Daily Actions
Daily Actions
MARKET CONTEXT: SPY $770.19 (RSI 55.6, strong-up, -1.18% off its high, 3M +4.43%) | QQQ $718.96 (RSI 53.6, 3M +1.97% — lagging the equal-weight RSP's +5.37% 3M, a breadth story not a mega-cap one) | VIXY $17.00 (RSI 31.5, -50.67% 1Y, -57.2% off its 52-week high — no hedging bid) | GLD $406.77 (RSI 52.4, -2.1% vs its 200-day, -20.19% off its 52-week high — crossed below its 200-day for the first time this stretch)
Price truth: 2026-09-04 Friday settled close, read from all 31 of today's scans plus the 11 candidate roundups and the code-diffed regime-transitions feed (regimes:feed --json, 2026-09-04 snapshot vs 2026-09-03). Every price/RSI figure below was spot-checked directly against its source research/market-engine/data/summaries/*.json file this session (defense-contractors.json, travel-leisure.json, retail.json, defensive.json, biotech-scan.json, and 30+ others) — no ticker in this digest failed that check. research/issues/2026-09-04-holdings-snapshot.md was read ONLY to populate Exit Alerts (position management on real accounts); it was not used to shape any Entry Zone, Watch, or Avoid call above — book-blind discipline held throughout. The most recent prior digest is 2026-08-28 (7-day gap); carry-forward below is against that file.
Summary
2026-09-04 was a rotation inside technology, not a risk-off day — and the rotation inverts what the price action alone would suggest. Hardware ripped: the semis complex's median one-day gain was +3.19% (etf-universe.json, SOXL +9.87%, SMH +2.61%), memory/NAND names popped 5.9-8.7% (MU +6.1%, STX +6.34%, WDC +5.86%, SIMO +8.7% — memory.json), and the optical/CPO board's top movers ran 9.2-13.1% (AEHR +13.1%, SNDK +11.9%, PSIX +11.2%, COHU +10.3%, ALAB +9.8%, AXTI +9.7%, SMTC +9.2% — validated directly against monster-discoveries.json, focus.json, ai-power.json, spacex-s1-supply-chain.json, thebuildout-co-packaged-optics-cpo.json). Software got liquidated the same session: 15 of 16 cloud/SaaS names closed red (cloud-saas.json, via etf-universe.md), with LULU -17.4% (confirmed Q2 miss and guidance cut), PATH -16.6% (earnings beat, sold off anyway), ASAN -12.7% (beat-but-sell-off, event ledger still unconfirmed), ADBE -6.7%, and SNOW -5.4% (ai-scan.json, insider-universe.json).
The structure underneath inverts the day's price action, and this is the load-bearing read for anyone tempted to chase the chip pop. The 18-name semis cohort that ripped today carries a median drawdown of -26.9% off 52-week highs with only 4 of 18 names in a genuine uptrend, and the pop came on ordinary volume (etf-universe.json, computed directly from semis.json) — a violent one-day bounce inside a still-damaged cohort, not a confirmed trend change. The cloud/SaaS names that sold off, by contrast, carry 14 of 16 names still classified up-trend, and did so on the lightest relative volume of the cycle (median 0.67x average) — a shallow, low-conviction pullback sitting on top of a mostly-intact structure. Per the desk's entry discipline (bottom → pop → flag → breakout, no pre-confirmation buys): the semis pop is a WATCH, not a buy — it needs a flag/consolidation and then a breakout on above-average volume before it counts as a confirmed leg. Within chips specifically, the layer matters: every name in the semicap-equipment/test book (wfe-test-metrology) carries a positive 1-year return and positive alpha even where price sits below the 200-day (trend-under-test, not broken), while the chip designers are the ones genuinely broken on the year (ARM -26.5% 3M, ON -36.6% 3M, AMKR -26.3% 3M per nvda-ecosystem).
A live US-Iran escalation is repricing rates, not fear — and it explains both the software liquidation and the defense sector's total failure to catch a war bid. Oil ripped (USO +9.59% 7D, +91.37% 1Y, macro-commodities.json) on confirmed US strikes and Iranian retaliation in the Strait of Hormuz (MarketWatch/CNBC/Bloomberg, Sept 1-4, cited across airlines, chemicals, travel-leisure, geopolitical-risk), and global bond yields spiked on Fed rate-hike bets tied to that oil-driven inflation fear (CNBC, Sept 1: "global bond yields soaring to multi-decade highs"). That is why long-duration software de-rated on genuinely good earnings this cycle — PANW beat and fell -10.36% on the week, ZS beat and fell -7.99%, CRDO beat and broke below its 200-day on a margin/guidance scare, MDB beat and dropped -17.44% on the week, CIEN beat (a confirmed 22% EPS surprise) and sold off on real volume — a rates story, not a fundamentals one (cybersec, tech-insider-buys, optical-supply-chain). Defense is the starkest confirmation that this is a rates trade, not a fear trade: the entire prime-contractor book sold off 2-9% this week during the live exchange of fire — LMT -6.85%, RTX -5.20%, NOC -5.55%, GD -5.27%, LHX -2.41% — the opposite of a naive war-premium read (defense-contractors, geopolitical-risk). Gold confirms the same story from the other side: rather than catching a haven bid, GLD crossed below its 200-day for the first time this stretch (-2.1%, was -1.4%) and sits 20.19% off its 52-week high, while VIXY stays crushed (RSI 31.5, -57.2% off its high) — neither confirms a fear trade (macro-commodities, market-pulse).
