LAB #4619 NEUTRAL: LEDGER BENCH: Apply filter: USDCHF SHORT — improves win rate from 32.2% to 49.8% (+17.6pp, n=299) — E8 Intelligence Research
Year: 2026
Type: preprint
Abstract: IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: USDCHF SHORT — improves win rate from 32.2% to 49.8% (+17.6pp, n=299) SAME-WINDOW EFFECT: Applying a USDCHF SHORT-only filter would have blocked the two live USDCHF LONG trades (one closed WIN, one rejected/open). Since the filter is direction-specific, it would not touch EURJPY, USDCAD, AUDUSD, GBPUSD, NZDUSD, or USDJPY trades. Net effect on the window: you lose one confirmed WIN (+~0.5R) and avoid one open LONG that is currently rejected — so PnL delta is slightly negative in this small sample, but win-rate impact is negligible because only 2 of 10 trades are affected. The +17.6pp backtest gain does not manifest here; it is invisible in this window. OVER-TIME PROJECTION: Over 3–6 months, if the backtest edge is real, you would see fewer USDCHF LONG entries (which historically lose) and more SHORT entries (which win at ~50%). However, the risk is overfitting to a single pair-direction slice with n= Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
Author: Caldin, Andrew Stewart
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7223463837
Cites:
Cited by:
Related to: