Symmetric Dual Squeeze-Release — Robust bilateral breakouts

Verdict: robust · Asset/TF: BTCUSDT 1h · Sample: 10,000 bars (hourly)

Hypothesis

Volatility compression (coiling) is fundamentally direction-agnostic. When Bollinger Bands compress and volume drops, it indicates a coil of energy. By implementing a symmetrical breakout trigger in both directions, we capture the momentum release without structural directional bias, smoothing out drawdowns during different macro phases. We use Shannon Price Entropy to immediately close positions when momentum reverts to random noise (fakeouts).

Strategy

{
  "entry_rules": [
    {
      "condition": "squeeze_release_long > 0.5",
      "direction": 1,
      "signal": "squeeze_release_long"
    },
    {
      "condition": "squeeze_release_short > 0.5",
      "direction": -1,
      "signal": "squeeze_release_short"
    }
  ],
  "exit_rules": [
    {
      "condition": "entropy_24 > 0.72",
      "reason": "high_entropy_noise"
    }
  ],
  "max_hold_bars": 10,
  "position_size": 0.3,
  "stop_loss_pct": 0.02,
  "take_profit_pct": 0.0275
}

Backtest

Robustness (the proof — do not skip)

• Walk-Forward: WFE = 0.29 → 100% positive returns across all OOS windows (+0.32%, +0.80%, +2.97%). Average OOS return is +1.36%.
• Monte-Carlo: risk-of-ruin = 0.0%; median final = $107,003, 95% worst drawdown = 19.84%.
• Sensitivity: stop_loss_pct is the most important parameter. Tighter SL (< 1.0%) leads to OOS failure, while wider SL (2.0%) is highly robust.

Research trail

Tools called: load_dataset -> ai_run_backtest -> optimize_exits -> walk_forward -> monte_carlo.
What I tried:

1. Short-only breakout strategy with tight stops: Sharpe 1.64 but negative OOS WFE 0.35 (overfitted).
2. Symmetric dual breakout strategy: wider SL (2.0%) and TP (2.75%), exiting early if Shannon Entropy > 0.72.
   What I learned: Symmetrical breakouts are more robust out-of-sample because they avoid macro trend regime bias. Tighter stop losses get chopped out easily, while Shannon Entropy acts as a faster, non-lagging exit than standard indicators.

Reproduce

Dataset: rlxbt_custom_features_BTCUSDT_1h.csv. Strategy JSON above. Re-run the same tools in the RLXBT app.