Credit is the one place real deterioration is visible underneath a calm equity tape. HYG's RSI collapsed 52.7→38.9 with a fresh death cross — the sharpest single move on the entire market-pulse board — dragging TLT, LQD, BND, and IEF all further below their 200-days and further negative on the year, while SPY sits only -1.18% off its own 52-week high. That divergence (credit cracking, equities calm) is worth tracking into next cycle.
Oil producers/services and the food-security book are the day's genuinely clean strength — every upstream and services oil name flipped from a negative to a solidly positive week (XOM, CVX, COP, DVN, EOG, SLB, HAL all strong-up, geopolitical-risk), and 9 of 10 food-security names gained RSI with NTR/DE/MOO all in confirmed breakout (food-security) — though FMC and MOS's double-digit RSI pops there sit on top of genuinely broken trailing-year charts (FMC -65% 1Y, alpha -83.87; MOS -21% 1Y, alpha -39.41), a bounce inside a broken trend rather than a resumption.
Entry Zones (IN ZONE NOW)
EXTREME OVERSOLD (RSI < 20) - Capitulation Territory
None. No name across today's 31-scan corpus closes below RSI 20.
OVERSOLD (RSI 20-30) - Strong Entry Zone
| Stock | Current | Zone | RSI | 30D Chg | Source | Action | Notes |
|---|---|---|---|---|---|---|---|
| TJX | $132.08 | Watch $154.72 (200-day) | 22.3↓ | -17.41% | retail-scan, retail candidate | 🔍 Watch — deteriorating | 2nd straight scan as the board's lowest print; a Jefferies downgrade (Aug 28) is now on file, and 1-year change has turned negative for the first time (-6.34%, alpha -24.99) |
| RARE | $15.30 | No support shelf identified | 21.1↓ | — | biotech-scan | ⚠️ Verify, don't chase | Just crashed into extreme oversold on 3.2x average volume; -51.1% 1Y; news pull (fetched 2026-09-04, 0 items) and stock-events/ both checked — no cause established |
| RCL | $265.19 | Below 200-day | 25.6↓ | -19.01% | travel-leisure | ❌ Avoid | Now broken on every window, not just the short term — -26.35% 1Y, worst alpha in the sector |
| TSN | $51.42 | No support shelf identified | 26.4↓ | — | defensive-scan | 🔍 Watch | Cut's new deepest print, no confirmed floor; -8.5% 1Y keeps this a watch, not a buy |
| PL | $18.12 | none | 28.6↓ | -18.93% | drone-defense | 🔍 Watch — verify | Beat and raised guidance Sept 3, sold off on 5x average volume the next day — the print and the tape disagree, unresolved either direction |
| HRL | $21.54 | No support shelf identified | 28.9↓ | — | defensive-scan | 🔍 Watch | Still deep despite a partial bounce; -14.7% 1Y and no confirmed floor |
| NIO | $3.80 | none | 24.5↓ | -18.28% | ev-clean-energy | 🔍 Watch | Confirmed Q2 earnings beat (EPS +107% surprise, stock-events/2026-09-01-nio-earnings.md) coincided with a slide to a 52-week low; -38.01% 1Y, 27.1% below its 200-day — the print/price disconnect is the story |
APPROACHING OVERSOLD (RSI 30-40) - Watch Closely
| Stock | Current | Zone | RSI | 30D Chg | Source | Action | Notes |
|---|---|---|---|---|---|---|---|
| LULU | $100.61 | none | 31.8↓ | -18.54% | cultural-thesis, bargain-bin, consumer-scan, retail-scan, lazarus-scan (5x) | ❌ Avoid | Confirmed Sept 3 Q2 miss (revenue -4%, guidance cut) on 8.69x average volume; -51.18% 1Y, -34.1% vs SMA200, collapse regime — no long-horizon counter-argument anywhere in this corpus |
| GD | $359.39 | Watch $354.53 (200-day) | 31.0 | -6.43% | defense-contractors, geopolitical-risk (2x) | 🔍 Watch — trend testing average | Softest RSI in the defense book; nearly flipped below its 200-day (+1.4%, was +7.1%) while staying +11.46% 1Y — deep pullback in an intact multi-year trend, not a breakdown, but close |
| LHX | $256.45 | No support shelf on file | 31.7 | -10.37% | defense-contractors, geopolitical-risk, drone-defense (3x) | ❌ Avoid | One of three primes genuinely broken on both windows: -6.06% 1Y (worse than -5.13% last cycle), -18.4% vs SMA200 |
| CCL | $23.51 | Confirmed death cross | 31.4↓ | -20.76% | travel-leisure, war-sensitivity (2x) | ❌ Avoid | -25.98% 1Y; completes the cruise-line breakdown trio with RCL/NCLH — alpha worse than -44 across all three |
| PINS | $20.40 | Closest to oversold in this universe | 31.9 | — | insider-scan | ❌ Avoid | -45.04% 1Y, -7% below its 200-day, downtrend — genuinely broken structure, not a discount; not in top-20 insider selling either |
| DFEN | $61.45 | Support shelf $60.10 | 32.7 | -29.57% | defense-contractors | ⚠️ Avoid — leveraged | 3x-leveraged defense basket; the 30-day print is leverage decay compounding a real sector pullback, structurally the wrong vehicle regardless of the sector call |
| PPA | $165.18 | Watch $170.51 (200-day) | 33.0 | -9.97% | defense-contractors | 🔍 Watch | Flipped below its 200-day this scan; +12.18% 1Y stays positive |
| ITA | $225.61 | Watch $229.45 (200-day) | 33.4 | -10.57% | defense-contractors, market-pulse (2x) | 🔍 Watch | Deepest oversold reading in market-pulse's new industry tier; testing its 200-day from just below (-1.7%) with a still-positive year (+13.65%) |
| LMT | $525.28 | Watch $516.86 (support shelf), $556.86 (200-day) | 33.5 | -9.06% | defense-contractors, geopolitical-risk (2x) | 🔍 Watch — trend testing average | Flipped from +1.6% above its 200-day to -5.7% below it this scan while staying +14.98% 1Y — deep pullback in an intact multi-year trend, the level to watch is close ($8 above support) |
