LAB #2788 NEUTRAL: LEDGER BENCH: Apply filter: USDCHF SHORT @ LONDON_OPEN with 6 conf — improves win rate from 31.0% to — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: USDCHF SHORT @ LONDON_OPEN with 6 conf — improves win rate from 31.0% to 71.4% (+40.5pp, n=7) SAME-WINDOW EFFECT: Over the current live window, the filter would have blocked the USDCHF SHORT that is currently [REJECTED/open] — that trade is not counted as a win or loss yet, so the immediate PnL delta is neutral. However, the filter would also have blocked the USDCHF SHORT from 2026-08-04 (entry 1.40358, CLOSED/WIN) — that was a winner, so applying the filter would have removed a confirmed win. The backtest claims +40.5pp win-rate improvement on n=7, but the live window shows at least one filtered-out winner, suggesting the filter may be overfit to a tiny sample. Net effect on this window: likely negative PnL if the rejected USDCHF later wins, and definitely negative if the 08-04 winner is excluded. OVER-TIME PROJECTION: Over 3–6 months, a filter with n=7 and a 40pp swing is statistically fragile — t Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com