J3P Playbook #1 — VWAP Reject Short SHORT
Why this one: across a 3-month, 8-name, out-of-sample test it was the only play positive in both halves — +0.19R train / +0.11R test (N=137 test). It is the short-side mirror of the VWAP pull-in, and it is the exact structure of your best trade of 7/2 (the 10:10 MU short). Everything else overfit; this survived.
① When to even look (all three gates required)
SIP — the stock is in play TODAY (on the daily SIP list / has a catalyst + volume). No SIP, no trade — the whole edge assumes an in-play name.
Morning — 09:30–11:00 ET only. The one window that isn't a graveyard; skip midday and the afternoon for this play.
Daily-aligned — the daily trend is DOWN (price below its 20-day average). Never short into a daily uptrend.
② The trigger (short-side mirror of the pull-in)
Price LOSES session VWAP from above — this bar closes back below VWAP after the prior bar was at/above it (a cross-down) — in a down-context (VWAP slope ≤ 0 or price below the 2-day anchored VWAP), with the close sitting in the VWAP → −1σ value zone and −2σ below the close. It's the reject back under value after a pop into it.
③ The trade
Entry
the VWAP-loss close
Stop
above the shakeout high (last 6 bars) or the 2-day AVWAP, + 0.25·ATR
Target 1
−2σ (cover ~60%)
Target 2
−4σ (cover ~20%)
Runner / exit
trail the last ~20% back toward VWAP; cover into VWAP if it stalls (this is a mean-reversion short — do NOT trend-trail it)
Size
round( Risk$ ÷ |entry − stop| ) — default $50 risk/trade
④ Discipline — the don'ts (learned from your own 7/2 tape)
- Don't short a name that isn't in play today.
- Don't take it after 11:00 — the edge is a morning edge.
- Don't short into a daily uptrend (daily-aligned gate).
- Don't over-trade it. On 7/2 you had this exact setup and covered the 20-share in ~3 seconds, then scalped the move you'd already called. The edge is in holding the size to the targets, not scalping.
⑤ Evidence (out-of-sample, 8 names × 3 months)
| window | N | avg R | win % |
| Train (Apr–May) | 59 | +0.186 | 49% |
| Test / out-of-sample (Jun–Jul) | 137 | +0.110 | 44% |
Status — LAB, not yet proven at scale. This is a durable candidate: it held out-of-sample on a fixed 8-name universe. It has NOT yet been tested on the correct SIP (daily in-play) universe or over more periods — the aligned-short holding could be partly a down-market in the test month. Next: re-test on the daily SIP universe; trade it small/live with strict discipline to confirm. Nothing here is promoted to the Pine.
J3P Playbook #1 · generated 2026-07-04 · see the
MU 7/2 replay for the live annotated example (your 10:10 entry).