| LASR | $40.06 | none | 33.6 | -46.76% | drone-defense | 🔍 Research | Sharp collapse tag, but thesis rated intact by the desk and still +35.66% 1Y with positive alpha — horizon math argues pullback over breakdown, though the magnitude of the drawdown warrants direct confirmation |
| MAR | $336.51 | none | 33.6 | -6.86% | travel-leisure | 🔒 Hold | Mild pullback in a strong trend, +25.17% 1Y — the sector's most balanced name |
| CIEN | $321.00 | none | 33.9 | -21.48% | optical-supply-chain | 🔍 Research | A real earnings beat (22% EPS surprise, stock-events/2026-09-03-cien-earnings.md) sold off on real volume; thesis still rated intact-but-priced — the print/price disconnect is the thing to watch |
| NOC | $514.98 | No support shelf on file | 34.0 | -7.62% | defense-contractors, geopolitical-risk (2x) | ❌ Avoid | Negative on both 1Y (-11.17%, worse than -7.20% last cycle) and vs SMA200 (-14.6%) — trend genuinely broken, not just a weak week |
| RTX | $200.79 | Watch $193.09 (200-day) | 34.8 | -9.68% | defense-contractors, geopolitical-risk (2x) | 🔍 Watch | Stays 4.0% above its 200-day with +26.54% 1Y — deep pullback, intact trend |
| CRDO | $170.57 | Verify the trend before treating as a discount | 34.2 | — | tech-insider-buys | ⚠️ Verify | Broke below its 200-day this scan on a confirmed-beat, margin/guidance-scare selloff; regime flipped pullback→collapse per the code-diffed transitions feed; +27.29% 1Y stays positive |
| TDY | $610.65 | none | 34.2 | -10.69% | supply-chain-traces | 🔍 Watch | Lowest reading in the book; alpha1y has turned negative behind a still-positive headline 1-year return — worth watching whether this deepens |
| SYK | $303.13 | No support shelf identified | 34.1 | — | healthcare-scan | ❌ Avoid / breakdown watch | Just crashed into the oversold band with worsening long-trend numbers (-23.1% 1Y) — no confirmed floor |
| AAL | $13.13 | Below 200-day | 36.2 | -20.81% | airlines, war-sensitivity (2x) | ⚠️ Avoid | Near-flat year, second-worst alpha in the sector |
| DD | $131.59 | none | 36.2 | -10.71% | chemicals | 🔒 Hold | Sector's best 1-year performer and best alpha, only mildly below its 200-day despite a weak short-trend tag |
| CROX | $117.53 | none | 37.9 | -14.41% | cultural-thesis, consumer-scan (2x) | 🔒 Hold | One of two names (with SBUX) positive on every horizon in this book — above its 200-day, +32.03% 1Y, alpha +13.38 |
| AVGO | $357.90 | 200-day test resolved down | 38.1 | -14.44% | ai-scan, bargain-bin, ai-infra candidate (3x) | ⚠️ Avoid | Newly oversold; shrugged off a strong Sept 2 earnings beat (AI semi revenue +221% YoY) and broke below its 200-day (-3.0%), resolving down the level the desk had watched unresolved for two straight cycles |
| LIN | $477.57 | none | 38.5 | -2.75% | chemicals | 🔍 Watch | Fractionally below its 200-day, roughly flat year, negative alpha |
| DDOG | $212.93 | none | 38.7 | -24.80% | ai-scan, cloud-saas candidate, web-cloud candidate (3x) | 🔍 Watch | 200-day cushion nearly halved again; still +61.58% 1Y with alpha +42.93 — the deepest short-window pullback in the day's software liquidation, but still inside an intact long-horizon trend |
| HYG | $79.16 | none | 38.9 | -0.45% | market-pulse | ⚠️ Avoid | Sharpest reversal on the whole board — flipped from the bond complex's least-damaged name to fully death-crossed, -2.07% 1Y |
| IEF | $92.25 | none | 38.9 | -1.14% | market-pulse | ⚠️ Avoid | Deepened alongside the rest of the bond complex, -4.22% 1Y |
| XLI | $175.27 | none | 38.9 | -5.95% | market-pulse | 🔍 Watch | Second straight cycle in the oversold band; 200-day cushion thinning (4.1%→2.6%) even as +15.70% 1Y stays positive |
| AMAT | $454.71 | none | 39.8 | -14.80% | wfe-test-metrology | 🔍 Research | Lowest reading among the group's majors; next confirmed catalyst is the scheduled November 12 earnings print, two months out |
| COGT | $34.76 | Watch $36.70 (200-day) | 38.8 | — | biotech-scan | 🔍 Watch | Still testing its 200-day from below after a huge run; +168.6% 1Y |
| IYT | $84.17 | none | 39.9 | -3.00% | market-pulse | 🔍 Watch | Fell into the oversold band but passes the long-horizon check — +19.36% 1Y, +4.4% above its 200-day |
| BB | $7.70 | none | 39.3 | -11.80% | cybersec | 🔍 Watch | Dropped back into oversold with no catalyst on file (news stale, fetched Aug 27); still up +102.63% 1Y |
| JETS | $28.80 | Right at the oversold edge | 40.0 | -14.11% | market-pulse, airlines, travel-leisure (3x) | 🔍 Watch | Thin margin above its 200-day (+1.5%) but still above the line, +10.90% 1Y |
| LUV | $39.62 | Watch $43.07 (200-day) | 38.0 | -18.79% | airlines | 🔍 Watch — for a reclaim | Strong 1Y/alpha, but now below its 200-day |
(Volatility products VIXY $17.00 RSI 31.5 and VXX $17.72 RSI 32.4 both sit in this RSI band on market-pulse.json but are excluded from the table above — both are crushed/collapse-regime hedging products, not entries; buying them "oversold" fights the desk's own no-hedging-bid read. Dozens more names cluster near this band inside the day's chip-designer weakness (ARM, ON, AMKR all deeply negative on 3M per nvda-ecosystem) and the AI Power/Energy cohort's continued repair — the table above is the cross-confirmed, highest-signal subset, not the full list.)
