Loop

Loop

How is the autopilot doing, what has it done, and what has it learned? The autopilot's report, in that order.

Projection

The autopilot's report. The desk runs its ideas as paper books — no real money trades itself; live execution stays human. Three questions, in order: how is it doing (the trust ladder), what has it done (open calls and their lifecycle), and what has it learned (attribution, you-vs-desk, live experiments). Your own graded decisions live in the Decision Log.

How it's doing · Trust ladder

Books earn trust rungs on completed outcomes — promotion is earned, demotion is automatic. As of the last weekly pass, 2026-08-12.

ark-innovation 11 45% -2.7% hold
autotrade 6 50% -0.1% hold
bench-signals 94 41% -1.2% hold
buffett-13f 4 75% +15.8% hold
burry-13f 1 0% -7.6% hold
claude-momentum 72 46% -0.7% hold
claude-trader 45 47% -0.3% hold
claude-trader-trail 3 33% -5.1% hold
dca-larry 4 50% +0.3% hold
druckenmiller-13f 45 71% +8.0% hold
gottheimer-trades 5 60% +16.0% hold
mccaul-trades 2 0% -3.2% hold
pelosi-trades 18 61% +0.8% hold
swing-kid 2 50% +0.6% hold
wasserman-schultz-trades 2 50% +9.0% hold
yolo 44 48% -0.5% hold
jan 1 100% +8.7% hold
theo 2 100% +13.6% hold

DIRECTION-SIGNED: negative = bad for the book regardless of side; a sell whose name fell after exit scores positive.

Not yet on the ladder: Robinhood · live — live state on Positions; a book earns a rung after its first weekly pass with scored outcomes.

What it's done · Open calls

Open calls are falsifiable desk stances with setup state, entry and invalidation triggers, a review-by date, and raw return windows.

