diff --git a/config/training/dqn-production.toml b/config/training/dqn-production.toml
index 30e7797a0..4b16c5df4 100644
--- a/config/training/dqn-production.toml
+++ b/config/training/dqn-production.toml
@@ -68,6 +68,7 @@ min_loss_improvement_pct = 0.1
 [experience]
 initial_capital = 100000.0
 tx_cost_multiplier = 0.0001
+min_hold_bars = 5
 
 [experience.fill_simulation]
 ioc_fill_prob = 0.85
diff --git a/config/training/dqn-smoketest.toml b/config/training/dqn-smoketest.toml
index 33148b004..46f414065 100644
--- a/config/training/dqn-smoketest.toml
+++ b/config/training/dqn-smoketest.toml
@@ -37,6 +37,7 @@ min_epochs_before_stopping = 5
 [experience]
 gpu_n_episodes = 32
 gpu_timesteps_per_episode = 100
+min_hold_bars = 3
 
 [experience.fill_simulation]
 ioc_fill_prob = 0.85
diff --git a/crates/ml/src/trainers/dqn/config.rs b/crates/ml/src/trainers/dqn/config.rs
index 2828d67e4..114a8c6cc 100644
--- a/crates/ml/src/trainers/dqn/config.rs
+++ b/crates/ml/src/trainers/dqn/config.rs
@@ -904,6 +904,11 @@ pub struct DQNHyperparameters {
     /// Derived from data pipeline's BarSize when available.
     pub bars_per_day: f64,
 
+    /// Minimum bars to hold a position before allowing exit or reversal.
+    /// Prevents per-bar churning. Trailing stop can override after this period.
+    /// Default: 5 (about 5 minutes for 1-min bars). Range: [2, 20].
+    pub min_hold_bars: usize,
+
     // P2-B Enhancement: Cash reserve requirement
     /// Cash reserve requirement as percentage of portfolio value (0-100)
     /// Default: 0.0 (no reserve, backward compatible)
@@ -1374,6 +1379,7 @@ 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: 5,
 
             // P2-B Enhancement
             cash_reserve_percent: 0.0, // Default: no reserve (backward compatible)
diff --git a/crates/ml/src/training_profile.rs b/crates/ml/src/training_profile.rs
index 7cb868c3b..635e682f0 100644
--- a/crates/ml/src/training_profile.rs
+++ b/crates/ml/src/training_profile.rs
@@ -195,6 +195,8 @@ pub struct ExperienceSection {
     pub initial_capital: Option<f64>,
     /// Maps to `DQNHyperparameters::transaction_cost_multiplier`.
     pub tx_cost_multiplier: Option<f64>,
+    /// Minimum bars to hold a position before allowing exit or reversal.
+    pub min_hold_bars: Option<usize>,
     /// Fill simulation sub-section for order routing realism.
     pub fill_simulation: Option<FillSimulationSection>,
 }
@@ -801,6 +803,9 @@ impl DqnTrainingProfile {
             if let Some(v) = ex.tx_cost_multiplier {
                 hp.transaction_cost_multiplier = v;
             }
+            if let Some(v) = ex.min_hold_bars {
+                hp.min_hold_bars = v;
+            }
             // [experience.fill_simulation]
             if let Some(ref fs) = ex.fill_simulation {
                 if let Some(v) = fs.ioc_fill_prob {