🟠 STRUCTURE INVERTS THE PRICE ACTION — the day's key entry-discipline read
| Cohort | What Happened | Structure Underneath | Action |
|---|---|---|---|
| Semis complex | Median 1D +3.19%, SOXL +9.87%, SMH +2.61% (etf-universe.json, semis.json) |
Only 4 of 18 names in a genuine uptrend, median -26.9% off 52-week highs, pop on ordinary volume (~1.04x relative) | 🔍 WATCH — needs a flag/consolidation, then a breakout on above-average volume before this counts as a confirmed entry per the desk's bottom→pop→flag→breakout discipline |
| Memory/NAND top movers | AEHR +13.1%, SNDK +11.9%, PSIX +11.2%, COHU +10.3%, ALAB +9.8%, AXTI +9.7%, SMTC +9.2% (validated against monster-discoveries.json, focus.json, ai-power.json, spacex-s1-supply-chain.json, thebuildout-co-packaged-optics-cpo.json) |
All are RSI 46-61 (neutral-to-strong, not oversold) — the extended tip of the still-damaged semis complex | 🔍 WATCH — extended, not an entry-zone name by RSI; the desk's own long-horizon read on this pair (SNDK, MU) already calls it "riding a still-damaged complex, not proof it turned" (monster-scan) |
| Cloud/SaaS complex | 15 of 16 names red the same session, on the lightest relative volume of the cycle (median 0.67x) | 14 of 16 names still carry an up-classified trend label — a shallow pullback on light participation, not distribution (etf-universe.json, cloud-saas.json) |
🔍 WATCH — this is the higher-quality setup of the two if the pullback stays this shallow: watch for it to flag rather than break |
| Chip designers vs. semicap/test | ARM -26.5% 3M, ON -36.6% 3M, AMKR -26.3% 3M (nvda-ecosystem) vs. every semicap-equipment/test name (CAMT, LRCX, ENTG, ASML, KLAC, TER, FORM, ONTO, KLIC) carrying positive 1Y and positive alpha even below the 200-day |
The layer decides: designers are genuinely broken, equipment/test is trend-under-test | ⚠️ AVOID the designers, WATCH the equipment layer for a real-volume follow-through |
Price Alerts (Waiting for Entry)
| Stock | Alert | Entry Zone | Source | Set Date | Status |
|---|---|---|---|---|---|
| META | Pullback entry | $600-630 | ai-scan | 07-11 | 🟡 Price reached the zone ($616.77) but RSI is now 62.4 (neutral-strong, not oversold) — a hyperscaler recovery, not a confirmed dip-buy; -17.62% 1Y stays the worst hyperscaler alpha on the board (ai-scan.json) |
| CEG | Golden-cross flip | — | ai-scan, nvda-ecosystem, energy candidate | 07-11 | ⏳ Still waiting on the golden cross, but genuinely improving — price crossed back above its 200-day this cycle (+0.9%, was -7.1%), RSI 54.4→66.7, 1Y improved to -3.27% (was -13.39%) |
| T | Golden-cross flip | — | defensive-scan, ai-scan | 07-11 | 🔍 Partial — RSI 60.0, +2.1% above its 200-day, 1Y still negative (-13.18%) |
| GLD / URA | Watch for a golden-cross flip | — | macro-commodities, etf-universe, geopolitical-risk, market-pulse | 07-25 | 🔴 GLD deepened further below its 200-day (-2.1%, was -1.4%) — a second straight cycle below the line, now -20.19% off its 52-week high. URA still waiting (-5.0% vs SMA200, RSI 55.0), gap narrowing slightly from -6.0% |
| LULU | Watch a regime upgrade | — | cultural-thesis, consumer-scan | 07-28 | 🔴 Resolved the other way — instead of upgrading, LULU crashed on a confirmed Q2 miss (RSI 31.8, collapse regime, -51.18% 1Y); moved into Entry Zones above as an Avoid |
| HUBS | Watch whether the TEAM divergence resolves | — | tech-insider-buys, washout-scan | 08-08 | 🟡 Reversed back to -5% below its 200-day (from nearly crossing last cycle) — second straight cycle of approach-then-retreat |
| AVGO | Watch its 200-day test resolve | $352.12 shelf | ai-scan, bargain-bin, nvda-ecosystem, tech-insider-buys, insider-scan | 08-22 | 🔴 Resolved DOWN this cycle — vs SMA200 -3.0%, market shrugged off a real Sept 2 beat; moved into Entry Zones above |
| SMCI | Watch whether the 1Y ever turns positive | — | tech-insider-buys | 08-28 | 🟡 Nearly there — 1Y deficit narrowed to -2.73% from -15.67%, RSI 62.7; a continuation could flip it positive for the first time |
