LAB #3904 NEUTRAL: LEDGER BENCH: Apply filter: USDCHF SHORT — improves win rate from 32.5% to 54.3% (+21.8pp, n=265) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: USDCHF SHORT — improves win rate from 32.5% to 54.3% (+21.8pp, n=265) SAME-WINDOW EFFECT: Applying a USDCHF SHORT-only filter would have blocked the single USDCHF LONG (rizzy7m, 0.80231, CLOSED/LOSS) — that trade was a loser, so removing it improves the window's raw PnL slightly. However, the filter does not touch any other pair or direction, so the other 9 trades remain unchanged; the win-rate delta in this tiny sample is +1 win out of 10 (from 2/10 to 3/10 if you count the open as pending), which is far below the claimed +21.8pp. The backtest's n=265 is not represented here — the live window has only one USDCHF trade, and it was LONG, not SHORT, so the filter would have zero effect on the SHORT side. Net effect: negligible PnL change, no meaningful win-rate shift in this window. OVER-TIME PROJECTION: Over 3–6 months, if the filter truly isolates a profitable USDCHF SHORT edge, you might see a mode Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com