Systematic Trading · Model
After the close, every name that just qualified under the book's validated rules becomes a plan. Accept it and it fills at the next session's open — the same fill the backtests use. This is a model: nothing here touches a broker.
Run this once after 16:00 ET. It scans the S&P 500 for today's qualifying closes. Momentum is capped at 3/week, ranked by 20-day dollar volume — the same cap-vs-rank test found that beats taking every signal (PF 1.42 vs 1.25) or ranking by breakout strength, which does worse than random (chasing bias). Reversal and Bollinger Band Reversion are both uncapped — their entry rules already do the quality-selection job a liquidity cap used to do (Reversal requires a confirmed 5%+ snap-back, ROC(10)>5%; BB Reversion requires an extreme 2.5+ std-dev dislocation inside a name already 20%+ above its own 200-day average). Re-running the scan also executes any plan you accepted on a prior session, at today's open — except Reversal and BB Reversion plans that gapped up past the signal's close, which are cancelled instead of filled: paying up for the gap means buying at a worse price than the one the signal was validated on, and this filter (PF 1.39 → 1.52 for Reversal, held up on both halves of a 10y temporal split) costs some trade count in exchange for the better edge.
A validated signal doesn't mean you have the capital for it. Every plan is sized off your account below (risk % of equity ÷ stop distance = shares), and accepting one checks it against your per-position cap and total margin — a plan that would breach either is blocked at accept time, not after.
Each strategy gets its own slice of the account, not the whole thing. A new position is sized by risk (risk % ÷ stop distance) same as always, but capped at sleeve budget ÷ slot count if that's smaller — a wide-stop signal never oversizes past its fair share, and one strategy can't crowd out the others or eat all your margin. Why this matters →
Priority (tier A) and over-cap (tier B) both passed the strategy's exact entry rule — tier is purely the weekly liquidity ranking, not a pass/fail gate. Accepting a tier-B name is fine if you want more exposure; the cap just marks where the evidence says quality drops off.