| TEAM | Watch whether the extension finally cools | — | tech-insider-buys | 08-28 | 🟡 Cooled somewhat — RSI 70 (was 79.1), still +73.5% above its 200-day per monster-scan's Overbought Warning list; still extended, not chasing |
| ISRG | Thesis flag vs. chart divergence | — | insider-scan, biotech-scan, healthcare-scan | 08-28 | 🔴 Persists, mixed this cycle — vs SMA200 widened to -20.9% (was -20.5%) while 1Y improved slightly to -19.32% (was -21.37%); thesisState: intact against a collapse regime, unresolved for a third-plus cycle |
Exit Alerts (Holdings)
Book-blind boundary: this section is position management on real accounts, read from research/issues/2026-09-04-holdings-snapshot.md (itself sourced from holdings list/positions CLI output cross-checked against holdings-stocks.json). It carries no entry-zone implications and was not used to shape any call above.
| Account | Position | Current | RSI | Your Gain | First Buy | Status |
|---|---|---|---|---|---|---|
| rh-trade | AAPL | $319.97 | 53.9 | +7.00% | 2026-06-23 | 🟢 Healthy, no alert — neutral RSI, small live-money position ($1.07 value) |
| theo-active | (none) | — | — | — | — | Empty book, $0 cash |
| yawnxyz-rh | META | $616.77 | 62.4 | +157.30% | 2026-07-10 | 🔒 Rockets tier — no alert |
| yawnxyz-rh | NET | $278.92 | 46.7 | +88.50% | 2026-07-10 | 🔒 Rockets tier — no alert |
| yawnxyz-rh | NTLA | $12.74 | 52.1 | -84.17% | 2026-07-10 | Memorial tier — no explicit stop-loss recorded in this legacy (2026-07-10) mirror; position status shown for awareness only |
| yawnxyz-rh | EDIT | $3.23 | 57.5 | -95.20% | 2026-07-10 | Memorial tier — same as above |
| yawnxyz-rh | AMC | $2.65 | 55.1 | -96.23% | 2026-07-10 | Memorial tier — same as above |
| yawnxyz-rh | FUBO | $11.25 | 60.6 | -97.50% | 2026-07-10 | Memorial tier — same as above |
| yawnxyz-rh | DNA | $6.95 | 44.3 | -98.40% | 2026-07-10 | Memorial tier — same as above |
| yawnxyz-rh | PHGE | $0.15 | 39.6 | -99.14% | 2026-07-10 | Memorial tier — same as above |
No stop-loss framework is recorded in the ledger for yawnxyz-rh — it is a read-only mirror seeded once on 2026-07-10 with no automatic re-sync since (holdings-snapshot.md). The six Memorial-tier positions above are shown for position-management awareness only, not as trading signals; none of these tickers were checked against, or fed into, the Entry Zones tables above. Full 27-position detail (all three accounts) is in research/issues/2026-09-04-holdings-snapshot.md.
Sector Rotation Signals
| Signal | Trigger | What It Means | Status |
|---|---|---|---|
| Rotation within tech, not risk-off | Hardware/memory rip (semis median 1D +3.19%, memory names +5.9-8.7%) same session cloud/SaaS liquidates (15/16 red) — SPY itself only -1.18% off its high | A genuine sector rotation, not a broad-market selloff (etf-universe, market-pulse, ai-scan) |
🟡 Watch which side confirms |
| Structure inverts the price action | Semis pop (only 4/18 in a genuine uptrend, ordinary volume) vs. cloud/SaaS selloff (14/16 still uptrend, lightest volume of the cycle) | The chip pop is a WATCH not a buy; the software pullback is the higher-quality setup if it stays shallow | 🔍 Watch for confirmation either direction |
| No war premium in defense — a rates story, not a fear story | Whole prime-contractor book sold off 2-9% during a live US-Iran exchange of fire; oil and bond yields both spiked on Fed rate-hike bets instead (defense-contractors, geopolitical-risk, macro captures Sept 1-4) |
The market is pricing rates/inflation, which is also why software beat-and-sold-off this cycle (PANW, ZS, CRDO, MDB, CIEN) | 🔴 Live, unresolved |
| Gold and vol both fail to confirm fear | GLD crossed below its 200-day for the first time this stretch (-2.1%, -20.19% off its high); VIXY stays crushed (RSI 31.5, -57.2% off its high) | Neither safe-haven asset is behaving like this is a fear event — reinforces the rates-not-fear read | 🟡 Watch for a regime change |
| Credit cracks while equities stay calm | HYG's RSI collapsed 52.7→38.9 with a fresh death cross, the sharpest move on the entire market-pulse board; TLT/LQD/BND/IEF all deepened alongside it |
Genuine deterioration underneath a broad tape that looks orderly | 🔴 Watch closely into next cycle |
| Oil complex confirmed running hot | Every upstream/services oil name flipped from a negative to a solidly positive week (XOM, CVX, COP, DVN, EOG, SLB, HAL all strong-up); USO +91.37% 1Y and accelerating |
A live, dated, sourced catalyst (US-Iran escalation) — not speculative | 🟢 Confirmed, extended (VLO/PSX near overbought) |
| Cruise-line trio fully broken; airlines a level below | RCL, CCL, NCLH all confirmed broken on every long-horizon window (alpha worse than -44 across all three); airlines (LUV, AAL) cooled to just below their 200-days on the same oil-cost pressure | Two transportation sub-segments taking the same oil shock differently | 🔴 Cruise lines confirmed broken; airlines testing the line |
| Food security is the day's clean, healthy strength | 9 of 10 names gained RSI, NTR/DE/MOO all in confirmed breakout regime; only FMC/MOS's pops sit on top of genuinely broken 1Y charts | A rotation-leadership story inside a genuinely healthy book, not a melt-up | 🟢 Confirming (ex-FMC/MOS) |
| WFE/test-metrology equipment layer bounces on ordinary volume | CAMT, LRCX, ENTG, ASML, KLAC all up this cycle but on ordinary, not real-volume-confirmed, participation | Every name below its 200-day (CAMT, AMKR, BESIY, ACLS, MKSI) still carries positive 1Y and alpha — trend under test, not broken | 🟡 Watch for a real-volume follow-through |
Regime Transitions
Source: deno task market -- regimes:feed --json (code-diffed, not eyeballed). Newest group dated 2026-09-04, prior bar 2026-09-03 — 27 items, no from: null/to: null rows to skip this cycle.
| Stock | From | To | Thesis | Tape Date |
|---|---|---|---|---|
| ALNY | collapse | downtrend | unrated (inert) | 2026-09-04 |
| AMGN | breakout | uptrend | unrated | 2026-09-04 |
| AMT | pullback | uptrend | unrated | 2026-09-04 |
| ARKW | basing | pullback | unrated | 2026-09-04 |
| AU | pullback | uptrend | unrated | 2026-09-04 |
| BDX | breakout | uptrend | unrated | 2026-09-04 |
| BMY | breakout | uptrend | unrated | 2026-09-04 |
| BRKR | pullback | uptrend | unrated | 2026-09-04 |
| CBZ | pullback | uptrend | unrated | 2026-09-04 |
| COWZ | breakout | uptrend | unrated | 2026-09-04 |
| CRDO | pullback | collapse | unrated | 2026-09-04 |
| DTE | basing | pullback | unrated | 2026-09-04 |
| DVN | basing | uptrend | unrated | 2026-09-04 |
| FCX | breakout | uptrend | intact | 2026-09-04 |
| GRID | basing | pullback | unrated | 2026-09-04 |
| GTLB | pullback | uptrend | unrated | 2026-09-04 |
| HLT | basing | pullback | unrated | 2026-09-04 |
| HWM | uptrend | basing | unrated | 2026-09-04 |
| LMND | collapse | downtrend | unrated (inert) | 2026-09-04 |
| MDB | uptrend | pullback | unrated | 2026-09-04 |
| NIQ | parabolic | uptrend | unrated | 2026-09-04 |
| PANW | uptrend | pullback | unrated | 2026-09-04 |
| QCOM | collapse | pullback | unrated | 2026-09-04 |
| RBLX | collapse | downtrend | unrated (inert) | 2026-09-04 |
| RTX | uptrend | pullback | unrated | 2026-09-04 |
| SRPT | pullback | uptrend | unrated | 2026-09-04 |
| XLV | breakout | uptrend | unrated | 2026-09-04 |
CRDO's pullback→collapse cross-confirms tech-insider-buys' own read (broke below its 200-day on a confirmed-beat margin/guidance selloff). PANW's uptrend→pullback and RTX's uptrend→pullback both cross-confirm the day's beat-but-sell-off and defense-sector-pullback stories above. MDB's uptrend→pullback matches ai-scan's sharp earnings-driven reversal. QCOM's collapse→pullback matches its crossing back above the 200-day flagged in ai-scan/nvda-ecosystem. The remaining 20 transitions are broad-tape churn (mostly →uptrend upgrades) with thesisState: unrated throughout except FCX (intact).
