Adaptive SuperTrend Factors Selected by K-Means Clustering

Summary

This indicator applies K-means clustering to a range of SuperTrend factor settings. For each factor, it tracks a performance measure based on whether the prior SuperTrend direction aligned with the latest price change. It groups those performance values into three clusters, initialized from quartiles, then selects the average factor from the chosen best, middle, or worst cluster to calculate an adaptive trailing stop.

The display includes the stop, a performance-weighted adaptive moving average, direction-change labels, candle coloring, and a dashboard showing cluster sizes, factors, and dispersion. The description suggests reading the stop like a conventional SuperTrend and notes that larger factor ranges produce longer-term signals. The supplied material contains implementation details, but no independent test results, comparison against a fixed-factor SuperTrend, or evidence that clustering improves trading outcomes. Results depend on the selected factor range, performance memory, and available historical bars, so the indicator should be evaluated across relevant markets and periods.

Key ideas

• The indicator evaluates multiple SuperTrend factors and scores their recent alignment with price changes.
• K-means groups factor performance into best, average, and worst clusters.
• The selected cluster’s average factor sets the adaptive SuperTrend trailing stop.
• A performance index weights an adaptive moving average and is also reflected in signal labels and candle coloring.
• The document provides no independent evidence that the adaptive method outperforms a fixed-factor indicator.

Tags

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