Entry triggerInvalidation trigger
AVAV watch watching a pullback that bases in the $163-$180 band — the reclaimed 20-day ($165.70), 50-day ($163.22) and the $176.42 shelf — not a chase into the $199.98 volume wall a close below $158, which gives back both the 50-day and the anchored VWAP from the low ($160.78) Sep 15 0.0% to date 0.0% to date
AXON watch watching a pullback that holds the $576.14 shelf or the 20-day ($556.43) a close below the 50-day ($525.63) Sep 30 0.0% to date 0.0% to date
ITA pass watching no entry 2.69% off a 52-week high with RSI 64.4; the wrapper prices the primes, not the drone chain a pullback into the $239.75-$243.94 moving-average band forces a re-look — as a defense position, not a drone one Sep 30 0.0% to date 0.0% to date
KTOS pass watching no entry while price sits 12.3% below the 200-day ($73.65) with a death cross intact and RSI at 70.9 a base built and held above the 200-day forces a re-look Sep 30 0.0% to date 0.0% to date
RCAT pass watching no entry after a +20.85% week that stalled at the anchored VWAP from the low ($12.25) with a death cross intact a base above $12.25 that holds the $10.38 shelf forces a re-look Sep 30 0.0% to date 0.0% to date
UMAC watch watching a pullback that bases in $21-$24, where the value-area high ($21.09), 50-day ($22.47) and 20-day ($23.21) converge; never the parabola a close below $18.50, or a third straight quarter of widening operating loss on higher revenue Oct 9 0.0% to date 0.0% to date
DUST pass watching n/a — the inverse side of the same instrument problem n/a Sep 11 -3.3% to date -3.3% to date
GDX watch watching don't chase RSI 70.3 after a +21.31% week — a first pullback that holds the 200-day ($87.37) and the $87.83 shelf is the entry the structure supports a close back below the 200-day turns the reclaim into a spike Sep 11 +1.6% to date +1.6% to date
GLD watch watching a close above the 200-day ($411.70) that holds with RSI above 55; $445 reopens the full recovery case a close below the July 16 low ($364.96) Sep 4 +0.7% to date +0.7% to date
MU watch watching reclaims and holds the 20-day ($888.26) with RSI back above 50 RSI falls back under 40 while price stays below the 20-day Sep 11 +17.5% to date +17.5% to date
NUGT pass watching n/a — leveraged miner exposure compounds the whipsaw that already makes miners the wrong instrument for timing this trend n/a Sep 11 +2.8% to date +2.8% to date
SIL watch watching above its 200-day by 2.4% but RSI 67.1 after +20.0% — same pullback discipline as GDX, entry on a hold of the 200-day ($86.39) a close back below the 200-day Sep 11 +0.5% to date +0.5% to date
SIMO watch watching price back above the 20-day ($264.25) with the 30-day loss narrowing RSI below 40 Sep 11 +18.3% to date +18.3% to date
SLV watch watching the laggard of the complex at 9.8% below its 200-day ($63.76) — a reclaim of that line is what would confirm bullion joining the miners loses the 50-day ($56.64), leaving the week a miner-only move Sep 11 +0.3% to date +0.3% to date
SNDK pass watching no entry while price sits 13.0% below the 20-day ($1,393.06) and 28.2% below the 50-day a base built back above the 20-day on rising volume forces a re-look Sep 11 +44.3% to date +44.3% to date
STX watch watching holds the 20-day ($834.21) and turns RSI back above 50 either drive maker signals added unit or greenfield capacity, or a close below the 200-day ($533.00) Sep 11 +24.2% to date +24.2% to date
UUP watch watching n/a — the dollar at RSI 39.5 is the tailwind, not the trade a UUP reclaim of its 20-day ($28.34) on rising RSI drains the rates channel this rally runs on Sep 11 -0.1% to date -0.1% to date
WDC watch watching not before the Aug 5 call's allocation and gross-margin language is public and read; then a reclaim of the 20-day ($511.47) any unit- or greenfield-capacity addition signalled by either drive maker — the supply-discipline pillar fails Sep 11 +22.3% to date +22.3% to date
GLD watch watching a close above the 200-day ($411.72) that holds with RSI above 55; $445 reopens the full recovery thesis; above 411.72 a close back below the July 16 low ($364.96) — bounce failed, downtrend resumes; below 364.96 Sep 4 +4.1% to date +4.1% to date
CVX watch watching strongest RSI in the upstream cohort (67.2) — hold above the 20-day (~$185.34) to stay constructive loses the 50-day (~$182.75) Sep 11 +3.0% to date +3.0% to date
GLD watch watching this is a signal to watch, not a trade — a reclaim of the 20-day (~$373.15) would be the first sign of the safe-haven bid this week's tape didn't produce n/a — continued weakness alongside a weakening dollar (UUP) is the non-confirmation already in the data Sep 11 +9.1% to date +9.1% to date
HAL avoid watching not a trend-hold candidate yet — wait for a base and a reclaim of the 50-day (~$36.18) before treating any bounce as more than a dead-cat pop n/a — the condition to flip constructive is the 50-day reclaim above; a fresh 52-week low confirms the derate instead Sep 11 +7.7% to date +7.7% to date
PSX watch watching RSI is a hair under overbought (69.7) — wait for a pullback to the 20-day (~$200.76) rather than chasing the extended print loses the 50-day (~$185.74), turning the refining leg's leadership into a rollover Sep 11 +13.6% to date +13.6% to date
