Lorentzian Matrix ORB Grid Hybrid

Family: hybrid · Regime: trending · Complexity: high · Asset classes: Equities, Forex, Crypto · Timeframes: 5m, 15m

Thesis

Market participants respond to psychological 'round numbers' (Grid Points) after initial session volatility (Opening Range) creates a trend; machine learning (Lorentzian Classification) can identify the specific momentum clusters that lead to profitable extensions beyond these ranges.

Components

• Supported for-in empty array slice result negative body history (regime) — Acts as a 'Runtime Integrity Filter'. Since this technical test returns 'na' on empty slices, it functions as a kill-switch: if the runtime environment fails this semantic test, the strategy assumes data corruption and remains flat.
• Supported matrix.swap_columns (direction) — Provides the 'True Price' reference. By swapping columns of a 1x2 matrix containing price, it isolates the current close in a matrix structure to ensure signal calculations are synced with the matrix-based Lorentzian entry.
• Lorentzian Classification (entry) — Uses k-Nearest Neighbors in a Lorentzian distance space to classify the current market state based on historical similarities. This provides the primary predictive signal.
• Momentum (MOM) (exit) — Used for mean-reversion exits; when momentum shifts against the trade, the position is liquidated before price hits the stop.
• Grid Points Utility (risk) — Establishes psychological support/resistance levels. These static round-number intervals define the physical Stop Loss and Take Profit levels, ensuring risk is pegged to structural 'psych levels'.
• Relative Strength Index (RSI) (confirmation) — Provides momentum confirmation to prevent entering Lorentzian 'Buy' signals in oversold conditions or 'Sell' signals in overbought conditions.
• Trepidity Opening Range with Extensions (volatility_filter) — Defines the volatility boundary. Entries are only permitted when price has expanded beyond the initial 30-minute opening range, identifying high-conviction intraday trends.

Known failure conditions

• Lorentzian classifier results in more than 5 alternating signals within a 10-bar window (overfitting/noise).
• Opening range width exceeds 2% of asset value, making grid-based stops mathematically irrelevant.
• Matrix operations return 'na' due to buffer overflows in the runtime.

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).