💰 Practice money
$87,773
−$12,227 (-12.23%) since 2026-05-08 · Today: +$74.62
Locked in: −$13,984 · On paper (open): +$1,757
The things it looks at most when deciding what to bet on:
It does better in some market moods than others — and it knows it:
55 bets in the latest run, 12 of them experiments to learn new things.
So far its experiments beat the market 50 / 100 vs 45 / 100 for its confident bets — measuring that gap is how it learns what actually works.
It has made 163 trades so far — 52 still open and 111 finished. On the finished ones it has lost $13,984 for real; its open positions are +$1,757 on paper.
Showing the 30 most recent.
Exact sell prices settle at the broker, so closed trades show how they ended, not a per-trade dollar figure.
Every day it writes down what it is learning, then later checks whether it was right. So far it has confirmed 82 ideas that genuinely beat the market and flagged 210 that didn't pan out — and it is dialing those back.
It also caught itself 15488 times trying to trade stocks too illiquid to fill safely — its plan is to stop recommending names that keep failing that check.
The notes below are the bot's own raw jottings — technical, but it's all here if you want to dig in.
# Learnings Playbook
Curated standing trading-rules document, distilled by the Sunday weekend
routine from `memory/learnings.jsonl`. Last rewritten: 2026-08-16 (weekend
prep for week of 2026-08-18, CLEAN macro week post-CPI/PPI/Retail stack;
Jackson Hole 8/21–8/23 is the only vol driver). Group by category, drop
expired, merge duplicates, keep ≤200 lines.
## Process
- Always read `learnings.summary` (this file) and `learnings.recent` from
the bundle before writing. Side with a recent lesson over default unless
today's data clearly invalidates it (record regime change as new lesson).
- Append 1–3 lessons every run (wins AND losses). Selection bias toward
only-corrections drifts the bot from validated approaches.
- **COVERAGE REQUIREMENT** — every run issuing an options OR direct-short
pick MUST write at least one lesson covering that horizon / direction
with a falsifiable predicate.
- **Caveat staleness as FIRST visible block** when context is older than
the run date by 4+ hours.
- For pre-event runs (FOMC/CPI/NFP/GDP/PPI), explicitly defer NEW entries
to **post-event range confirmation**. Tight stops in the 30 min before
release get shredded by 1.5–2× ATR bars.
- **Pre-event breakout-conditional triggers**: gate on VIX/sector
confirmation through ALL multi-event windows; never chase-at-reference
pre-event.
- **CLEAN-week post-stack rule (added 2026-08-16)**: the trading week
immediately after a multi-event stack settles T+0 favors *pullback-entry*
(SMA20 + XLK green + rvol < 1.2× + RSI < 72) over *breakout-chase* from
Friday-close reference. Chase-behavior into a CLEAN week gets punished by
the late-week Jackson-Hole-style vol driver that always shows up.
- **Weekend brief checklist**: conditional triggers only, this-week-vs-
next-week macro split, gap-risk + 52w-extremes inventory, calibration
+ sizing rule in footer, non-Tech diversifier when concentration ≥ 80%.
- **Futures aggressive-mode override**: futures-* labels bias toward
action, not watch. ≥ 3 directional rows per action plan; ≤ 1 watch row;
direct medium shorts allowed on liquid futures (NQ/ES/YM/RTY/CL/GC) with
hard stop when at least one confirmation present.
- **Trader-signal noise filter**: the trader-live log spams the same
`TRV/MET/ZBH/OVV blocked at max_per_position/max_gross` lesson every 30s
during trading hours. Ignore these in the `learnings.recent` block —
operational not strategic. If a trader gate lesson recurs ≥ 20 times in
a window, remove the ticker from the advisor's recommendation list rather
than re-appending the lesson.
## Regime
- **VIX gates (binary, daily close)**: < 17.0 = deep full-unlock (also
requires IWM 1w lead vs SPY ≥ +1pp — see IWM gate); 17.0–17.5 = fragile
full-unlock (sizing cap 1/3 floor / 1/6 non-floor); 17.5–19.0 = dead-zone,
equity direct-short window LOCKED + long-bias default; > 19.0 = tactical
short ARMED via defined-risk put debit spreads on triple-confirm; > 22.0
= FULL tactical-short unlock (defined-risk only). Equity direct shorts at
conf ≥ medium PROHIBITED (calibration medium/short 37.5% n=32 AND
low/short 34.7% n=98 both below 40% floor).
- **Current regime 2026-08-16**: VIX 14.25 (-4.36% 1w deep-unlock ceiling),
SPY 1w +0.40%, QQQ +1.11%, IWM +1.17%, DXY 99.67 (-0.29% 1d, +0.07% 1w
flat). Regime = **risk_on, posture stand_aside, FRAGILE-full-unlock** —
VIX would unlock deep-full BUT IWM lead gap +0.77pp still below +1pp
threshold. **2nd consecutive week of IWM lead ≥ +0.5pp** (0.55pp → 0.77pp)
= provisional non-floor cap relaxation to 1/4 permitted (still not full).
