LAB #2302 NEUTRAL: LEDGER BENCH: Apply filter: NZDUSD LONG — improves win rate from 31.5% to 36.8% (+5.2pp, n=68) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: NZDUSD LONG — improves win rate from 31.5% to 36.8% (+5.2pp, n=68) SAME-WINDOW EFFECT: Over the current 10-trade window, no NZDUSD trades appear, so the filter would have changed nothing — zero trades affected, zero PnL delta. The claimed +5.2pp win-rate improvement is derived from a separate backtest (n=68) that does not overlap with the live window shown. Therefore, this adjustment is untestable against the provided live data; it neither helps nor hurts the current open positions. OVER-TIME PROJECTION: Over 3–6 months, a NZDUSD-LONG-only filter could plausibly improve win rate if the backtest is not overfit, but the sample (n=68) is small and the +5.2pp gain is within noise for a 31.5% baseline (standard error ≈ 5.8pp). The filter also reduces trade frequency and may introduce selection bias — NZDUSD long conditions may correlate with specific market regimes (e.g., risk-on, commodity strength) tha Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com