Detrended Volume Profiles with Linear and Polynomial Regression
Summary
This indicator rebuilds a volume profile around a fitted linear or quadratic regression curve rather than fixed horizontal price levels. It assigns each bar's volume to a bin based on its distance from the curve, producing a profile of trading activity relative to the trend. The busiest bin defines a Point of Control that follows the fitted curve, so its location remains interpretable as prices rise or fall. The implementation also draws standard-deviation channel curves and reports trend direction, POC information, and channel extremes.
The explanation describes least-squares fitting with a centered, normalized time axis, which simplifies the polynomial calculations and improves numerical conditioning. It also flags a key limitation: the default grid width uses the source series' standard deviation, which can be dominated by the trend and compress the profile into very few occupied bins. Synthetic examples illustrate this effect; a switch allows grid sizing from residual dispersion instead. These examples demonstrate a design issue, not trading performance. The indicator is an analytical chart tool, and the note does not establish that its POC or bands generate profitable signals.
Key ideas
- The profile bins volume by each bar's distance from a fitted linear or quadratic trend curve.
- Its Point of Control is a curve representing the most active distance-from-trend band, rather than a fixed price level.
- Centering and normalizing the regression time coordinate simplifies least-squares calculations and improves numerical stability.
- Using standard deviation of the raw source can make bins too wide in trending markets and reduce profile resolution.
- A residual-dispersion setting addresses that sizing issue, but the examples do not establish trading performance.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.