MODEL ADOPTION #2578: REJECT — TG AUTO: Institutional metrics refreshed — Sharpe 1.66 · Sortino 3.88 · Calmar 2.11 — E8 Intelligence Research

DECISION: REJECT ELEMENT: Nothing survives. The 80.3% win rate and 94.7% no-loss rate are artifacts of a 224-month backtest with proprietary vaulted filters — this is an unfalsifiable black box. The 4.3% CAGR is structurally incompatible with our 15s–7m hold scalping model; we cannot adopt a strategy whose per-trade edge is thinner than our spread costs on XAUUSD. The "OOS Sharpe 1.9" is meaningless without knowing the exact config weights, and the 2.05% max DD over 224 months implies the system never faced a true regime break — our current live window shows exactly that correlated chop on USDJPY/XAUUSD that this backtest would have smoothed over. EXPECTED IMPACT: Zero. Adopting this would add no concrete rule to our E8 timing geometry. The only transferable element — "weighted multi-config portfolio" — is a portfolio construction technique, not a tradeable edge. Our live model already uses `rizzy`-family entries; this candidate adds nothing but a claim of higher win rate backed by h Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com