Adapting SuperTrend with Performance Clustering of ATR Multipliers

Summary

This document describes an adaptive trend indicator that evaluates multiple SuperTrend variants using different Average True Range multipliers. It scores their recent performance, groups the variants into three performance clusters using a k-means-inspired procedure, and forms a final SuperTrend from the selected cluster’s average multiplier. Traders can select the best, middle, or worst-performing group, making the resulting trend level more or less responsive according to recent behavior.

The indicator uses price crossings of its dynamic level to signal possible bullish or bearish trend changes, and exposes settings for ATR length, multiplier range and step, performance smoothing, and clustering convergence. The document supplies implementation code but no backtest results or evidence that clustering improves trading outcomes. Its performance depends on the evaluation window, cluster construction, and market regime; the signals should therefore be treated as an indicator method rather than proof of a profitable strategy.

Key ideas

• The method calculates several SuperTrend lines using different ATR multipliers.
• Recent performance scores are grouped into best, middle, and worst clusters.
• The selected cluster determines the multiplier used to construct the adaptive trend level.
• Price crossing the resulting level is interpreted as a possible trend reversal signal.
• The document provides no performance validation for the adaptive indicator.

Tags

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