LAB #2876 NEUTRAL: LEDGER BENCH: Apply filter: USDCHF LONG — improves win rate from 31.1% to 37.5% (+6.4pp, n=96) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: USDCHF LONG — improves win rate from 31.1% to 37.5% (+6.4pp, n=96) SAME-WINDOW EFFECT: Applying a USDCHF LONG-only filter would have blocked the single USDCHF SHORT trade in the window (entry 0.81122, currently REJECTED/open). That trade is not yet resolved, so the immediate win-rate impact is ambiguous — if it closes as a loss, the filter would have improved the window's win rate from 7/10 to 7/9 (77.8% vs 70%); if it wins, the filter would have reduced it to 7/9 (77.8% vs 80%). The net PnL delta is roughly zero to slightly positive, but the sample is too small to infer anything about the +6.4pp backtest gain. OVER-TIME PROJECTION: Over 3–6 months, a USDCHF LONG-only filter could plausibly reduce losses if the pair's short-side signals are systematically weak in your model — but the backtest gain is based on only 96 trades, which is marginal. The risk is overfitting to a specific regime (e.g., USD Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com