Sizing cap 1/3 floor / 1/6 non-floor holds through CLEAN week 8/17–8/22.
- **IWM lead-gap gate** (validated 2026-08-09, 6th sighting 2026-08-16):
deep-full-unlock requires VIX < 17 AND (IWM 1w % − SPY 1w %) ≥ +1pp.
6-week trajectory: -0.5 / -1.0 / -0.9 / -1.09 / +0.05 / +0.55 / +0.77 —
consistently repairing but not yet clearing +1pp. **2-consecutive-week
rule**: two weeks of ≥ +0.5pp lead confirms directional trend and permits
provisional non-floor cap relaxation to 1/4 (not full-unlock). Retested
every weekend brief.
- **Multi-event stack settled 2026-08-14 T+0**. Week 8/17–8/22 is CLEAN.
Only vol driver = **Jackson Hole 8/21–8/23** (Powell speech usually Fri).
Trim aggressive-conviction entries into Thu 8/21 close if unhedged. If
VIX weekly close > 19 Fri 8/22, tactical-short-armed regime activates.
- **Next multi-event stack begins week of 9/1**: NFP 9/4 + CPI 9/10 + PPI
9/11 + Retail 9/15. Pre-NFP T-3 sizing pivot activates Fri 8/29 EOD;
T-1 tightening after Wed 9/3 14:00 ET.
- **VIX one-day-spike bounce predicate** (validated 2026-06-18, 2026-06-21):
when VIX close jumps ≥ 8% in one session into 17.5–19.0 dead-zone WITHOUT
a new catalyst emerging in the following week, next 3–5 sessions
structurally favor VIX crush back < 17.5. MYM long is the preferred
expression (YM ticker_memory 53% n=87 highest floor). SUSPENDED when a
new catalyst emerges within 5 sessions.
- **Post-FOMC/NFP T+1 chop 4/4 stop-out rule**: when post-event T+1 + slate
near 52w highs, trigger placement ≥ session high + 1.25% MINIMUM + second
5min K close + 1.5× volume; stop = SMA20 buffer NOT 0.5×ATR tight.
Applies to any post-print breakout re-entry.
- **Attribution 14d STABLE 2026-08-16**: 61.5% n=26. Mean_return +0.77%.
Best/worst both DDOG (+11.4% recovery pick / -9.8% earlier fade) = the
[[post-earnings-soft-forward-guide-fade]] rule now validated BOTH
directions (fade the initial, recover on weekly reclaim). Long_bias 66.7%
n=24 clean. Full relaxation still requires ≥ 5 sessions ≥ 60% hit-rate
AND IWM lead gap ≥ +1pp AND VIX < 17 (only 2 of 3 met currently: hit-rate
ok, VIX ok, IWM gap fail).
## Sector
- Cite top 2 / bottom 2 sectors from `sector_leadership` in exec summary —
rotation surfaces here.
- **Current rotation (2026-08-16)**: **XLE +7.67% top-1** (crude USO +7.31%
+ refiner-parabolic tailwind — but refiner basket HARD AVOID, non-refiner
E&P sanctioned expression), XLU +1.61% #2 (defensive), XLC +1.59% #3,
XLP +1.14% #4, XLK +1.09% #5 (from top-1 last week +7.20%, slowing);
XLY -1.39% bottom-1 (2nd consecutive week Consumer weakness), XLB -0.61%
#10.
- **1w-rotation-flip instability rule** (added 2026-08-02, refined
2026-08-09, EXCEEDED 2026-08-16): when a sector's 1w % swings > 8pp
week-over-week, treat current rotation as unstable — require 2nd week of
confirmation before shifting overweight allocation. **XLE +11.11pp swing
this week** (from -3.44% bottom-1 last week to +7.67% top-1) EXCEEDS the
trigger — do NOT tilt overweight to Energy until confirmed 8/22.
- **Defensive triple warning fired 2026-08-16**: XLE + XLU top-2 AND
XLY bottom-1 = late-cycle rotation warning per Aug 14 midday-issued
lesson. Downgrade all MEDIUM-conf tech-long single-name picks to
HOLD-only; drop DAY_TRADE tech breakout triggers to defer; elevate XLF
sector-diversifier options as the sanctioned expression. Do not chase XLE
(refiner HARD AVOID) or XLU (yield-sensitive noise).
- **Sector-ETF diversifier XLF sleeve** (validated 14 consecutive weeks
through 2026-08-16): XLF as the validated non-Tech ETF sleeve when
Financials 1w positive AND +3.5% 1m intact (14-week streak permits
provisional floor relaxation to 3.0%). Aug 22 58/60 option-debit-spread
is the sanctioned expression through the CLEAN week + Jackson Hole.
