MODEL ADOPTION #1940: REJECT — LEDGER BENCH: Apply filter: XAUUSD LONG @ ASIAN with 5 conf — improves win rate from 32.0% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No element survives — the "XAUUSD LONG @ ASIAN + 5 conf" filter is a time-of-day heuristic, not an E8 geometry timing rule. It lacks a causal mechanism tied to the Gods-Clock rhythm or FADE polarity edge. EXPECTED IMPACT: Win-rate lift is illusory — n=21 backtest has ±20pp CI, and the live window's "improvement" is driven by removing 3 losing shorts that happened to cluster in Asian hours. Over 3–6 months, this filter would cut trade frequency by 15–25%, increase variance, and likely degrade profit factor because it also blocks profitable Asian breakouts (gold can trend strongly during London open overlap). Net expected impact: win-rate +5–10pp at best, but total R-multiple per month likely flat-to-negative. RISK: Overfits to recent gold consolidation regime; introduces a non-E8 time dependency that conflicts with the model's rhythm logic; reduces sample size for future edge validation; could mask real short signals during Asian momentum bursts. TEST: Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com