LAB #3975 PROMISING: LEDGER BENCH: Apply filter: Phase distance TIGHT (<=2deg) — improves win rate from 32.5% to 37.6% (+ — E8 Intelligence Research

Abstract

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Phase distance TIGHT (<=2deg) — improves win rate from 32.5% to 37.6% (+5.1pp, n=593) SAME-WINDOW EFFECT: Applying the phase-distance filter (≤2°) to the current live window would have removed the EURJPY LONG (185.603, loss) and the USDCHF LONG (0.80231, loss) — both likely had phase distances >2° at entry. It would also have removed the AUDUSD SHORT (0.71765, currently open/rejected) and possibly the NZDUSD LONG (0.59515, open) if their phase distances exceeded the threshold. Removing two confirmed losses and one open/rejected trade from a window that already contains 4 losses and 3 wins would shift the win rate from 3/9 (33%) to roughly 3/6 (50%) — a meaningful improvement, but the sample is tiny and the two removed losses are not independent of the filter's selection bias. OVER-TIME PROJECTION: Over 3–6 months, a +5.1pp win-rate improvement from a tight phase filter typically comes at the cost of Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com