James Three Plays (Pull-In / Trend / Backside) on the SIP watchlist · 10 trading days ending 2026-06-26 · 5-min bars · generated 2026-06-27 ET
Verdict — where the edge is (and isn't)
- EDGE: Trend Trade (Dale 1st-dev pullback) — the only profitable play: $361, 59.6% win, +0.078R over 47 trades; best at grade B (+0.383R, 75% win). Building Trend as the Dale 1st-deviation model (not the old 9-EMA trigger) was the right call.
- BLEED: VWAP Pull-In over-fires in ROTATION — 94 of 106 pull-ins fired in
ROTATION regime for $-2,429 (-0.257R). In COMPRESSION / TREND regimes the same play is small but positive. Action: gate Pull-In on VCET regime (require not-ROTATION / expansion_score>0) — this is the VCET expansion-gate thesis confirmed on live data.
- UNPROVEN: Backside Scalp — only 4 trades, -1.669R, with runaway MAE. Too few to trust and the structural stop ran wide; needs a regime gate + tighter stop before it's even a fair test.
Edge gate for hotkeys (Stage 4): only positive-expectancy cells (avg_R>0, with enough n) advance. Today that's Trend / grade B–C / TREND_UP. Pull-In advances only once regime-gated. Backside does not advance.
Symbols: CALY, SOXL, LLY, CASY, VMI, BH, AYI, BE, PRAX, FTAI, SQQQ. Bars: real 1-min from karb/kti, aggregated to 5-min. Regime tag: VCET v1.1 engine (
vcet_state) per session. Plays mirror
J3P_v2.2_SIP_Webhook.pine: Pull-In = VWAP recapture after a shakeout (LB3 stop, +2σ/+4σ ladder, 80/20); Trend = Dale 1st-deviation pullback in an uptrend (stop = reaction point); Backside = ≥2·ATR extension above VWAP + rejection (stop above failed high, cover into VWAP). Entries fill next-bar open with conservative slippage (≥1¢ or 0.10×bar range); stops add adverse slip; 0.5¢/sh commission both sides.
Constant $100 risk/trade sizing so an $18 name and a $248 leveraged ETF are comparable — total P&L therefore reads as "if you'd risked $100 per trade." R = net P&L ÷ initial risk. Entries gated to RTH 09:30–15:55 ET; force-flat 15:45 ET; one long + one backside slot per name (no pyramiding).
Sample is 10 days — treat single-digit-n cells as directional, not conclusive; the meaningful signals here are Trend (n=47) and the ROTATION pull-in bleed (n=94). Bands use the engine's session VWAP+σ (premarket-inclusive, same yardstick as the regime tag). Roadmap:
SIP→Indicators→g8trade plan.