Using Hierarchical Clustering of Price Ranges for Volatility-Based Position Sizing

Summary

This article explains how to apply agglomerative hierarchical classification (AHC) and dendrograms to historical price bar ranges. It treats each range as an observation, groups similar observations into clusters, and uses their hierarchical relationships to estimate a coming bar’s range. The proposed application is an MQL5 Wizard money-management class that adjusts trade volume according to the volatility estimate; the same information could also inform exposure changes or entry and exit levels.

The discussion motivates range forecasts with volatility clustering: periods of wider ranges may be followed by further wide ranges, and calmer periods by calmer ones. It outlines the use of ALGLIB’s clustering facilities and describes a comparison against fixed-margin sizing using an Awesome Oscillator signal on EURUSD four-hour data over a stated one-year test period. The author reports potential for volatility-based volume adjustment, but the supplied excerpt gives no numerical performance results or detailed robustness analysis. The findings are presented as preliminary testing, so they do not establish that the approach will work across markets, periods, or strategies.

Key ideas

• AHC groups observations by similarity and represents relationships among clusters in a dendrogram.
• The proposed model uses historical high-to-low bar ranges to estimate future ranges.
• Volatility estimates can guide position volume as well as trade levels and exposure changes.
• The article compares volatility-based sizing with fixed-margin sizing in a limited EURUSD test.
• The reported potential is preliminary and does not demonstrate broad or robust profitability.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.