- **Validated non-Tech diversifier roster**: VRT, AKAM, NXT, GNRC, LNTH,
AA, CIEN, DY, XLF, XLI, XLV, NVT. AI-DC infrastructure non-Tech triple =
DY + GNRC + CIEN when concentration > 80% AND XLI top-5 1w.
- **Sector concentration escalation**: when top sector share ≥ 85%, review
MUST cite non-sector diversifier picks; ETF expression preferred over
single name when ≥ 5 single-name diversifier candidates fire parabolic
predicate. Current top-25 sector share **56% Technology / 12% Consumer
Cyclical + Energy tied / 8% Communication + Healthcare** — under 85%
escalation this week.
- **Non-refiner Energy exposure rule**: when XLE top-3 1w AND refiner
basket in parabolic-5-way HARD AVOID, express XLE-broad via SLB
(services) or OVV (E&P) at low conf, NOT the refiner basket. Currently
active: XLE +7.67% top-1 + refiner 11-week HARD AVOID = OVV/SLB is the
only sanctioned Energy expression (still LOW conf per XLE rotation
instability).
## Setup
- **High-volume breakouts** (> 1.5× 20d avg) with RSI < 70 follow through
better than RSI-extended breakouts.
- **RSI sweet spot 55–72**: cleanest entry filter when parabolic basket
fires simultaneously elsewhere. NBIS RSI 67 this week's cleanest sweet-
spot example within top-25.
- **Ticker-memory ≥ 77% priority** (highest-signal single feature):
CRL 100% n=7 (perfect record top-of-universe), OKTA 94.12% n=17, CRWD
91.67% n=12 (ELITE tier), QLYS 83.33% n=12 (but cluster-monitor binding),
ZBRA 80% n=5 (but cluster core avoid), NTAP 75% n=4 (but cluster core
avoid). Use ticker-memory rank as tie-breaker between otherwise similar
setups; downgrade when cluster-monitor overrides.
- **Mean-reversion needs trigger + level + catalyst**, not RSI alone.
Recipe: RSI < 40 + price 0–5% ABOVE 50-day SMA (reclaim) + fundamental
catalyst (Q2 beat / raised outlook / regulatory approval).
- **Post-earnings soft-forward-guide fade** (BOTH directions validated
2026-08-16 via DDOG recovery): when software/AI-adjacent Q2 print
delivers headline beat BUT forward-guidance framing raises AI-monetization
questions AND stock opens down > 8% AND sector-peer trades same-direction,
DOWNGRADE the swing pick to WATCH ONLY regardless of pre-print thesis
strength. Wait for pullback to SMA20 + XLK stability AND next-day close
reclaim before re-engaging. DDOG cycle: -19% fade (Q2 8/6) → +11.4%
recovery (10d later) = confirms both the fade AND the recovery legs.
- **Storage / memory / refiner parabolic predicate** = HARD avoid filter:
RSI > 75 + sector-relative-momentum 1m > 40% (or 3m > 80%) + analyst
target < spot. Tactical shorts ONLY on triple-confirm (5d lower-high +
VIX > 19 + RTY new low).
- **10-way parabolic cluster HARD AVOID escalation** (validated 2026-08-16,
9th consecutive week): NTAP/VEEV/ZBRA/SN/QLYS/HRB/HALO/CRL + Everpure(P?)
+ PSX-adjacent simultaneously firing RSI > 73 + 52w-high proximity +
analyst upside NEGATIVE. When cluster ≥ 10 AND crosses ≥ 5 sectors, HARD
AVOID escalates from cluster-only to 'trim adjacent longs same sectors
AND downgrade single-name Tech/Healthcare to defined-risk-only for
remainder of the week'. Cluster is currently at 10+ names across Tech +
Healthcare + Materials + Industrials + Consumer Cyclical + Comm Services.
- **Refiner parabolic 5-way** (validated 2026-07-19, 11-consecutive-week
validation 2026-08-16): PBF/DINO/MPC/VLO/PSX simultaneously firing RSI
60–76 + 52w-high T0-3 + analyst upside NEGATIVE + WSJ record-margins
cycle-peak print. **NOW with USO +7.31% 1w crude tailwind amplifying**.
HARD AVOID posture unit. Direct shorts prohibited. Wait signal: ≥ 10%
pullback + RSI < 60 + weekly close reclaims prior swing high with volume.
- **Semi-contagion cleared (unchanged)**: STX 7/28 AMC + AMD post-FOMC
crush + MU sympathy = XLK 2-green gate cleared through 7/31, held through
week ending 8/14. If any Aug semi (INTC/NXPI/ON) guides down Mon–Fri
8/17-8/22, RE-ARM the +2 XLK green session gate before adding to Tech.
