Convex Hull and GRU Trailing Stops for Momentum Breakouts
Summary
The article proposes a custom trailing stop for momentum trading that combines a price boundary with a sequential momentum filter. A rolling convex hull is used to locate structural price extremes, while a compact gated recurrent unit (GRU) evaluates recent price changes. The intended effect is to keep a stop aligned with the outer structure of a trend and temporarily pause stop adjustments when a counter-move appears inconsistent with the broader momentum.
The discussion contrasts this design with a smoothed, historically anchored approach that may lag during breakouts. It frames the method as a possible way to reduce exits caused by short-lived volatility, while acknowledging that it may struggle in tight ranges. The article describes an MQL5 Wizard implementation and test runs, but the supplied text gives limited detail about their results. It also notes that the runs cover a narrow period and that changing other inputs makes it difficult to isolate the contributions of the hull and neural network; broader testing across symbols and conditions is needed.
Key ideas
- A rolling convex hull supplies price extremes for a structure-based trailing boundary.
- A GRU processes recent price changes and acts as a filter for stop adjustments.
- The combined method aims to avoid exits during brief counter-moves in a continuing trend.
- The approach is designed for momentum and breakout conditions and may perform poorly in ranges.
- Limited test coverage and changing inputs prevent a clear assessment of each component's value.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.