LAB #2070 NEUTRAL: LEDGER BENCH: Apply filter: NZDUSD LONG — improves win rate from 32.0% to 35.9% (+4.0pp, n=64) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: NZDUSD LONG — improves win rate from 32.0% to 35.9% (+4.0pp, n=64) SAME-WINDOW EFFECT: Over the current live window, no NZDUSD trades appear in the list, so this filter would have changed nothing in the observed 10 trades. The claimed +4.0pp win-rate improvement (32.0%→35.9%) is based on n=64 backtest samples, which is thin and likely dominated by a few favorable NZDUSD sessions. If applied live, it would only affect future NZDUSD LONG entries, not the current XAU/AUD/USD/JPY/EUR mix, so PnL delta in this window is zero. OVER-TIME PROJECTION: Over 3–6 months, a filter that only allows NZDUSD LONGs would reduce trade frequency and concentrate risk on a single pair and direction. The +4.0pp edge is within noise for n=64 (standard error ~6pp), so it could easily be a false positive from regime-specific momentum. Risk: NZDUSD is commodity-linked and prone to sharp reversals on risk-off days; a long-only Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com