**NVDA Aug 27 earnings** is next semi catalyst; pre-print positioning
starts Fri 8/22 AMC.
- **Four-fold / five-fold resonance high**: today + 5d + 20d + 52w highs
collapsing to one price = reaction zone, not chase zone. Trim existing
longs into the level; continuation only on second-day 5min close above +
1.25× volume.
- **NQ/ES ratio z20 mean-reversion window**: when nq_es_ratio_z20 ≤ -1.5,
NQ mean-reversion long window opens. When z ≤ -1.75, window is fragile —
require strict 5min confirmation + XLK 1d not breaking further -1%.
- **CL structural short thesis FAILED post-NFP T=0 8/7**: CL bounced through
8/6 morning short stops. Rule: pre-print short setups on CL (mixed_trend
+ RSI 45-50 boundary) MUST be closed/downgraded to WATCH before the
print — T=0 event chop DEFAULT is a 1.5-2× ATR reversal in whichever
direction removes the crowd's stops. Rule refreshed 2026-08-14 midday:
attribution 14 days shows CL 11/11 stop-out streak — HARD AVOID any CL
entries without 3-session structural rebreak.
- **GC persistent-extreme-rvol distribution predicate 6-day validation**
(refined 2026-08-07): when GC rvol ≥ 10x for 6+ consecutive sessions AND
session_range/ATR ≥ 1.5 on the extreme day, next-session pullback
probability rises to ≥ 65%. **BUT if DXY < 100 AND weakening, distribution
predicate suspended and the extreme rvol is a base-formation-for-extension
signal instead** (validated by GLD +7.25% during DXY 99.60 weakening).
Current: GLD +0.76% 1w cooling, DXY 99.67 flat — extension mode
neutralizing.
- **Friday-weekend forced-trim discipline for futures BINDING** when: (a)
Fri is a T=0 event print day OR (b) next week is multi-event stack.
ALL Friday-midday new entries MUST include an explicit '15:30 ET trim
2/3, remainder only YM or NQ six-contract highest floor' clause; GC/CL
always full-flat before 15:30 ET regardless of P&L.
- **CLEAN-week post-stack pullback-entry preference** (added 2026-08-16):
after a multi-event stack (CPI+PPI+Retail or FOMC+NFP+CPI) settles T+0,
the following clean-week best long-entry approach is pullback-to-SMA20 +
XLK green + rvol < 1.2× + RSI < 72, NOT breakout-chase from Friday-close
reference. Chase-behavior into a clean-macro week gets punished by
Jackson-Hole-style late-week vol.
## Risk
- Every pick MUST include confidence (Low/Medium/High) and ≥ 1 invalidation
level. No exceptions.
- **Post-stack CLEAN-week sizing (2026-08-16)**: 1/3 floor / 1/6 non-floor
sizing HOLDS through the clean week 8/17-8/22. Provisional 1/4 non-floor
relaxation permitted on best-signal single names (ticker-memory ≥ 80% +
RSI 55-72 sweet-spot + XLK green) per 2-consecutive-week IWM lead ≥
+0.5pp rule.
- **Jackson Hole 8/21-8/23 event risk**: any Powell-hawkish signal spikes
VIX out of deep-unlock ceiling. Trim aggressive entries into Thu 8/21
close if unhedged. If VIX weekly close > 19 Fri 8/22, tactical-short-armed
regime activates.
- **Trim-winners operational rule**: at +0.4×ATR take 1/3, +0.7×ATR take
2/3, last 1/3 trail-stop.
- **Equity direct shorts prohibited at conf ≥ medium (BINDING)**:
calibration medium/short 37.5% n=32 AND low/short 34.7% n=98 both below
40% floor. Short bias expressed ONLY as defined-risk debit spreads gated
on VIX > 19 daily close + triple-confirm. Futures aggressive-mode
overrides this for liquid contracts.
- **HIGH/long calibration floor-fail BINDING 2026-08-16**: high/long 33.3%
n=33 (below 40% floor). Do NOT stack size into any "high-conviction"
single-name long. H/M-conviction net-long expression sanctioned = XLF
options debit spread only.
- **No persistent unhedged hold across NFP/CPI/PPI/Retail**: T-2 close all
non-defined-risk positions with expiry inside the print week. Print day
= no new entries until 09:45 ET after print.
- **3-day-weekend / federal holiday gap risk**: previous-day entries face
multi-day cash-tape gap. Use defined-risk options (debit spreads cap
downside at premium paid).
## Attribution
- 14d equity attribution 61.54% n=26 STABLE (2026-08-16 bundle) — settled
from prior 60.32% n=63. Mean_return +0.77%. Best DDOG +11.4% (10d,
post-earnings-fade recovery); worst DDOG -9.8% (11d, earlier fade of
the same name — Q2 8/5 print). Long_bias 66.7% n=24 clean. Maintain
1/3 floor / 1/6 non-floor fragile-unlock sizing.
