Sigmoid-JB Statistical Breakout System

Family: breakout · Regime: trending · Complexity: medium · Asset classes: FX, Equities, Crypto · Timeframes: M15, H1

Thesis

Market participants often underestimate tail risk. By using the Jarque-Bera test to detect non-normal price distributions, we can filter for breakouts that occur during momentum 'surges' (Sigmoid) while dynamically reducing exposure when the market enters a statistically unstable (high kurtosis) state, thereby capturing trends while surviving black swan events.

Components

• Logistic Function (SIGMOID) (regime) — Determines the momentum cycle regime; values above 0.5 suggest an accelerating bullish cycle, while values below 0.5 suggest a bearish acceleration.
• SuperTrend MT5 (FxGeek) (direction) — Provides the primary trend filter and volatility-adjusted support/resistance levels.
• Donchian Channels (entry) — Acts as the tactical entry trigger by identifying local price breakouts (market structure shifts).
• GoldenChart MACD (exit) — Used to identify momentum exhaustion via histogram color changes for early profit taking.
• Jarque-Bera Test (JB) (risk) — Quantifies tail risk; high JB values (non-normal distribution) trigger reduced position sizing due to unpredictable volatility.
• ProfitRobots Dashboard Template (confirmation) — Confirms trend alignment on the H1 timeframe to ensure the intraday breakout is supported by higher-level structure.
• Standard Deviation (volatility_filter) — Filters out low-volatility 'dead zones' where Donchian breakouts are prone to false signals.

Known failure conditions

• The Jarque-Bera value remains consistently above 20, indicating a permanently fat-tailed regime where stop-losses are statistically unreliable.
• Price spends more than 60% of the time in Sigmoid saturation (values 0.95+ or 0.05-) without price movement, indicating a failure of the nonlinear mapping.

Explore the full interactive blueprint, parameter ranges and evidence on WOBR StrategyVerse, or generate this strategy as an MT4/MT5 Expert Advisor with QuantMogul AI Engine.