MODEL ADOPTION #3584: REJECT — LEDGER BENCH: Apply filter: Quarterly phase = ACCUMULATION — improves win rate from 32.4% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No element survives — the ACCUMULATION filter is a lagging, regime-dependent overlay that conflicts with E8's fade-polarity timing (which already exploits counter-trend exhaustion at precise rhythm nodes). Adding a quarterly phase filter would delay entries past the optimal geometric trigger, degrading R:R even if win rate rises. EXPECTED IMPACT: Win rate might rise +5–8pp in backtest, but average R:R drops by ~0.4R due to late entries (accumulation confirmed after price already left the zone). Net PnL likely flat to negative over 3–6 months, with 30–40% fewer trades — reducing absolute profit and compounding. RISK: Overfits to a specific 2026 quarterly cycle; "accumulation" is not a leading signal — it's a hindsight label. In live trading, the filter would exclude high-R:R fade entries at exact E8 nodes (e.g., AUDUSD SHORT at weekly judas rejection) that are the model's core edge. Also, n=128 is statistically weak; the +12.2pp could be noise from a sing Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com