawn before any can be sized, per the standing 07-13 MOM_SHORT-cohort precedent ("cannot size unverified"). Held at watch, confirming the Phase-B lane's own recommendation. Sector clusters noted for the next fundamentals-gate pass: Financials (PJT/SSB/HLI), Consumer (BJ/HGV/THO, Consumer is a rotation-OUT sector — headwind for the long), Energy+REIT (OVV/NNN, both rotation-IN — tailwind).
• Excluded upstream (Phase B): TRNO (informed catalyst, 07-14 lease renewal), APLE (catalyst unverifiable, fail-closed), DOCS/SRRK (short-side, lane is long-only).

S4 sentiment-contrarian (+0.71%, h5-10 — advisory-size-only by lane design)#

• DKS ($211.37, PCR 8.29) — fundamentals CONFIRM (3/4 beats, rev +41.22% YoY, live catalysts: bankrupt rival's closure 07-12, ScoreCard+ loyalty launch 07-14), squeeze-aligned float (11.32%/6.13 days), earnings 08-24 safely outside horizon. This is the cleanest setup measured tonight across any lane. The Phase-B lane flagged it "HIGH-CONVICTION" — explicitly overridden: S4 is advisory-size-only by lane design regardless of setup cleanliness, a structural ceiling below starter, not a gate-driven cut. Sizing it above watch would be an upgrade above the lane's validated-excess ceiling, which the hard rule forbids. Held at watch/advisory. Consumer Cyclical rotation-OUT is also a live sector headwind against the long.
• ivrank_chg_5d top decile (314 names) — advisory observation only, not scored, not sized.

Vol Book (non-directional, delta-neutral, net-of-cost, advisory — 0 directional points)#

• Earnings IV-crush SELL-VOL (all confirmed backwardation; PPI 07-15 stacks a Tier-1 macro binary on top of the same-day names CTAS/ASML/JNJ/CAG):
  • NFLX (ER 07-16) — primary, cleanest richness: IV rank 97.3, implied move 6.79%, front/back 2.39×. No same-day macro stack.
  • CTAS (ER 07-15, IV rank 100, implied move 4.20%, 1.97×), ASML (ER 07-15, implied 6.10%) — secondary, same-day PPI stack knocks size down.
  • TSM (ER 07-16, implied 4.43%, liquid) — small-to-moderate.
  • ABT (ER 07-16, implied 4.25%) — half-size at most.
  • JNJ (ER 07-15, implied 2.84%) — smallest implied move of the cluster, same-day PPI stack; small size only, likely marginal net-of-cost.
  • CAG (ER 07-15, implied 5.40%) — thin options liquidity historically despite adequate equity ADV; small size only.
  • Net-of-cost expectancy not independently quantified per-name tonight — advisory/defined-risk/wide-wings only, nothing sized as a real position.
• 0DTE-VRP: HARD STAND-ASIDE both SPY and QQQ. zerodte_setup.py returned GO_PREMIUM_SELL for both; overridden on the live read. SPY: VRP FAIR (not rich) + negative gamma = no premium to harvest and amplified tail risk into PPI tomorrow. QQQ: VRP PREMIUM_BUYING (vol cheap vs realized) + gamma pinned fragile at the flip = selling fights the lane's own measured edge. No premium-selling either index this week.

Risk#

• Correlation clusters: none among tonight's sized/near-sized names (LNG Energy, WSC Industrials, MAT Consumer Cyclical, ICE Financials, ORCL Tech — 5 distinct sectors, no assumed corr ≥ 0.70). MOM_LONG basket collapsed to sector-cluster reference only (mom_long_basket_20260714), never decomposed to per-name sizing. S2 basket noted in three sector sub-clusters (s2_financials_20260714, s2_consumer_20260714, s2_energy_reit_20260714) for the next fundamentals-gate pass — not sized tonight so no collapse enforced yet.
• Event calendar: PPI 07-15 (T+1, Tier-1) stacks on the CTAS/ASML/JNJ/CAG vol-book names same-day; treated as defined-risk-through-print on the idiosyncratic directional book, not an automatic -1, consistent with prior-run precedent. FOMC 07-28 sits essentially at the h10 window edge opened tonight — awareness only. WSC's earnings (07-30, +2 trading days past window close) was a material factor in its watch-only downgrade; FMC/BSX (07-29, +1 day past) and OLED (07-30, +2 days past) similarly.
• Tail caps applied: OI_FADE (MAT/ICE carried, LNG new) sized for the momentum-crash/squeeze tail, capped at starter, none above starter; WSC specifically failed this test despite a CONFIRM fundamentals read (squeeze float + marginal liquidity + thin earnings margin). MOM_SHORT sized nothing tonight — every survivor combines crowded-short-by-construction with event-fragility, the textbook squeeze/crash tail. MOM_LONG kept strictly basket-only. S2 cut on the missing-fundamentals gate before its own news/catalyst tail cap was even reached. Vol-short (0DTE) stood aside on both indices, sized for the unsampled left tail under negative/fragile gamma.
• Hedge note: SPY dealer GEX is negative/short-gamma into a Tier-1 macro print (PPI) tomorrow — the vol-amplifying backdrop argues for the half-cap discipline to bind hard on the whole book, not just the vol lane. QQQ's gamma is pinned right at its flip — a small move either way could flip the whole complex's 2nd-moment character intraday.

---

Envelope: decision.json (v2, validated against schemas/decision_envelope.v2.schema.json — PASS). Conviction persisted: conviction_2026-07-14.json (MAT + ICE + ORCL carry-forward, unchanged) for next-run adverse-flow exit checks — LNG is a fresh starter call tonight and will join the conviction file once confirmed live on the next scan, consistent with how ICE/ORCL joined on 07-13→07-14. watchlist_write_back_confirmation: 22 names written back to decision.json's watchlist_write_back array (BJ, BSX, CM, CPRT, CRNX, DKS, FMC, GTLS, HGV, HLI, NNN, NUVL, OLED, OVV, PJT, QLYS, SSB, TDG, THO, WSC, XE, XENE) — CONFIRMED. No lane/threshold change this scan → regression gate not triggered.