LAB #4986 NEUTRAL: LEDGER BENCH: Apply filter: Phase distance FAR (>10deg) — improves win rate from 31.7% to 31.9% (+0. — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Phase distance FAR (>10deg) — improves win rate from 31.7% to 31.9% (+0.2pp, n=927) SAME-WINDOW EFFECT: The filter would only have touched trades whose phase-distance reading sat in the FAR (>10°) bucket, and the live window gives us no phase-distance values — so the affected set is indeterminate from the data shown. The three closed losses (EURUSD, XAUUSD, AUDUSD shorts) and two closed wins (USDCHF, USDCAD) are all rizzy5m/rizzy7m shorts/longs with no phase annotation, meaning the filter's marginal +0.2pp cannot be mapped onto any of them. Best case it removes one of the three losses; worst case it removes one of the two wins, and the net PnL delta is within noise either way. No trade in this window is demonstrably improved by the filter. OVER-TIME PROJECTION: +0.2pp on a 31.7% base is a 0.6% relative improvement — statistically indistinguishable from zero at n=927 (the 95% CI on a 0.2pp shift at t Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com