MODEL ADOPTION #2605: REJECT — LEDGER BENCH: Apply filter: USDCHF SHORT — improves win rate from 31.0% to 59.1% (+28.0pp, — E8 Intelligence Research

DECISION: REJECT ELEMENT: No element survives — the USDCHF SHORT filter is a statistical artifact, not a geometric edge. E8 timing does not encode pair-direction polarity; the +28pp delta is a backtest mirage from n=171, and the live window shows zero evidence (1 trade, already rejected). EXPECTED IMPACT: Win rate unchanged (~55.6% in current window); profit unchanged. Adopting would add a non-E8 rule that could block high-RR USDCHF SHORTs during dollar strength, reducing expectancy by an unknown but likely negative amount. RISK: Overfit to a CHF-specific regime (e.g., SNB intervention or dollar-franc trend) that has already reversed; false-positive filter removes future profitable trades; degrades model discipline by mixing statistical noise with E8 rhythm logic. TEST: Must pass a 60-day shadow replay with ≥30 new USDCHF SHORT signals, filtered vs unfiltered, requiring ≥10pp win-rate improvement AND positive net expectancy AND identical E8 entry/exit timestamps. Additionally, Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com