USO watch watching the -5.5% weekly pullback holds above the 20-day (~$121.96) — a normal pullback inside a strong-up trend, not a reversal closes back below the 20-day and takes out the $126.55 support shelf Sep 11 +0.9% to date +0.9% to date
VLO watch watching same setup as PSX — a pullback to the 20-day (~$295.63); note this row is one session stale (2026-07-30 close), confirm Friday's print before acting loses the 50-day (~$269.30) Sep 11 +11.0% to date +11.0% to date
CEG watch review-overdue reclaim the 20-day (~$256) with hyperscaler spending language intact on the Jul 29–30 prints capex language softens and the July shelf breaks — then the contracted-electrons bid is repricing too Aug 8 +9.7% to date +9.7% to date
ELAN watch invalidation-flagged a volume-confirmed (at least average volume) close through the 52-week high (~$27.72), or a post-Q2-print hold above it — the breakout the caution-overridden board prompt has been flagging, finally confirmed instead of chased; above 27.72 a close below the 50-day (~$23.68) — the near-high structure failed; back to priced; below 23.68 Aug 14 -6.1% to date -6.1% to date
GEV watch review-overdue held the AI unwind at −5%; reclaim the 20-day (~$1,085) and print a higher low loses the July low — the complement bid was narrower than the month suggested Aug 8 -0.0% to date -0.0% to date
LEU watch review-overdue first higher low after a clean Jul 29–30 capex read — the −18% month is a beta markdown, and the entry is print-gated, not chart-gated lower lows after capex language holds firm — then the fuel-cycle derate is regime, not beta Aug 8 +17.7% to date +17.7% to date
CLMT pass watching RSI 83 on 2.7x volume, +165% in a year — parabolic-chaser material, not trend-hold; reopen only on a first pullback that holds $36.99 (20-day) with volume drying up n/a — not owned; wrong if $36.99 holds on the first test and the trend resumes Sep 4 +7.2% +11.3% to date
DK pass watching reopen if it builds a multi-week base above the 20-day ($52.58) and resumes — a small refiner +53% in 30 days and +20% over its own 20-day is the gap, not the setup n/a — not owned; the pass is wrong (and will be scored wrong) if that base forms and the resumption runs without us Sep 4 +0.2% +4.4% to date
MPC watch watching the lane's quality large-cap; interest begins on the first pullback that HOLDS the 20-day ($272.00) after the Aug 4 print validates Q2 margins — at +15% above the 20-day, buying today buys the gap a close below the 50-day ($261.52) breaks the momentum regime and the lane read resets; below 261.52 Aug 21 +9.5% +13.6% to date
MU watch in-zone Friday printed a reversal bar (low $804, closed $848.95); a close above that bar's high ($903.93) confirms it — the trend-hold entry itself still needs the base and a 20-day reclaim ($1,023); above 903.93 a close below Friday's low ($804) voids the reversal; the knife is still falling; below 804 Aug 14 +12.8% +16.9% to date
NVDA watch in-zone the general never broke (−3.7% in 30 days vs the ETF's −15.7%); a reclaim of its 50-day ($209.81) is the complex-repair signal — the leader turns first; above 209.81 a close below the $186.48 volume shelf means the general finally fell and the whole complex read resets bearish; below 186.48 Aug 14 +6.6% +10.7% to date
PANW watch watching the lane's strongest structure (-2.7% from its 52-week high, +83% on the year); first pullback that holds the 20-day ($327.50) after this +10% week a close below the 50-day ($283.88) breaks the momentum regime for the lane's leader; below 283.88 Sep 4 +3.6% +7.8% to date
PL watch invalidation-flagged untouchable while the knife falls (RSI 30, −14% on the week); interest begins when BOTH hold: a base that reclaims the 20-day (~$27.72 today) AND the latest quarter's net-loss spike (−$0.40 EPS vs −$0.04 a year ago) explained at the next print or in the filing — quality-on-sale requires the quality confirmed first a close below the 30-day low ($21.03) opens the next leg down and resets the whole watch; below 21.03 Sep 11 +6.0% +10.1% to date
PSX watch watching the diversified laggard of the three (+70% on the year, midstream/chem mix dilutes the crack exposure); first 20-day hold ($181.86) after the Aug 5 print a close below the 50-day ($178.65) — barely below the 20-day, so this one resolves fast; below 178.65 Aug 21 +11.1% +15.2% to date
S watch watching the laggard catching up (+30% in 30 days, still only +9% on the year and -8.6% from its high); after this week's +9.5% the entry is the first pullback that HOLDS the 20-day ($17.39) — chasing 12.5% above it buys the gap a close below the 50-day ($16.85) — the 20 and 50 sit close together here, so a failed hold resolves the read fast; below 16.85 Sep 10 +8.0% +12.1% to date
SMH watch invalidation-flagged the complex switch: first close back above the 50-day ($597.33) after a higher low, confirmed through the 20-day ($610.93); until then every bounce is a bounce inside a downtrend; above 597.33 a close below Friday's low ($536.81) means the bounce attempt failed and the next leg down is open; below 536.81 Aug 14 +2.2% +6.3% to date
STX watch in-zone the complex's only accumulation-shaped bar: +5.7% on 1.35× volume, closed near the high after touching $700.39; above Friday's high ($808.65) with the July 28 print ahead, the bounce has legs; above 808.65 a close below $700.39 means Friday was a dead-cat, not a turn; below 700.39 Aug 14 +19.9% +24.0% to date
VLO watch watching best 12-month tape of the pure refiners (+117%); same frame — first 20-day hold ($271.29) after the Jul 30 print, the earliest verification event in the lane a close below the 50-day ($257.97) breaks the momentum regime; below 257.97 Aug 21 +7.1% +11.2% to date
ARM watch in-zone reclaim the pre-registered $270 line and hold it — the trigger the thesis flag broke this week, now the recovery tell; above 270 lower lows continue below $270 after the prints — then the derate is regime, not scare Aug 8 -2.4% +1.6% to date
CCJ watch watching washed out at RSI 30 — wait for the $84 shelf to hold and a first higher low; the 50-day (~$105) reclaim is the confirmation, not the entry the $84 shelf fails — then the fuel-cycle unwind has further to run; below 84 Sep 15 +11.2% +15.1% to date
CEG watch in-zone the merchant bellwether needs a turn signature, not a dip-buy — a base under the 50-day (~$269), then a reclaim that holds with a higher low; above 269 a PJM-style revenue cap lands on its core markets, or the base breaks to new lows below the July shelf Sep 15 +6.3% +10.2% to date
DUK watch invalidation-flagged the one working expression — regulated rate-base growth from the same demand; a close at a fresh 52-week high that holds confirms the leadership loses the 200-day (~$123) — the defensive bid is gone; below 123 Sep 15 -5.1% -1.1% to date
LEU watch in-zone only after a full-quarter base — the enrichment monopoly story is intact but the stock is in a −68%-from-high collapse; a 50-day (~$177) reclaim after a multi-week base is the earliest signal worth acting on; above 177 HALEU contract flow stalls, or the base never forms and lower lows continue Oct 1 +14.0% +18.0% to date
MU watch review-overdue gated on the prints, not the chart — a storage/memory print (STX Jul 28, WDC Jul 29, SNDK Aug 5) showing DRAM/NAND pricing intact, plus a first higher low a print that shows pricing momentum actually rolling — then the derate was early, not wrong Aug 8 +15.2% +19.2% to date
NVDA watch review-overdue the R1 playbook leg — hold the 20-day retest (~$202) through the July 28–29 storage prints, then a higher low; the efficiency-scare component tends to mean-revert if the fundamentals hold the retest fails and the pullback becomes a breakdown below the July shelf, or hyperscaler capex language actually softens on the July 29–30 prints Aug 8 +7.0% +10.9% to date
TLN watch invalidation-flagged the strongest tape of the merchants — hold the $360 shelf and the 200-day (~$371), then clear the 20-day (~$394) to signal the pullback resolved up; above 394 a close below $360 turns the pullback into a breakdown; below 360 Sep 15 -8.1% -4.2% to date
VST watch invalidation-flagged reclaim the 200-day (~$166) and hold it — first sign the derate is done; above 166 capacity-revenue caps spread to ERCOT/PJM exposure, or price breaks down from the current shelf instead of basing Sep 15 -9.9% -6.0% to date
ABCL watch in-zone the pop already happened (Feb low $2.78 to a $8.44 high) — this is the first flag; interest is the 52-week-high reclaim above $8.45, which would say the pullback was digestion, not distribution; above 8.45 a finish under the 50-day (~$5.56) — the four-month turn structure is broken there and it goes back to being a runway story; below 5.56 Aug 31 +25.3% +53.6% to date
CBRS watch in-zone a close above the post-flush bounce high ($221.30) confirms the higher low and resolves the debut range up — the first clean post-debut base signal, not the open; above 221.3 a finish under the flush low ($160.81) — no base; the debut air pocket is still open; below 160.81 Aug 31 +13.0% +26.9% to date
META watch invalidation-flagged second reclaim attempt off a higher low — buy interest is a close above ~$665, through the falling 200-day ($642) and the year's volume shelf ($662) together; above 665 a finish under $563 says the June higher low failed and the range is still broken; below 563 Aug 31 -8.6% -9.9% to date
MSFT watch in-zone no turn signature yet — interest begins at the 200-day reclaim (~$442) after a base; until then this is a falling mega-cap, not a setup; above 442 a finish under the June 25 low ($352.83) opens a fresh leg down and resets the whole watch; below 352.83 Oct 1 +28.8% +25.0% to date
NTLA watch invalidation-flagged golden cross, +31% month, and a fresh impulse into $18.48 on July 2 — buy interest is the breakout above that swing high while still ~40% below the 52-week high; above 18.5 a finish under the ~$14.20 cluster (50-day and the nearest volume support shelf) says the turn failed; below 14.2 Aug 31 -25.0% -21.7% to date
ORCL pass watching reopen when the knife stops: the February low ($136) holds on this retest AND a +7% week prints off it — the bottom-pop signature the desk buys, instead of the falling edge it doesn't n/a — not owned; the pass is wrong (and will be scored wrong) if this retest of $136 holds and becomes the durable low Oct 1 +2.3% +2.0% to date
AXTI pass watching reopen only if it bases above its 200-day (~$42.55) for a full quarter AND revenue inflects — a collapsed 25x parabolic with sub-$100M revenue and operating losses is a lottery aftermath, not a trend n/a — not owned; the pass is wrong if the substrate cycle produces a durable base and profitability Oct 1 +43.6% +59.6% to date
CIFR pass watching reopen only on revenue inflecting up AND visible de-leveraging of the $4.4B debt stack n/a — not owned; the pass is wrong if hosting revenue ramps against falling leverage Oct 1 -25.1% -15.3% to date
CORZ pass watching reopen on a recapitalization that restores positive equity with the CoreWeave tenancy intact n/a — not owned Oct 1 -14.3% -14.3% to date
CRDO buy in-zone elite fundamentals (33% operating margin on $1.34B FY26 revenue) with the least-damaged tape in the sweep — already flagging under its 20-day; buy the reclaim (~$255.50); above 255.5 a finish under the 200-day (~$160) — until then pullbacks are digestion in an intact uptrend; below 160 Sep 15 -7.1% +9.3% to date
CRWV pass watching reopen only on a financing regime change — interest expense stabilizing while revenue keeps compounding, and a reclaimed $98–100 (the point-of-control / 200-day cluster) n/a — not owned; the pass is wrong if CRWV re-rates through $100 on improving coverage Oct 1 -3.0% +17.8% to date
EQT watch watching the fundamentals are the best in the sweep (net income nearly quadrupled in two quarters) but the tape trades below its 200-day — watch for the reclaim of the ~$55.60 cluster (50-day, 200-day, one-year VWAP); above 55.6 gas strip rolling over while price stays below the 200-day through the October print Oct 1 -2.4% +3.9% to date
HUT pass watching reopen on a financed capacity deal that closes the gap between $160M of cash and gigawatt-scale ambitions n/a — not owned; the pass is wrong if the power-first strategy lands funding without heavy dilution Oct 1 -20.3% -17.0% to date
KGS buy invalidation-flagged the sleeper — record Q1, raised full-year guidance to include the Distributed Power acquisition, and a 300–500MW/yr power-generation growth outlook through 2030; buy the 50-day reclaim (~$70); above 70 a finish under $57.50 (top of the prior value area) says the datacenter-power re-rate failed; below 57.5 Aug 15 -18.0% -5.3% to date
MXL watch in-zone wire-flagged inflection but already a 4.5x over the year and −16% this week on no news we can find — needs the flag to hold; interest returns above the 20-day (~$91); above 91 a finish under the one-year VWAP (~$50.66) unwinds the entire re-rate; below 50.66 Aug 15 -16.6% -1.5% to date
NBIS buy in-zone pullback within an intact uptrend — buy the 50-day reclaim at ~$217, which says the seven-day air pocket resolved; above 217 a finish under the 200-day (~$134) — at that point the hypergrowth story is repricing, not consolidating; below 134 Aug 31 -16.9% +24.2% to date
OKTA pass watching the rotation's confirming leader — but RSI 78 and 3% off its high is chasing; reopen on the first orderly pullback that holds the 50-day n/a — not owned; the pass is a timing pass, not a thesis pass Sep 15 -2.7% -2.4% to date
TER buy in-zone the −21% week has no company event behind it — Q1 revenue rose 87% with a 21% EPS beat and the wire flagged the margin inflection; buy the 50-day reclaim (~$383), with the July 28 print as the catalyst; above 383 a finish under the 200-day (~$270), or the July 28 print showing the 'lumpy, concentrated among large customers' risk arriving early; below 270 Aug 15 +4.2% +26.1% to date
WDAY watch watching the earnings proof-point of the rotation (beat + raised subscription guide) but still in a downtrend, 43% off its high — watch for a base; this is a lane signal, not yet an entry the bounce fades below the June low with the lane's 7-day breadth rolling back under 50% Sep 15 +26.6% +38.7% to date
WULF watch invalidation-flagged the $2.63B war chest has to start converting into hosting revenue on the P&L — watch the Q2 print; a 50-day reclaim (~$24.30) alongside it upgrades this to a buy candidate; above 24.3 cash burn without contracted revenue landing by the Q3 print, or a breakdown under the $15.67 support shelf; below 15.67 Aug 15 -28.9% -22.9% to date
SKHY watch in-zone first clean post-debut base — not the open (debut ~2026-07-10); above 151.05 DRAM ASP momentum turns negative, HBM4E timeline slips, or the debut air-pocket breaks down without basing Sep 1 n/a n/a
ZM watch in-zone 2 consecutive quarters of accelerating revenue; base over $92; above 92 revenue still flat through the Q3 FY27 print, or a breakdown below the $77.50 March low; below 77.5 Sep 1 +16.0% +24.3% to date
MU buy watching own the up-cycle with cycle sizing — record DRAM/HBM pricing power, HBM booked through 2027; size so a 20-30pt margin reversion is survivable DRAM ASP momentum turns negative — reduce first, ask questions later; also watch the SKHY ADR closing the proxy-premium gap Sep 1 -12.0% +2.7% to date
SNDK watch watching only on evidence the +69% NAND operating margin is a baseline, not the peak — two more prints holding pricing NAND ASP momentum rolls — the −111%→+69% swing is a cycle-top signature, not a new run-rate Sep 1 -20.8% +2.4% to date
STX buy watching own the cleanest un-hyped inflection — nearline-HDD margins compounding 23%→32% over four consecutive quarters on AI data-storage demand nearline margin progression stalls or reverses for two consecutive quarters, or DRAM ASP momentum turns negative (the cycle tell) Sep 1 -5.3% +14.6% to date
WDC buy watching same nearline thesis as STX — margins 10%→36% on the same AI storage driver, no hype premium nearline margin progression stalls or reverses for two consecutive quarters, or DRAM ASP momentum turns negative (the cycle tell) Sep 1 -7.7% -7.2% to date

RAW symbol-vs-SPY since the call: for trim/sell calls a negative number means the call worked.

What it's done · Call history

Per-call lifecycle timeline; each chip is date, stance, status, and setup state.

AVAV The Defense Trade Is at Its Highs. The Drone Trade Is Half Off.
Aug 15 watch open watching
AXON The Defense Trade Is at Its Highs. The Drone Trade Is Half Off.
Aug 15 watch open watching
GD the end of armor again
Aug 15 pass open watching
ITA The Defense Trade Is at Its Highs. The Drone Trade Is Half Off.
Aug 15 pass open watching
KTOS The Defense Trade Is at Its Highs. The Drone Trade Is Half Off.
Aug 15 pass open watching
NBIS The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 buy open watching Jul 12 buy open in-zone Aug 13 buy TRIGGER FIRED ON 2026-08-12 CLOSE - NBIS opened above the old ~$217 reclaim at $226.00 and closed $259.20, 16.8% above its $221.83 SMA50, on a +34.14% day. The condition fired, but the gap skipped the intended pullback entry; the 2026-08-13 deep dive supersedes the setup with a $220-235 held-retest/base zone and a $216 setup stop. n/a Aug 15 buy open in-zone
RCAT The Defense Trade Is at Its Highs. The Drone Trade Is Half Off.
Aug 15 pass open watching
RTX the end of armor again
Aug 15 pass open watching
TLN Power Is Short, the Power Trade Is Down — Both Are Right
Jul 17 watch open watching Jul 26 watch open invalidation-flagged Jul 31 watch INVALIDATED - TLN $334.10 has broken BELOW both the $360 support and the ~$371 200-day (current SMA200 $366.38), RSI 39.4, down trend, downtrend regime, -7.4% 30D, -25.97% off 52wk high (research/market-engine/data/summaries/ai-power.json, 2026-07-31 close). The hold condition this call was tracking has failed. n/a Aug 15 watch open invalidation-flagged
UMAC The Defense Trade Is at Its Highs. The Drone Trade Is Half Off.
Aug 15 watch open watching
MU AI-semi derate: the tape says sector derate not election lull — turn signature armed
Jul 18 watch open watching Jul 22 watch open in-zone Jul 29 watch open invalidation-flagged Aug 12 watch open in-zone
VST Power Is Short, the Power Trade Is Down — Both Are Right
Jul 17 watch open watching Jul 23 watch open in-zone Aug 12 watch open invalidation-flagged
ABCL ABCL — watch above $8.45
Jul 9 watch open watching Jul 26 watch open invalidation-flagged Aug 10 watch open in-zone
CEG What to Own When Intelligence Is Free
Jul 19 watch open watching Aug 10 watch open review-overdue
DUK Power Is Short, the Power Trade Is Down — Both Are Right
Jul 17 watch open watching Aug 10 watch open invalidation-flagged
GEV What to Own When Intelligence Is Free
Jul 19 watch open watching Aug 10 watch open review-overdue
LEU What to Own When Intelligence Is Free
Jul 19 watch open watching Aug 10 watch open review-overdue
MU The K3 Scare Is Not the R1 Scare
Jul 17 watch open watching Aug 10 watch open review-overdue
NVDA The K3 Scare Is Not the R1 Scare
Jul 17 watch open watching Aug 10 watch open review-overdue
DUST The Break Arrives, and It Arrives Overbought
Aug 8 pass open watching
ELAN Weekly money-loop pass — 2026-07-19
Jul 19 watch open watching Aug 8 watch open invalidation-flagged
GDX The Break Arrives, and It Arrives Overbought
Aug 8 watch open watching
GLD The Break Arrives, and It Arrives Overbought
Aug 8 watch open watching
MU Memory Splits: the Shortage Story Is Intact, the Single Names Are Not
Aug 8 watch open watching
NUGT The Break Arrives, and It Arrives Overbought
Aug 8 pass open watching
SIL The Break Arrives, and It Arrives Overbought
Aug 8 watch open watching
SIMO Memory Splits: the Shortage Story Is Intact, the Single Names Are Not
Aug 8 watch open watching
SLV The Break Arrives, and It Arrives Overbought
Aug 8 watch open watching
SNDK Memory Splits: the Shortage Story Is Intact, the Single Names Are Not
Aug 8 pass open watching
STX Memory Splits: the Shortage Story Is Intact, the Single Names Are Not
Aug 8 watch open watching
UUP The Break Arrives, and It Arrives Overbought
Aug 8 watch open watching
WDC Memory Splits: the Shortage Story Is Intact, the Single Names Are Not
Aug 8 watch open watching
GLD Gold's Second-Best Day of the Year Is Not a Safe-Haven Story
Aug 6 watch open watching
CBRS CBRS — candidate base, not a base
Jul 9 watch open watching Aug 4 watch open in-zone
SKHY SK Hynix Lists on Nasdaq as SKHY: the HBM Monopoly's Record ADR, the Wonky Seoul Tape, and Why It's US-Listing at All
Jul 8 watch open watching Jul 18 watch open review-overdue Jul 29 watch open invalidation-flagged Aug 4 watch open in-zone
TER The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 buy open watching Aug 4 buy open in-zone
LEU Power Is Short, the Power Trade Is Down — Both Are Right
Jul 17 watch open watching Aug 3 watch open in-zone
CVX Oil Quietly Reaccelerates While Gold Sits Out the Unwind
Aug 1 watch open watching
GLD Oil Quietly Reaccelerates While Gold Sits Out the Unwind
Aug 1 watch open watching
HAL Oil Quietly Reaccelerates While Gold Sits Out the Unwind
Aug 1 avoid open watching
PSX Oil Quietly Reaccelerates While Gold Sits Out the Unwind
Aug 1 watch open watching
USO Oil Quietly Reaccelerates While Gold Sits Out the Unwind
Aug 1 watch open watching
VLO Oil Quietly Reaccelerates While Gold Sits Out the Unwind
Aug 1 watch open watching
META META — watch above $665
Jul 9 watch open watching Jul 12 watch open in-zone Jul 30 watch open invalidation-flagged
MSFT MSFT: no signal. ORCL: falling knife.
Jul 9 watch open watching Jul 30 watch open in-zone
