trend_aggressive on aggressive KILL
Did not survive the pre-registered gate. This is a first-class result, not an error: the honest thing to do with a hypothesis that fails is publish the kill.
Trend overlay (200-day SMA, downtrend tilt x0.5) tested alone against the plain aggressive baseline. Prior evidence (data/backtest/trend-overlay-comparison.md) shows a small in-sample Sharpe uptick (+0.035) and drawdown reduction; this run asks whether that uptick survives the full statistical gate (deflated Sharpe, walk-forward-with-reselection out-of-sample Sharpe, PBO) on its own, isolated from the other two overlays in the deployed stack.
The pre-registered gate
| Deflated incremental edge | +0.020 PASS (threshold +0.000) |
| Walk-forward out-of-sample uplift | -0.023 FAIL (threshold +0.000) |
| Probability of backtest overfitting (PBO) | 0.63 FAIL (threshold 0.50) |
No-lookahead probe: PASSED (divergence date 2024-03-27)
The underlying comparison
In-sample figures are the best-of-menu figure on one price history and are upward-biased by selection. Where they disagree with the deflated or walk-forward out-of-sample figures above, those win.
| In-sample Sharpe, candidate vs baseline | +1.058 vs +1.038 (uplift +0.020) |
| Walk-forward out-of-sample Sharpe, candidate vs baseline | +0.977 vs +1.000 |
| PBO combinations evaluated | 924 |
| Correlation to the plain baseline book | +0.974 (n=3851) |
Trust ladder
Current rung: UNVERIFIED. A hypothesis reaches a stakeholder surface as a claim only once it clears VERIFIED; see the verification page for what each rung means.
| validated | FAIL research loop verdict: KILL |
| lookahead_probe | PASS probe passed (divergence 2024-03-27) |
| panel_pinned | PASS panel_manifest_hash=f75ad5dbd3528e36 |
| harness_calibrated | FAIL calibration battery did not pass (null false-ship rate 13.3%, MDE None) |
Check the source
Every number above is read straight from these committed files, nothing recomputed for this page:
data/research/trend_aggressive-aggressive.json # the ORIGINAL pre-registered verdict, machine readable (never rewritten) data/research/trend_aggressive-aggressive.md # the same original verdict, written out data/research/trend_aggressive-aggressive.verification.json # trust ladder rung for the original verdict