Lorentzian-Sweep Momentum Hybrid
Family: hybrid · Regime: trending · Complexity: high · Asset classes: FX, Equities, Crypto · Timeframes: 15M, 1H
Thesis
Local price action clusters in multi-dimensional feature space (Lorentzian) provide a superior regime filter than simple moving averages. By combining this non-linear classification with linear regression and a liquidity sweep filter, the strategy capitalizes on momentum breakouts that occur after weak-hand liquidations, using objective fractal levels for risk management.
Components
- Lorentzian Classification (regime) — Classifies the current market regime by finding the k-nearest neighbors in historical data using Lorentzian distance to determine if current features favor bullish or bearish outcomes.
- Linear Regression (LINREG) (direction) — Ensures the entry aligns with the statistical 'line of best fit' slope, filtering out momentum signals that are counter to the immediate linear trend.
- Rate of Change (ROC) (entry) — Provides the tactical trigger; a zero-line cross indicates that current price velocity is shifting in favor of the established regime.
- ZigZag Color (Classic) (exit) — Acts as a trend-following trailing exit by identifying major structural shifts in swing direction.
- Williams Fractals (risk) — Provides objective, market-structure based levels for stop loss placement, utilizing the 2-bar lag to ensure the local extreme is valid.
- GFRMa Pivot HTF3 Dashboard (confirmation) — Provides multi-indicator confirmation (RSI/CCI) to ensure momentum and volatility are synchronized across internal oscillators.
- Reversal Engine Score v1 (volatility_filter) — Filters for liquidity sweep events, ensuring the strategy enters after a 'stop run' or exhaustion point to increase the probability of a reversal.
Known failure conditions
- Lorentzian classifier feature selection (RSI, WT, CCI, ADX) becomes correlated during low-volatility regimes, leading to signal noise.
- Price moves significantly past the exit point before a ZigZag leg can confirm a reversal, leading to heavy profit giveback.
- The hardcoded 15M constraint in the Volatility Filter creates data desynchronization when trading on higher timeframes.
Explore the full interactive blueprint with parameter ranges and evidence on WOBR StrategyVerse, or generate this strategy as an MT4/MT5 Expert Advisor with QuantMogul AI Engine (free download).