UDT-Sweep MAPE Adaptive Hybrid

Family: hybrid · Regime: high_vol · Complexity: high · Asset classes: FX, Equities, Crypto · Timeframes: M5, M15

Thesis

Institutional participants utilize liquidity sweeps to fill large orders; by entering when this behavior is detected (via Daily Sweep/FVG) during high-complexity regimes (UDT Array logic), and scaling size based on prediction error (MAPE), a quantitative edge can be harvested from the subsequent volatility expansion.

Components

• User Type Array Scalar Tree Test Fixture (regime) — Acts as a complexity filter; higher static offsets (e.g., >9000) represent regimes where data persistence and institutional 'nested' orders are theorized to be active.
• Unsupported ta.macd Fast Length (Test Case) (direction) — The ultra-fast (1.5 length) EMA crossover serves as a high-frequency momentum trigger, filtering for immediate price pressure following a sweep.
• The Daily Sweep (entry) — Identifies institutional liquidity raids and Fair Value Gaps (FVG) specifically within the high-volume New York window.
• Bollinger Bands (Standard) (exit) — Provides a volatility-adjusted mean reversion target for exiting positions.
• Mean Absolute Percentage Error (MAPE) (risk) — Calculates the 'unpredictability' of the current price action against a baseline EMA; used to scale position size inversely to model error.
• Volumatic Support/Resistance Levels [BigBeluga] (volatility_filter) — Filters out trades that are entering directly into high-volume supply or demand zones to prevent 'buying the top' of a range.

Known failure conditions

• MAPE remains consistently above 5% on major pairs, indicating the underlying EMA model is failing to track price.
• Daily Sweep bias fails to provide directionality during low-volatility summer months.
• StaticOffset regime gate (9000+) is never triggered due to lack of market 'depth' or data complexity.

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.