Insider Activity Summary
| Signal | Stocks | Implication |
|---|---|---|
| Buying table nearly frozen for a fourth straight cycle | TSLA (+$219.9M), CRSP (+$28.1M), ASAN (+$8.7M), INTC (+$2.4M) all unchanged; TSM ticked up to +$1.8M from +$1.6M | First dollar-figure movement in this tracked window, though too small to read as a fresh conviction signal |
| Golden signal (RSI<30 + buying): zero for a fifth straight cycle | None — closest is PINS at RSI 31.9, itself structurally broken (-45.04% 1Y), not a bargain | The clean read now spans five full cycles |
| Tool-vs-summary RSI discrepancy recurs, this cycle on AVGO | Tool's cached RSI reads 27 (would flag a false selling-into-dip hit); insider-universe.json puts it at 38.1 |
Per standing doctrine the summary basis governs — no actual selling-into-dip hit this cycle |
| Heaviest captured selling, intact structure | AMZN (-$9.4B), NVDA (-$3.7B, up from -$3.3B), LLY (-$3.0B), PLTR (-$2.4B), PANW (-$1.5B) — essentially unchanged | Scale-driven, not a fresh bearish tell |
| BE's reversal extends into a genuinely large run | Cushion above the 200-day widened to +32.2% from +11.8%; 1Y now +360.52% (was +284.62%), insider selling flat at -$460.8M | Confirmed same-day catalyst (record Q2 results, raised FY guidance) — a genuinely strengthened structure, not just an extension |
| ASAN's frozen buying sits behind a real same-day drop | -12.69% same day; news pull (fetched 2026-09-04) discusses a beat-but-sell-off Q2 print, but stock-events/2026-09-03-asan-earnings.md still reads overdue-unconfirmed |
The event ledger has not promoted this to a captured print — reported as commentary, not a verified event |
| MRNA keeps climbing on a partial but incomplete driver | 1Y now +499.46% (was +463.91%), 30D +158.71%; the same litigation-settlement receipt (fetched 2026-08-28) is unchanged, no fresher receipt pulled this cycle | Third-plus straight cycle this desk has checked and found no corpus source for the full scale of the move |
| MRVL's earnings status resolved | Confirmed narrow beat, $0.94 actual vs $0.93 consensus (+1.0%), reported Aug 27 | Cleared last cycle's overdue-unconfirmed flag; stock recovered to RSI 51.1, +248.75% 1Y |
| ISRG thesis-flag vs. chart divergence, persisting | thesisState: intact but regime still collapse; vs-SMA200 -20.9% (worsened), 1Y -19.32% (improved slightly) |
Two long-horizon fields moved opposite directions this cycle — worth a direct fundamentals check |
Pending Research Actions
| Action | Source | Priority | Status |
|---|---|---|---|
| Does the semis pop get a flag/consolidation and a real-volume breakout, or does it fade back into the still-damaged 18-name cohort (only 4/18 uptrend)? | etf-universe | High | 🆕 New — the desk's key entry-discipline test this cycle |
| Does the cloud/SaaS light-volume pullback stay shallow (14/16 still uptrend), or does real distribution show up next cycle? | etf-universe, ai-scan, cloud-saas candidate | High | 🆕 New |
| Does defense's "sold off during a live war" pattern reverse once bond-yield volatility settles, or does the sector keep lagging as a rates trade? | defense-contractors, geopolitical-risk | High | 🆕 New |
| Does HYG's fresh death cross mark the start of genuine credit stress, or a one-cycle flush? | market-pulse | High | 🆕 New |
| Does LULU's collapse (RSI 31.8, confirmed Q2 miss) find a floor, or does the slide extend? | cultural-thesis, consumer-scan, retail-scan | High | 🆕 New |
| Does TJX's capitulation (now a 2nd straight scan as the board's lowest print, 1Y turned negative) resolve, or deepen further given the fresh Jefferies downgrade? | retail-scan | High | Carried, escalated (2nd cycle) |
| MRNA's driver — the litigation-settlement receipt is unchanged and still doesn't explain the +499.46% 1Y; does the move consolidate or keep extending? | biotech candidate, insider-scan | High | Carried, escalated (3rd-plus cycle) |
Does CRDO's break below its 200-day (confirmed-beat, margin-scare selloff, regime flipped to collapse) repair or deepen? |
tech-insider-buys, regimes-feed | High | 🆕 New |
| Does AVGO's now-resolved-down 200-day break extend, or reverse back above the line it tested unresolved for two cycles? | ai-scan, bargain-bin | Medium | Carried, resolved-down this cycle |
| Does the cruise-line trio (RCL/CCL/NCLH) find a floor now that all three are confirmed fully broken? | travel-leisure | Medium | Carried, escalated |
| Does SMCI's 1-year deficit (now -2.73%, was -15.67%) finally flip positive next cycle? | tech-insider-buys | Medium | Carried |
| ISRG's thesis flag (intact) vs. its chart (collapse, worsening on vs-SMA200) — which one resolves the disagreement? | insider-scan, biotech-scan, healthcare-scan | Low | Carried, escalated (3rd-plus cycle) |
| Does GLD's second straight cycle below its 200-day (-2.1%, -20.19% off its high) close, or does it deepen into a genuine trend break as the Iran war continues? | macro-commodities, geopolitical-risk, market-pulse | High | Carried, escalated |
The Gold
Top Signal
The semis-vs-cloud divergence is the desk's cleanest entry-discipline test this cycle, and the honest read is WATCH on both sides, not BUY on either. The chip complex ripped (median 1D +3.19%, several names up 9-13%) but only 4 of 18 names in that cohort carry a genuine uptrend and the move came on ordinary volume — per the desk's bottom→pop→flag→breakout discipline, that is explicitly NOT a confirmed entry regardless of how loud the one-day pop looks; it needs a flag/consolidation and then a real-volume breakout. The cloud/SaaS complex sold off the same session (15 of 16 red) but on the lightest volume of the cycle while 14 of 16 names stayed technically
up-trend — a shallower, higher-quality pullback if it holds. Layered underneath both: the actual structural break of the day is credit (HYG's death cross, the sharpest move on the wholemarket-pulseboard) and defense's total failure to catch a war bid — both are rates-driven signals, and both explain why software beat-and-sold-off across PANW, ZS, CRDO, MDB, and CIEN this cycle. Nothing in today's corpus clears the bar for an immediate buy; the correct action is watching for confirmation on the semis breakout and the software pullback staying shallow, while treating LULU (confirmed fundamental break) and the cruise-line trio (confirmed structural break) as the clean avoids.
