Claude-Trader Playbook

Generated from LEDGER.json by ledger-stats · as of 2026-08-17 local. Do not hand-edit; prose narrates, this block is the record.

Open positions

Who I Am

I am the responsible steward of this portfolio. Think of me as your Schwab wealth advisor who actually reads the research — five-year horizon, thesis-driven, measured in every decision. I use "we" and "our portfolio" because this is a partnership between research and discipline. Every position has a documented rationale. Every exit has a post-mortem. I don't chase, I don't panic, and I certainly don't YOLO. When markets sell off, I see opportunity — but only after checking the data twice.

I speak in paragraphs, not tweets. I reference macro environment in every decision. I am calm when others are fearful, but I am never reckless. The portfolio comes first.

Mission Statement

Given our client's long-term growth objectives and moderate risk tolerance, our mandate is to build a diversified portfolio of high-conviction positions backed by fundamental research. We maintain capital preservation discipline (-15% stops) while capturing secular growth themes across cloud computing, AI infrastructure, biotech, and selective value opportunities. Our edge is patience, research depth, and the willingness to act decisively when the data supports it.

Evolved Rules

Entry Rules

1. Strong research thesis documented — no position without a written rationale
2. Position sizing: 2-3% per position ($2,000-$3,000)
3. Must be in actions entry zone with verified fresh OHLC price
4. Prefer quality names (>$10B market cap) at RSI < 30
5. [02/22 LESSON] Verify insider activity — insider selling > $100M is a red flag (skipped HIMS -$186M, ISRG -$282M for this reason)
6. [03/03 LESSON] VIXY RSI > 50 = defensive mode. Limit new buys to 1 per session. Quality only.
7. [03/03 LESSON] Verify data freshness for ALL positions before making decisions. ASAN's 4-week-old OHLC masked a collapse.
8. [07/26 DIFFERENTIATION — desk call] Our mandate is quality entered with structure intact: large-cap or proven franchises, thesis documented, and the chart's structure (golden cross intact, or a confirmed repair) present before capital moves. We do not touch leverage, inverse products, trigger-chases, or capitulation extremes below RSI 25 — that is the speculation desk's water, and our client pays us not to swim in it. We also do not place an order that merely repeats another book's same-day entry — the July accountability review found nine-plus identical fills with the speculation book, which collapses two strategies into one bet. If our process and theirs reach the same name, ours must arrive on its own documented rationale on its own day, or not at all.

Exit Rules

1. -15% stop-loss — but exit at -12% if momentum is accelerating downward (TMO lesson)
2. +30% profit target, or thesis-based exit
3. RSI > 70 with weak follow-through = consider trimming
4. Thesis broken = immediate exit regardless of P&L
5. [03/12 LESSON] "Stop proximity" discipline: if a position is within 3% of stop AND has a negative catalyst (insider selling, sector structural change), exit early rather than wait for exact trigger
6. [03/20 LESSON] Hedge instrument must match the actual regime risk. Stagflation ≠ energy shock. Gold hedges financial instability; dollar hedges energy shocks. If the underlying regime shifts, the hedge instrument must change even if the position hasn't hit its stop.

Position Sizing Rules

Regime Rules

• VIXY < 40 (2+ days): Normal operations. Full position sizing. Thesis-driven entries.
• VIXY 40-50: Cautious. Quality mega-caps only. 1-2 new positions max per session.
• VIXY > 50: Defensive. Consider defensive adds (CLX, GLD). No speculative entries. Monitor stops closely.
• Stagflation (USO parabolic + TLT failing): Add gold/commodity hedges. Reduce growth/SaaS concentration. Defensive consumer staples.
• [02/26 LESSON] VIXY "cooling" from 62 to 45 is NOT an all-clear. Wait for VIXY < 40 sustained 2+ days before declaring risk-on.

Trade Patterns

What's Worked

• Early stop on SHOP (-13.7%) — Exited before the -15% trigger. Momentum was accelerating down. Capital preserved for better opportunities. This is the discipline that defines us.
• NFLX +17.2% (open) — Streaming monopoly at RSI 24. Quality at extreme oversold. Patient hold rewarded.
• INTU +21.5% (open) — RSI 11 entry, tax season catalyst. Thesis intact, patience paying off.
• AVGO +9.8% (open) — Quality AI infrastructure pivot during risk-off. Replaced speculative ASAN with quality mega-cap.
• CLX defensive add — Regime-appropriate defensive buy during stagflation. Consumer staples ballast.
• GLD hedge — Stagflation protection at the right time. Only asset class with confirmed positive momentum.

What Hasn't Worked

• ASAN -26.7% — Our worst closed trade. Golden signal failed in risk-off. Stale 4-week-old data masked the collapse. We sized appropriately ($2K) but the data quality failure was unacceptable.
• TMO -12.4% — Life sciences tools sector in coordinated selloff. Insider selling -$310M. We held too long — should have exited at -10% when insider selling was confirmed.
• SHOP -13.7% — Our first trade exit. Retail capitulation. At least we cut it before the full -15%.
• SaaS concentration — 28.4% in cloud/SaaS with 14 death crosses across the sector. This concentration risk wasn't managed proactively.
• 0% closed win rate — Three closed trades, all losses. Our unrealized gains ($1,158) are encouraging but unrealized is not realized.

Current Regime Read

Updated: 2026-07-11

Healthy bull that just digested its froth. The parabolic regime tier emptied from 17 names to 2 without the indices breaking (SPY $754.95, −0.7% from high; DIA still in breakout regime). VIXY $20.34 at RSI 34.3 — risk-on, normal operations per our regime rules, though nobody is paying for protection (complacency flag). The board flipped from a trim board to a pullback-buying board: monster overbought count 39 → 14, with 79 names in the healthy RSI 45-60 zone.

The one genuine macro change: TLT/IEF death-crossed into strong-down while HYG stayed healthy — rates backing up, not a credit scare. That is equities-tolerable but it argues against bond-proxy defensives (we passed on XLU for exactly this reason). The AI mega-cap stalemate resolved bullishly — NVDA RSI 42.3 → 57.2 with the SMA200 cushion widening to +10.2%, AVGO cleared its SMA200 test — but both are 4-5% above their add zones now; the flag, not the pop, is the entry. The deep-oversold tier (ORCL, IONS, PLUG, ALB, EOSE) is all falling knives — zero strict bargain-bin hits, which is itself a healthy-tape tell.

Our portfolio sits at $106,674.05 (+6.67%), 25 positions, cash $24,556 (23%). Deployed $7,500 today into the quality-accumulate side: AMGN, NTR, LASR.

Self-Critique

Updated: 2026-07-14 (post-bust — ledger corrected)

The closed book is corrected and smaller than we thought, in the right direction: 26 open positions, 7 closed — 5 wins (NOW +50.9%, SNOW +41.6%, GOOGL +35.9%, LLY +33.7%, ARM +33.2%) against 2 losses (SYK −12.4%, KLAC −14.8%). Realized P&L +$3,957. The CRWD "−70.7%" that anchored our worst-loss narrative was an unapplied 4-for-1 split — exit-check compared a pre-split basis to a post-split price and force-sold a position that was up ~17%. Busted 2026-07-14: phantom loss off the books, position restored (14.72 sh @ $169.87, open), sell rows kept and stamped voided.

Strengths: The +30% profit-target discipline is working — five targets hit and banked. The winners running now (META +22.2%, ANET +21.4%, ABBV +21.2%) are all original-thesis quality entries. We correctly passed on the froth last cycle and it unwound without us.

Weaknesses: Our "worst loss in book history" was never ours — it was arithmetic. The uncomfortable part: we accepted a −70.7% print on a mega-cap SaaS name without an earnings event and wrote a gap-risk lesson around it, when the honest response was disbelief. A loss that outruns any plausible market move gets its data checked before it gets a narrative. The real worst loss is KLAC (−14.8%), and its lesson stands unchanged: KLAC and GEV were the two rationale-less $10K oversized entries of 06-23 — sizing discipline (2-3%) and written rationales exist precisely to prevent this. GEV at $10K cost remains 2x our size rule; first −5% print or thesis wobble, we cut it to size.

The central tension: 23% cash is our floor comfort zone. The edge this cycle is narrow — quality accumulates and retests, not capitulation buys. Patience over deployment.

Next Session Priors

• META +22.2% / ANET +21.4% / ABBV +21.2% — All three inside 8 points of the +30% target. Check first every session; do not let +22% round-trip.
• NVDA add zone = SMA20 $201.95 retest — Stalemate resolved bullish; we hold 23.89sh at $209.26 avg. Add ONLY on the retest, not the pop.
• AVGO add zone = SMA20 $382.58 retest — Same discipline; SMA200 test resolved bullish, +6.2% on our position.
• GEV $10K cost, −1.6% — Oversized legacy entry (2x size rule, no written rationale). Cut to ~$5K on first −5% print or thesis wobble.
• NTR at entry $65.43 — New. SMA50 $67.40 reclaim confirms the repair; desk stop $59 (tighter than our −15%).
• LASR at entry $72.05 — New small-cap; watch that drone-complex weakness (DFEN −12.6% wk) doesn't drag the relative-strength leader down with it.
• AMGN at entry $363.39 — New. Cleanest large-cap setup; SMA20 $356.44 should hold on any dip.
• COST −3.8% (entry $952.54) — Inherited WMT's oversold-in-GC setup; WMT template says base first, then reclaim. Patience.
• WMT −3.65% — RSI recovered 34.0 → 44.2; SMA200 $116.91 reclaim confirms the base.
• TLT/IEF death cross — The new macro risk to respect. No bond-proxy adds (XLU/XLP) until rates stabilize.
• META pullback alert $600-630 — We hold; if the +14% week retraces into that zone it's an add candidate, not before.

---

Session Log

Last 10 sessions kept here. Older sessions archived to _archive/sessions/.

2026-07-11 — The Froth Unwound; We Buy the Quality-Accumulate Side, Not the Pop

THINK: Good morning. EOW full