diff --git a/crates/ml/src/trainers/dqn/config.rs b/crates/ml/src/trainers/dqn/config.rs
index 038a5d84c..5f84ab20c 100644
--- a/crates/ml/src/trainers/dqn/config.rs
+++ b/crates/ml/src/trainers/dqn/config.rs
@@ -1341,7 +1341,7 @@ impl DQNHyperparameters {
             // P2-A Enhancement
             initial_capital: 100_000.0, // $100K default
             bars_per_day: 390.0, // 1-minute bars (6.5h × 60min)
-            min_hold_bars: 1,  // was 5 — model learns hold timing via urgency branch
+            min_hold_bars: 5,  // 5-bar minimum prevents coin-flip exits (1-2 bar churn)
 
             // P2-B Enhancement
             cash_reserve_percent: 0.0, // Default: no reserve (backward compatible)