The numbers
behind the playbook.
The full strategy suite, simulated against five years of daily history across roughly two thousand US stocks. The same detectors that fire on a live scan, walked bar-by-bar to their realized exit. Win rate, R-multiple, profit factor, hold time, exit mix — every number a setup has earned the right to claim.
How to read this. Tuned strategies pass every anti-overfit gate strictly and ship with live-adopted parameters. Tuned · provisional passes robustness + holdout PF + Sharpe haircut but fails the permutation p-value gate (either marginally or because the detector's pre-gates exhaust the sweep filter and signal-vs-null can't be measured) — params are published for inspection; live-engine adoption waits for the next regime change or quarterly re-tune to promote to full Tuned. Dead strategies failed holdout — dropped from live picks by the recommender, kept here for transparency. Inconclusive means sample too thin or edge indistinguishable from random label assignment; registry defaults stay in effect until the next quarterly re-tune.
Metrics: PF = profit factor net of $1.50 fee + 10 bps slippage per trade. p = shuffle-returns permutation p-value (n=200) on holdout. r = robustness score (fraction of ±10% neighbor configs that also clear PF ≥ 1.0). Sh = annualized Sharpe of trade returns with multiple-testing haircut (`SR × 1 / sqrt(2 log N_trials)`, an informal approximation — see methodology).
| # | Strategy | Win | Avg R | PF | Hold | Exits · T1 · T2 · Stop · TO |
|---|---|---|---|---|---|---|
| 01 | Cup & Handle Momentum·3,582 trades | 29.9% | +0.50R | 1.57 | 4d | 24% · 2% · 69% · 5% |
| 02 | Trend Pullback Pullback·1,092 trades | 26.5% | +0.33R | 1.45 | 2d | 15% · 4% · 72% · 9% |
| 03 | 52-Week-High Proximity Pullback Pullback·15,415 trades | 42.0% | +0.18R | 1.33 | 14d | 8% · 1% · 51% · 40% |
| 04 | Highest-High Breakout Momentum·36,728 trades | 46.5% | +0.13R | 1.31 | 17d | 15% · 1% · 42% · 42% |
| 05 | Power Earnings Gap Momentum·328 trades | 42.7% | +0.15R | 1.31 | 8d | 28% · 1% · 50% · 21% |
| 06 | ADX Trend Momentum Momentum·190,407 trades | 36.8% | +0.16R | 1.27 | 13d | 10% · 1% · 57% · 33% |
| 07 | Residual Momentum (Poor-Man's) Momentum·131,575 trades | 39.7% | +0.14R | 1.26 | 12d | 12% · 2% · 51% · 36% |
| 08 | ROC Breakout Momentum·120,823 trades | 32.3% | +0.21R | 1.26 | 5d | 19% · 2% · 64% · 15% |
| 09 | Snapback Z-Score Mean Reversion·124,882 trades | 50.0% | +0.10R | 1.25 | 10d | 7% · 1% · 29% · 64% |
| 10 | Quality 12-1 Momentum Momentum·217,217 trades | 33.5% | +0.20R | 1.23 | 4d | 29% · 3% · 64% · 3% |
| 11 | RSI Reversion Mean Reversion·30,398 trades | 51.2% | +0.09R | 1.23 | 7d | 14% · 1% · 25% · 61% |
| 12 | Volume-Weighted Trend Momentum·122,709 trades | 32.2% | +0.18R | 1.22 | 5d | 16% · 3% · 64% · 17% |
| 13 | RSI(2) Leader Dip Mean Reversion·189,123 trades | 49.8% | +0.08R | 1.21 | 10d | 7% · 1% · 25% · 67% |
| 14 | Multi-Period Strength Momentum·151,038 trades | 31.8% | +0.16R | 1.21 | 5d | 16% · 3% · 64% · 17% |
| 15 | MA Stack Confluence Momentum·102,531 trades | 35.1% | +0.13R | 1.19 | 6d | 12% · 3% · 59% · 26% |
| 16 | ATR Stretch Reversion Mean Reversion·146,981 trades | 51.0% | +0.06R | 1.18 | 7d | 8% · 0% · 18% · 73% |
| 17 | Qullamaggie Breakout Momentum·16,943 trades | 27.2% | +0.11R | 1.10 | 4d | 23% · 2% · 72% · 3% |
| 18 | Frog-in-the-Pan Momentum Momentum·206,175 trades | 32.8% | +0.09R | 1.09 | 5d | 28% · 3% · 65% · 4% |
| 19 | Trend Template Fresh-Pass Momentum·18,382 trades | 40.0% | +0.13R | 0.99 | 2d | 35% · 11% · 52% · 2% |
| 20 | Swing Condor Range-Bound·24,767 trades | 52.8% | +0.03R | 0.97 | 2d | 17% · 41% · 38% · 5% |
| 21 | RSI Overbought Mean Reversion·20,835 trades | 47.4% | +0.01R | 0.95 | 7d | 12% · 1% · 28% · 60% |
| 22 | Bear Flag Pullback·25,913 trades | 35.3% | -0.16R | 0.69 | 3d | 22% · 7% · 60% · 12% |
| 23 | VCP Breakout Momentum·1,259 trades | 36.9% | +0.06R | 0.38 | 1d | 42% · 38% · 18% · 2% |
5 years, ~2,000 stocks
Daily history end-to-end. Same TS detectors that power live screening — no parallel backtest fork.
Realized walk-forward exits
Entries simulated at next-day open. Exits walk bar-by-bar until stop, target, or max hold elapses.
No fees, no slippage
Reported P&L is raw price difference. Survivorship-uncorrected — true historical edge is several points lower.
Full methodology — survivorship bias, regime approximation, sample-too-thin handling, and the raw JSON payload — is at /methodology/strategy-metrics. Past performance, simulated or actual, is not indicative of future results.
Run these strategies on tomorrow's setups.
Daily setups graded A+/A/B+/B/C, regime-gated to the current market, with built-in journaling and R-multiple tracking. Same detectors as the backtest.