RLS-LSTM Structural Liquidity Engine

Family: hybrid · Regime: trending · Complexity: high · Asset classes: Equities, Forex, Crypto · Timeframes: 15m, 1h, 4h

Thesis

Markets oscillate between noisy inefficiency and structured institutional trends. By using an LSTM-optimized SuperTrend to identify the regime and an RLS filter to remove price noise, we can isolate institutional 'footprints' (Order Blocks/Fair Value Gaps) that represent real supply/demand imbalances. The edge exists because the strategy only enters when ML-confirmed volatility, momentum, and structural liquidity align, exiting via statistical overextension (Z-Score).

Components

• ML SuperTrend (Ultimate) - Auto-Optimized AI with LSTM (regime) — Defines the primary volatility-adjusted regime; the LSTM/DQN architecture attempts to adapt to changing market micro-structures faster than static parameters.
• Heiken Ashi (direction) — Provides a synthetic trend filter to ensure the immediate directional momentum aligns with the ML regime before commitment.
• Order Block / FVG Detector (entry) — Identifies institutional footprints (liquidity gaps and supply/demand zones) as high-probability entry triggers within the broader trend.
• Z-Score (exit) — Acts as a statistical 'stretch' gauge, identifying when price has moved too far from the mean to justify holding a position.
• Grid Bot [Grid range plugin] Two Moving Avarages [psyll] (risk) — Calculates dynamic volatility-based boundaries for risk management. The Grid High/Low serves as a logic-driven stop-loss level.
• Relative Strength Index (RSI) (confirmation) — Ensures that the price movement triggering the OB/FVG is backed by sufficient (but not exhausted) momentum.
• Recursive Least Squares Adaptive Filter (RLS) (volatility_filter) — Filters high-frequency noise and provides a 'clean' trend line; price proximity to the RLS filter determines trade validity.

Known failure conditions

• Market enters a 'low-volatility grind' where the ML SuperTrend whipsaws despite LSTM optimization.
• RLS filter experiences numerical instability during extreme black-swan price gaps.
• RSI stays in overbought/oversold territory for extended periods while Order Blocks continue to form (strong parabolic trends).

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.