WULF The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 watch open watching Jul 29 watch open invalidation-flagged
KGS The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 buy open watching Jul 28 buy open invalidation-flagged
SMH AI-semi derate: the tape says sector derate not election lull — turn signature armed
Jul 18 watch open watching Jul 28 watch open invalidation-flagged
PL PL: the focus name the desk never read — candidate question answered
Jul 18 watch open watching Jul 26 watch open invalidation-flagged
ARM The K3 Scare Is Not the R1 Scare
Jul 17 watch open watching Jul 20 watch open watching Jul 22 watch open in-zone
CEG Power Is Short, the Power Trade Is Down — Both Are Right
Jul 17 watch open watching Jul 22 watch open in-zone
NVDA AI-semi derate: the tape says sector derate not election lull — turn signature armed
Jul 18 watch open watching Jul 22 watch open in-zone
STX AI-semi derate: the tape says sector derate not election lull — turn signature armed
Jul 18 watch open watching Jul 22 watch open in-zone
CLMT Refining: the coverage miss, read properly — crack-spread lane, trigger-or-pass
Jul 18 pass open watching
DK Refining: the coverage miss, read properly — crack-spread lane, trigger-or-pass
Jul 18 pass open watching
MPC Refining: the coverage miss, read properly — crack-spread lane, trigger-or-pass
Jul 18 watch open watching
PANW Cybersec entry conditions: the rotation-IN lane gets its written triggers
Jul 18 watch open watching
PSX Refining: the coverage miss, read properly — crack-spread lane, trigger-or-pass
Jul 18 watch open watching
S Cybersec entry conditions: the rotation-IN lane gets its written triggers
Jul 18 watch open watching
VLO Refining: the coverage miss, read properly — crack-spread lane, trigger-or-pass
Jul 18 watch open watching
APLD The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 watch open watching Jul 15 watch open invalidation-flagged Jul 17 watch 🔴 STOPPED/INVALIDATED — APLD $25.79, down, RSI 26.6 (oversold), closed under the $28 invalidation (1-yr VWAP $28.92, -10.8% vs VWAP). Price-based invalidation triggered; entry-interest level ($34.21) never reached. n/a
CCJ Power Is Short, the Power Trade Is Down — Both Are Right
Jul 17 watch open watching
CGNX The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 buy open watching Jul 12 buy open in-zone Jul 17 buy ✅ HIT (07-11), holding — CGNX $64.07, weak-down, RSI 47.5, -3.3% vs SMA20 ($66.24) — pulled back below the 20-day again but stays +27.9% above the ~$49 200-day invalidation. Thesis intact, just consolidating post-trigger. n/a
IREN The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 buy open watching Jul 16 buy open invalidation-flagged Jul 17 buy 🔴 STOPPED/INVALIDATED — IREN $33.62, down, RSI 30.8, regime collapse. Closed under the $35 invalidation level; the 07-07 bounce never became a base and the reclaim of $44.50 was never reached. n/a
LASR The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 buy open watching Jul 9 buy open in-zone Jul 17 buy ✅ HIT (07-11), holding — LASR $69.25, weak-down, RSI 51.7, +4.8% vs SMA20 ($66.08), still above the $65.50 trigger level. Pulled back mildly from the $72.05 07-11 read but well clear of the ~$51 invalidation cluster. n/a
MDB The Anti-AI Trade Is Turning: Cloud Software Is the Emerging Rotation
Jul 8 buy open watching Jul 17 buy 🔴 STOPPED/INVALIDATED — MDB $312.33, strong-down, death cross confirmed, now below all three MAs (-6.9%/-6.3%/-6.4% vs SMA20/50/200). The "above all 3 MAs" buy condition broke and the ~$320 50-day invalidation (now $333.4) was breached. n/a
ZM Zoom (ZM): The Anthropic-Stake Re-Rating vs a Flat-Revenue Cash Cow
Jul 8 watch open watching Jul 16 watch open in-zone
NTLA NTLA — watch above $18.50
Jul 9 watch open watching Jul 13 watch open invalidation-flagged
ABCL the great company shit stock turn abcl and ntla go on the board with triggers
Jul 9 watch open watching
CRDO The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 buy open watching Jul 9 buy open in-zone
MXL The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 watch open watching Jul 9 watch open in-zone
NTLA the great company shit stock turn abcl and ntla go on the board with triggers
Jul 9 watch open watching
ORCL MSFT: no signal. ORCL: falling knife.
Jul 9 pass open watching
AXTI The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 pass open watching
CIFR The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 pass open watching
CORZ The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 pass open watching
CRWV The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 pass open watching
EQT The Unwind Is Washing Out Our Picks-and-Shovels Lanes While Their Fundamentals Inflect Up
Jul 8 watch open watching
HUT The AI-Compute Landlords Are All On Sale — Only Two Balance Sheets Deserve the Dip
Jul 8 pass open watching
MU The Memory Supercycle Is Broader — and More Dangerous — Than the Narrative Admits
Jul 8 buy open watching
OKTA The Anti-AI Trade Is Turning: Cloud Software Is the Emerging Rotation
Jul 8 pass open watching
SNDK The Memory Supercycle Is Broader — and More Dangerous — Than the Narrative Admits
Jul 8 watch open watching
STX The Memory Supercycle Is Broader — and More Dangerous — Than the Narrative Admits
Jul 8 buy open watching
WDAY The Anti-AI Trade Is Turning: Cloud Software Is the Emerging Rotation
Jul 8 watch open watching
WDC The Memory Supercycle Is Broader — and More Dangerous — Than the Narrative Admits
Jul 8 buy open watching