- 14d futures attribution: prior 7/26 reading 87.5% n=18 (medium 77.8% /
low 100%) GOOD. Continue baseline sizing.
- **CL attribution FAILED 2026-08-14**: 14 days 11/11 stop-out streak.
HARD AVOID any CL entries without 3-session structural rebreak.
## Calibration discipline
- **Monotonicity NON-monotonic 2026-08-16** (90d n=2253, low 43.5% /
medium 43.0% / high 52.83%). Brier 0.2534 (still worse than coin-flip
0.25). HIGH tier is only bucket materially above 50%. Confidence LOW
and MEDIUM barely separable — the useful separator is `high` vs
`not-high`. Merge candidate remains open. Monotonicity was BRIEFLY
restored 2026-08-14 (low 43.44 / med 42.99 / high 53.35) but broke again
by 2026-08-16.
- **Direction warnings (BINDING)**:
- HIGH/LONG 33.3% n=33 (below 40% floor). Do NOT stack size into
"high-conviction" single-name long.
- LOW/SHORT 34.7% n=98 (below 40% floor). Direct short prohibited.
- MEDIUM/SHORT 37.5% n=32 (below 40% floor). Defined-risk options
wraps only.
- **Options bucket calibration by conf**: high/options 50% n=2 (small n),
medium/options 52.63% n=38 (sanctioned bucket, mean_excess_5d +2.32%),
low/options 39.56% n=182 (mean_excess_5d -6.02% pooled — auto-withhold
binding BUT overridden by XLF 14-week diversifier sleeve). Options
entries require triple-confirm at conf ≥ medium AND defined-risk only
through catalyst windows.
- **Issue picks as conditional 'watch'** when they fail triple-confirmation
rather than 'buy/short' at reference. Pre-event breakout-conditional
triggers preferred structure over chase-at-reference.
## Pattern playbook
- **Dual-trigger event-day brief**: two mutually-exclusive conditional
orders, dovish-long + hawkish-short (short leg = defined-risk put debit
spread only).
- **Pre-event futures-midday brief** (≥ 1h before release): defer NEW
entries to post-event range confirmation. Manage existing only.
- **Defined-risk options across catalyst-stack windows**: WATCH-ONLY
through catalyst windows (options bucket auto-withhold binding —
mean_excess_5d -6.02% n=182 low/options bucket).
- **Single-binary-catalyst pre-event positioning**: equity entries
DEFERRED to post-print confirmation. Pre-print: defined-risk debit
spread only at LOW WATCH, cost cap 33% of spread width.
- **Weekend prep with sector-flip risk overlay** (refined 2026-08-16):
when a sector's 1w reverses > 8pp from prior-week ranking, mark
current-week rotation as UNSTABLE — require 2nd week of confirmation
before shifting overweight allocation. XLE +11.11pp swing this week is
fresh-fired instability trigger — Energy overweight NOT confirmed until
8/22 weekly close shows XLE top-5.
- **Post-parabolic no-chase-at-reference** (validated 2026-08-07 midday
TWLO +28%, ABNB +9%, extended 2026-08-14 HRB post-Q4-blowout): when a
Q2-beat print drives session-gain > +8% AND stock at/near 52w-high AND
RSI > 65, DO NOT chase at close reference. Wait for pullback to SMA20
zone + 5min stabilize + sector ETF green + rvol < 1.5x. Existing longs
trim 1/3 at reference + 0.5×ATR. HRB, CRL, ABNB, SN currently in this
bucket.
- **Post-NFP/CPI/PPI T+1 chop widen-stops rule**: when post-print T+1 +
slate near 52w highs, breakout stop = SMA20 buffer NOT 0.5×ATR tight.
4/5 breakouts stopped-out under tight ATR stops = validated recipe.
---
This file is rewritten by the Sunday weekend routine after distilling the
recent `memory/learnings.jsonl` tail. Keep ≤200 lines, group by category,
drop expired entries, merge duplicates, demote what recent runs have
invalidated.
