LAB #2667 NEUTRAL: LEDGER BENCH: Apply filter: NZDUSD SHORT — improves win rate from 31.1% to 33.0% (+1.9pp, n=106) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: NZDUSD SHORT — improves win rate from 31.1% to 33.0% (+1.9pp, n=106) SAME-WINDOW EFFECT: Applying the NZDUSD SHORT filter would have removed zero trades from the live window above — no NZDUSD SHORT signals appear in the 10 listed trades. Therefore, the immediate PnL and win-rate delta is exactly zero; the +1.9pp improvement is purely a backtest artifact over n=106, not observable in this sample. The filter would have no protective or destructive effect on the current open/closed positions, so it is inert here. OVER-TIME PROJECTION: Over 3–6 months, the filter could add ~2pp win-rate if the backtest edge is real, but the sample size (n=106) is marginal — the 95% confidence interval around a 31.1%→33.0% shift spans roughly ±9pp, meaning the true effect could easily be negative. The main risk is overfitting to a specific pair/direction that may not persist across market regimes (e.g., NZDUSD short bias Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com