What it's learned · Attribution mix

Attributions are human-labeled causes for notable wins or misses; total attributed rows: 140.

within-plan 69
research 58
execution 8
discipline 5

What it's learned · You vs desk

Followed and overrode counts come from the decision ledger; override win rate is the one-month score for decisions taken against advice.

_autotrade-v2-restricted-1785950079721-64322d45 1 followed 0 overrode override n/a
autotrade 1 followed 1 overrode override n/a
bench-signals 22 followed 3 overrode override n/a
bench-signals-trail 1 followed 1 overrode override n/a
buffer-trader 1 followed 0 overrode override n/a
claude-momentum 20 followed 1 overrode override n/a
claude-trader 12 followed 3 overrode override n/a
claude-trader-trail 2 followed 0 overrode override n/a
conviction 19 followed 10 overrode override n/a
degen 6 followed 1 overrode override n/a
gottheimer-trades 5 followed 0 overrode override n/a
mccaul-trades 3 followed 1 overrode override n/a
swing-kid 4 followed 0 overrode override n/a
wasserman-schultz-trades 2 followed 0 overrode override n/a
yolo 17 followed 0 overrode override n/a
yolo-insider-probe 0 followed 1 overrode override n/a
jan 1 followed 0 overrode override n/a
theo 2 followed 0 overrode override n/a

What it's learned · Live experiments

Each row is an A/B fork: one variant book versus its parent, with an out-of-sample readout date and a predeclared hypothesis.

Hypothesis
bench-signals-trail bench-signals Jul 8 Oct 8 trailing stop (15% give-back, armed at the old +20% cap) beats the fixed profit cap on the same entries
claude-trader-trail claude-trader Jul 8 Oct 8 trailing stop (15% give-back, armed at the old +30% cap) beats the fixed profit cap on the same entries