LAB #2874 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.1% to 31.8% ( — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 31.1% to 31.8% (+0.7pp, n=1105) SAME-WINDOW EFFECT: Applying the MANIPULATION-phase filter to the 10 live trades above would have removed 4 trades (GBPUSD SHORT 1.34407, GBPUSD SHORT 1.34461, XAUUSD LONG 4270.17, AUDUSD LONG 0.70264) — all of which were wins. That would flip the live record from 6W/4L to 2W/4L, a catastrophic −4 wins. The +0.7pp backtest gain is not visible here; instead, the filter would have destroyed PnL in this specific window, suggesting the filter is regime-dependent and not uniformly beneficial. OVER-TIME PROJECTION: Over 3–6 months, the filter may add a small edge if MANIPULATION phases are genuinely low-quality, but the backtest improvement is marginal (+0.7pp) and within noise for n=1105. The main risk is that the filter removes high-conviction winners during trending phases misclassified as manipulation, as seen abov Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com