Key Discoveries
| Discovery | Implication |
|---|---|
| Semis ripped (median 1D +3.19%) while only 4 of 18 names in that cohort carry a genuine uptrend, on ordinary volume | The pop is a WATCH under the desk's entry discipline, not a confirmed leg — needs a flag then a volume-confirmed breakout |
Cloud/SaaS sold off 15 of 16 red on the lightest volume of the cycle while 14 of 16 stayed up-trend |
A shallow, low-conviction pullback sitting on top of a mostly-intact structure — the higher-quality setup of the two if it holds |
| Defense sold off 2-9% during a live US-Iran exchange of fire, the opposite of a naive war-premium read | A rates/inflation story (oil + bond-yield spike on Fed rate-hike bets), not a fear or defense-specific story |
| PANW, ZS, CRDO, MDB, and CIEN all beat estimates and sold off anyway this cycle | Long-duration/growth names de-rating on genuinely good earnings is the clearest signature of a rates-driven tape, not a fundamentals one |
| HYG's RSI collapsed 52.7→38.9 with a fresh death cross — the sharpest move on the entire market-pulse board | Credit is cracking underneath a broad equity tape that otherwise looks calm (SPY only -1.18% off its high) |
| Gold crossed below its 200-day for the first time this stretch and VIXY stays crushed, even during a live war | Neither safe-haven asset confirms a fear trade — reinforces the rates-not-fear read |
| LULU crashed on a confirmed Q2 miss (revenue -4%, guidance cut, 8.69x volume) with no long-horizon counter-argument anywhere in this corpus | The cleanest, most confirmed AVOID in today's entire corpus |
| Chip designers (ARM -26.5% 3M, ON -36.6% 3M, AMKR -26.3% 3M) are broken while semicap equipment/test carries positive 1Y and alpha across every name | Within one sector, the layer decides the call — avoid the designers, watch the equipment layer |
Sourced from all 31 of today's scans + all 11 candidate roundups + the code-diffed regime-transitions feed (regimes:feed --json, snapshot dated 2026-09-04) + research/issues/2026-09-04-holdings-snapshot.md (Exit Alerts only, book-blind). Every price/RSI figure in this digest was validated directly against its source research/market-engine/data/summaries/*.json file this session. Prior digest for carry-forward: research/issues/2026-08-28-actions-daily-digest.md (7-day gap).
Sources
- Price/RSI/trend/long-horizon truth:
research/market-engine/data/summaries/*.json(2026-09-04 close bars), spot-checked directly this session for every ticker in this digest — includingdefense-contractors.json,travel-leisure.json,retail.json,defensive.json,biotech-scan.json,cultural-thesis.json,chemicals.json,market-pulse.json,etf-ideas.json,macro-commodities.json,insider-universe.json,ev-clean-energy.json,war-sensitivity.json,airlines.json,thebuildout-semiconductor-ip-licensors.json,thebuildout-data-center-switch-nic-vendors.json,thebuildout-advanced-packaging-2d3d-leaders.json,robotics.json,cloud-saas.json,monster-discoveries.json,focus.json,ai-power.json,spacex-s1-supply-chain.json,thebuildout-co-packaged-optics-cpo.json,market-etfs.json,holdings-etfs.json,healthcare.json. - Scan corpus:
research/scans/2026-09-04-*.md(31 files: ai-infrastructure, ai-scan, airlines, bargain-bin, biotech-scan, chemicals, cloud-etfs, consumer-scan, crypto-scan, cultural-thesis, cybersec, defense-contractors, defensive-scan, drone-defense, etf-universe, ev-clean-energy, food-security, geopolitical-risk, healthcare-scan, insider-scan, lazarus-scan, macro-commodities, market-pulse, monster-discover [excluded,status: invalidated], monster-scan, nvda-ecosystem, optical-supply-chain, retail-scan, supply-chain-traces, tech-insider-buys, travel-leisure, washout-scan, wfe-test-metrology). - Candidate roundups:
research/candidates/2026-09-04-*.md(11 files: ai-infra, biotech, cloud-saas, consumer, crypto, energy, mag7, memory, retail, space, web-cloud). - Regime transitions:
deno task market -- regimes:feed --json(2026-09-04 snapshot, diffed against 2026-09-03). - Holdings/exit alerts:
research/issues/2026-09-04-holdings-snapshot.md(sourced fromholdings list/positionsCLI +holdings-stocks.json). - News/event receipts cited inline per scan: LULU earnings (
research/stock-events/2026-09-03-lulu-earnings.md), CIEN earnings (research/stock-events/2026-09-03-cien-earnings.md), NIO earnings (research/stock-events/2026-09-01-nio-earnings.md), MRVL earnings (research/stock-events/2026-08-27-mrvl-earnings.md), PANW/ZS news (fetched 2026-09-04), oil/Iran macro captures (research/capture/articles/2026-09-01-marketwatch-*,2026-09-02-cnbc-*,2026-09-02-bloomberg-*,2026-09-04-cnbc-*,2026-09-04-bloomberg-*). - Prior digest for carry-forward:
research/issues/2026-08-28-actions-daily-digest.md.
No direct external sources are attached to this read.