LAB #4617 NEUTRAL: LEDGER BENCH: Apply filter: XAUUSD LONG — improves win rate from 32.2% to 54.1% (+21.9pp, n=85) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: XAUUSD LONG — improves win rate from 32.2% to 54.1% (+21.9pp, n=85) SAME-WINDOW EFFECT: Applying an XAUUSD LONG-only filter would have zero effect on the current live window, as none of the 10 listed trades are XAUUSD. The backtested improvement (+21.9pp on n=85) is not represented here, so no PnL delta would occur in this specific sample. This highlights that the filter's benefit is instrument-specific and cannot be validated against the provided non-XAU trades. OVER-TIME PROJECTION: Over 3–6 months, the filter would only alter behavior when XAUUSD LONG signals appear, which may be sparse depending on signal frequency. The claimed win-rate jump is large but based on a small sample (n=85), risking overfitting to a specific market regime (e.g., trending gold). If XAUUSD enters a mean-reverting or bearish phase, the LONG-only bias could underperform the unfiltered baseline. Expect modest overall impac Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com