# Morning Review — 2026-08-20 **Regime:** neutral / posture stand_aside. **VIX 15.19** deep-unlock ceiling (BUT +3.83% 1w drift up + Jackson Hole 8/21–8/23 vol driver ahead). **XLV +4.30% 1w top-1** for a second consecutive week + XLK -2.76% bottom-1 = defensive rotation confirmed, downgrade tech-long single names to HOLD-only. **DXY 98.73 (-1.23% 1w) breaking below 99** = GLD +3.84% 1d extension mode active. **Attribution 14d has slumped to 40% n=15** (from 61.5% n=26 last read) — AMD -7.82% (post-semi contagion), MSFT +0.82% + XLF -0.17% + CRL +9.22% top. Sizing cap 1/3 floor / 1/6 non-floor binding through the CLEAN week + Jackson Hole. ## Today's Action Plan | # | Ticker | Action | Trigger | Confidence | Stop | |---|--------|--------|---------|------------|------| | 1 | **MRNA** | AVOID | RSI 93.9 EXTREME + T-0 52w-high + analyst upside -70.8% + +177% cancer-vaccine parabolic. NO chase, NO direct short (calibration LOW/short 35.6% floor-fail) | High | n/a | | 2 | **Cluster (BDX/RGEN/VEEV/A/DXCM/CRL/HALO/MRK/HAE/P/NTAP)** | AVOID | Parabolic cluster week 10 escalation — 11 healthcare + tech names RSI 71–89 at 52w extremes with negative analyst upside. Trim adjacent longs same sectors | High | n/a | | 3 | **NEM** | Watch (add) | Weekly close > $128 on 1.25× vol + DXY < 99.00 hold + GLD > $410. Small starter only (1/6 non-floor sizing); XLV/gold defensive rotation confirmed | Medium | $118 | | 4 | **TGT** | Watch | Reclaim >$162 SMA20 zone on 1.25× vol + hold 2 sessions. NO chase pre-print-reclaim (Q2 raised guidance released 8/19, initial fade classic post-earnings soft-guide setup) | Low | $155 | | 5 | **MRVL** | Watch | Pullback to $220 SMA20 + XLK green + 5min stabilize + RSI < 70. NO chase at $237+ per CLEAN-week pullback rule; Google $12.2B TPU deal creates AI-monetization tailwind but semi-contagion overhang unresolved | Low | $208 | ## Executive Summary - **Defensive rotation confirmed week 2**: XLV +4.30% 1w top-1 (biotech breakthrough + drug-manufacturer beats driving XLV rank), XLE +4.18% #2, XLP +1.72% #3; XLK -2.76% bottom-1 (semi contagion continues — AMD lost -7.82% 4d after our 8/16 pick). Late-cycle defensive triple: XLE + XLV top-2 AND XLY -1.39% recent-bottom = tech-long single-name conviction is capped this week. - **MRNA is not a repeatable setup**. +177% on melanoma vaccine Phase 3 primary endpoint met is a single-binary catalyst, not a screen signal. RSI 93.91 top of universe. The trader-advisor already tested and validated the extreme-RSI biotech mean-reversion predicate (36 samples, -0.54% edge vs SPY 5d) but the pull is modest and shorts prohibited at conf ≥ medium — best action is stand aside, not fade. - **Gold structural breakout arming**: DXY 98.73 broke below 99 with 1w -1.23% weakness; GLD +3.84% 1d; NEM top-of-mining-ranks +5.88% analyst upside room. Non-Tech diversifier expression as XLV faded from XLF (XLF 1w -0.76% BREAKS the 14-week diversifier gate; sleeve suspended this run). - **Attribution slump requires posture check**: 14d hit-rate 40% n=15 (from 61.5% n=26 last read) is now BELOW the 60% relaxation floor. This blocks any progression to full-unlock even if VIX/IWM-lead conditions were to align. Confidence discipline binding: HIGH/LONG 25% n=28 = do NOT stack size into "high-conviction" tech long. - **Calendar map**: this week is CLEAN post-CPI/PPI/Retail settled 8/14. Only vol driver = **Jackson Hole 8/21–8/23** (Powell speech typically Fri). Next multi-event stack begins 9/1 (NFP 9/4 + CPI 9/10 + PPI 9/11 + Retail 9/15 + FOMC 9/16). T-3 sizing pivot activates Fri 8/29 EOD. ## Swing Setups **NEM — Newmont Corp. — Watch/add (Medium, LONG)** - Reference $125.08 (+30% 1m, RSI 79.6, 140d off 52w-high — recovery leg, not chase). - Trigger: weekly close > $128 on 1.25× vol AND DXY < 99.00 hold AND GLD > $410. Ticker not in memory (no prior signal) but sector rotation + gold breakout structure clean. - Stop: $118 (below prior swing low, ~5.7% invalidation). - Target: $138 (T1 first resistance ~10%), $148 (T2 = pre-2023 highs). - Thesis: DXY structural breakdown below 99 + gold extension mode 6-day rvol persistence + Newmont Q2 beat 4/4 + insider net -6k (immaterial) + analyst upside +5.9%. Defensive expression during XLV top / XLK bottom rotation. - Sizing: 1/6 non-floor cap (not floor ticker), starter only into $128 breakout. Trim 1/3 at +0.4×ATR ($131.5), 2/3 at +0.7×ATR ($134). - Invalidation: DXY close > 100.00 for 2 sessions, OR GLD close < $400, OR NEM daily close < $122 SMA20 zone. **HL — Hecla Mining — Watch (Low, LONG)** - Reference $20.54 (Treasury buyback plan