LAB #4900 NEUTRAL: LEDGER BENCH: Apply filter: USDCHF LONG — improves win rate from 31.9% to 41.6% (+9.7pp, n=154) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: USDCHF LONG — improves win rate from 31.9% to 41.6% (+9.7pp, n=154) SAME-WINDOW EFFECT: Over the listed live window, the filter touches exactly two trades: the USDCHF LONG on 2026-09-04 (already a WIN) and the USDCHF LONG on 2026-09-08 (currently REJECTED/open). It would have kept the winner and, depending on whether the 09-08 entry is still live, either left it untouched or converted a rejected signal into an open position — so the realized PnL delta on closed trades is zero, and the only real change is one additional open exposure. The claimed +9.7pp win-rate lift cannot be validated here because the window contains a single closed USDCHF LONG, which is already a win; n=154 is nowhere in this sample. OVER-TIME PROJECTION: If the +9.7pp lift is real and out-of-sample, it moves USDCHF LONG from a losing edge (31.9%) to a marginal one (41.6%) — still below breakeven for most R:R profiles, so it is a

Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com