silver rally, +34% 1m, RSI 75.3). - Trigger: pullback to $19.00 SMA20 zone + weekly close reclaim of $20.50 on 1.5× vol. - Stop: $17.50 (below SMA50 zone). - Target: $23.00 (T1 = analyst target mean $23.53). - Thesis: HL/NEM 0.90 pairwise correlation = redundant to NEM — pick ONE, prefer NEM (higher analyst confidence, larger cap, more liquid). - Sizing: If NEM not filled, HL as alternative 1/6 non-floor. NOT additive. **TGT — Target — Watch (Low, LONG)** - Reference $159.00 (Q2 print 8/19: raised guidance released AM, initial fade sub $152 then intra-day reclaim). - Trigger: 2-session hold above $162 SMA20 reclaim on 1.25× vol (per [[post-earnings-soft-forward-guide-fade]] rule — wait for weekly reclaim before re-engaging). - Stop: $155 (below Wed intra-day low). - Target: $175 (T1 = analyst target high). - Thesis: XLP +1.72% 1w defensive rotation + earnings beat + raised guidance framing. Ticker-memory not available. WMT print 8/20 = tape risk overhang. - Sizing: 1/6 non-floor starter only. ## Long-Term Ideas **LLY — Eli Lilly — Watch (Low, LONG)** - Reference $1,280.34 (52w high, RSI 68.97, XLV top-sector). Beat 4/4 last 4 quarters (+27.3% last surprise accelerating). - Trigger: consolidation to $1,240 zone (SMA20) + weekly close > $1,285 on volume. - Stop: $1,180 (2 ATR). - Target: $1,371 (T1 = analyst median). - Thesis: Multi-quarter GLP-1 leadership + Q2 blowout + XLV defensive rotation. Ticker-memory 33.3% n=12 mean_excess_5d -2.07% is a HEADWIND — the bot's historical LLY picks have underperformed. Downgrade to LOW conf and treat as long-term structural exposure only, not swing entry. - Sizing: 1/6 non-floor cap. **MRVL — Marvell Technology — Watch (Low, LONG — pullback-only)** - Reference $237.27 (Q2 print upcoming 8/28 AMC. Google $12.2B TPU warrant deal announced. RSI 71.2 stretched. XLK 1d -1.07% contagion). - Trigger: pullback to $220 SMA20 zone + XLK 5min stable + RSI < 68 (per CLEAN-week rule). - Stop: $208 (below prior swing low). - Target: $260 (T1 analyst mean $260). - Thesis: Ticker-memory 56.4% n=39 mean_excess_5d +4.59% is the elite tier — validated bull sleeve. Google TPU deal 'greatly expands' long-term AI opportunity per analyst. AMD sympathy risk unresolved. Q2 print 8/28 = defined-risk only through print. - Sizing: 1/6 non-floor starter only on pullback. Pre-print (Fri 8/22 close) reduce to 1/3 of full size if held. ## Day-Trade Watchlist - **MRVL breakout** (only if pullback fills): break > $240 on 5min close + 1.25× vol + second bar hold + XLK green. First target $246, day-trade stop $234. - **NEM continuation**: break > $128 with GLD > $412 confirmation + DXY < 99.00. First target $130 (round-number), day-trade stop $124. - **TGT reclaim**: break > $162 on 1.25× vol + hold 15min. First target $166, day-trade stop $158. Skip if WMT prints negative gap-down before 09:30. - **HARD avoid intraday**: MRNA, HAE, RGEN, MRK, CRL, A, HALO, VEEV, BDX, DXCM, P, NTAP — parabolic-cluster names getting whipsaws in either direction. ## Options Ideas **No new option positions this run.** Rationale: - XLF diversifier sleeve **suspended**: XLF 1w -0.76% breaks the "Financials 1w positive" 14-week gate. Aug 22 spread cannot be re-issued. - Options bucket calibration: LOW/options 39.8% n=181 mean_excess_5d -6.02% (auto-withhold binding); MEDIUM/options 52.6% n=38 sanctioned but no clean medium-conf setup today. Healthcare parabolic cluster is not a sanctioned defined-risk long expression; direct-short calibration prohibits put debit spreads at conf ≥ medium. - Wait signal for new options: post-Jackson Hole 8/22 close if VIX remains < 17 AND XLF 1w reverses positive (re-arm XLF Aug 29 or Sep 5 debit spread), OR MRVL post-print 8/29 for defined-risk cybersecurity-cohort read-through. - Existing XLF Aug 22 58/60 debit spread (issued 8/19 morning): now stops closer to expiry with XLF -0.62% 1d cutting into extrinsic; if paid ≤ $0.60 and XLF spot 57.48, position is negative delta at expiry — manage via close-at-market during 09:45–10:15 ET window if XLF stays below $58.00. ## Risks / What Would Invalidate - **Jackson Hole hawkish (Fri 8/22)**: Powell signaling delayed cuts = VIX weekly close > 19 arms tactical-short regime (defined-risk puts on parabolic-cluster only). Trim all aggressive-conviction entries into Thu 8/21 close if unhedged. - **Semi contagion escalation**: AMD -7.82% 4d = validated warning shot. Any NVDA-negative headline before Aug 27 print re-arms the +2 XLK green-session gate for tech longs. - **Attribution slump not stabilizing**: 40% n=15 hit-rate — if next 5 sessions do not repair to ≥50%, downgrade sizing cap to 1/4 floor / 1/8 non-floor next weekend brief. - **DXY reversal**: DXY close > 100.00 for 2 sessions kills the gold-extension thesis (NEM/HL invalidation). - **MRNA cluster fade**: if MRNA closes < $120 (30% pullback) in the next 5 sessions, the entire biotech mean-reversion predicate ARMS for the wider healthcare cluster (VEEV/CRL/BDX/A/DXCM = adjacent picks). Do NOT initiate defined-risk shorts unless VIX also crosses > 19. - **Coverage gap**: no direct shorts issued today (calibration binding); if healthcare cluster fades > 5% into next 3 sessions, the [[parabolic-rsi-cluster-hard-avoid-10way]] rule needs post-fade re-scoring. --- Disclaimer: This newsletter is for educational and informational purposes only. It is not financial advice, not a solicitation to buy or sell any security, and reflects algorithmic screens and model-written commentary that may be wrong. Do your own research. Past performance does not guarantee future results.
Armed to trade 6 stocks (plan from 2026-08-20): A, DXCM, LLY, MRNA (betting it falls), MRVL, NTAP. It only buys if a price trigger is hit.
The things it looks at most when deciding what to bet on:
The robot hasn't settled on what matters yet.
It hasn't changed its mind much yet.
It does better in some market moods than others — and it knows it:
Not enough days watched yet.
30 bets in the latest run, 6 of them experiments to learn new things.
So far its experiments beat the market 48 / 100 vs 39 / 100 for its confident bets — measuring that gap is how it learns what actually works.
It has made 176 trades so far — 96 still open and 80 finished. On the finished ones it has lost $8,421 for real; its open positions are +$2,790 on paper.
Showing the 30 most recent.
Exact sell prices settle at the broker, so closed trades show how they ended, not a per-trade dollar figure.
Every day it writes down what it is learning, then later checks whether it was right. So far it has confirmed 4 ideas that genuinely beat the market and flagged 3 that didn't pan out — and it is dialing those back.
The notes below are the bot's own raw jottings — technical, but it's all here if you want to dig in.
Armed to trade 6 stocks (plan from 2026-08-20): MRVL, NBIS, LITE, AAPL, XEL, PEP. It only buys if a price trigger is hit.
The things it looks at most when deciding what to bet on:
The robot hasn't settled on what matters yet.
It hasn't changed its mind much yet.
It does better in some market moods than others — and it knows it:
Not enough days watched yet.
22 bets in the latest run, 5 of them experiments to learn new things.
So far its experiments beat the market 46 / 100 vs 43 / 100 for its confident bets — measuring that gap is how it learns what actually works.
It has made 216 trades so far — 116 still open and 100 finished. On the finished ones it has lost $1,171 for real; its open positions are +$3,607 on paper.
Showing the 30 most recent.
Exact sell prices settle at the broker, so closed trades show how they ended, not a per-trade dollar figure.
Every day it writes down what it is learning, then later checks whether it was right. So far it has confirmed 4 ideas that genuinely beat the market and flagged 4 that didn't pan out — and it is dialing those back.
It also caught itself 1092 times trying to trade stocks too illiquid to fill safely — its plan is to stop recommending names that keep failing that check.
The notes below are the bot's own raw jottings — technical, but it's all here if you want to dig in.
Armed to trade 6 stocks (plan from 2026-08-20): MRNA (betting it falls), REGN (betting it falls), ABNB (betting it falls), TGT (betting it falls), LLY, NEM. It only buys if a price trigger is hit.
The things it looks at most when deciding what to bet on:
The robot hasn't settled on what matters yet.
It hasn't changed its mind much yet.
It does better in some market moods than others — and it knows it:
Not enough days watched yet.
0 bets in the latest run, 0 of them experiments to learn new things.
No bets placed in the latest run.
Not holding any stocks right now — waiting for one of its triggers to hit.
It hasn't placed any trades yet.
No open trades right now.
No closed trades yet.
Every day it writes down what it is learning, then later checks whether it was right. So far it has confirmed 0 ideas that genuinely beat the market and flagged 0 that didn't pan out — and it is dialing those back.
The notes below are the bot's own raw jottings — technical, but it's all here if you want to dig in.
Nothing logged yet.
Nothing logged yet.
Nothing armed to trade at the moment.