Linear Regression Channel Indicator: Formula, Settings and How to Read It
Linear Regression Channel fits a least-squares straight line through the last N closes and places two bands a multiple of the residual standard deviation on each side. Here it is computed as a rolling series rather than as a drawing, so the chart shows how the right-hand end of the channel evolved bar by bar.
Senzoukria · Indicators · Updated September 2026
Linear Regression Channel ships with the Senzoukria desktop app, in the Averages & volatility group of the indicator catalogue. It is drawn on the price chart.
What Linear Regression Channel measures
At every bar the fit is redone over the window, with the oldest bar at t = 0 and the current one at t = N - 1, using closed-form sums for the regression. The plotted basis is the terminal point of the fitted line, a + b x (N - 1), not the line across the whole window. The half-width is the root of the mean squared residual of that window, dividing the sum of squared deviations from the fitted line by N, the same population convention used for the Bollinger bands in this catalogue. The edges are the basis plus and minus the multiplier times that value, and nothing is plotted until N bars are available.
The formula, as implemented
This is not a description of how the indicator is usually defined elsewhere — it is what the shipped code computes, documented next to the implementation:
Linear Regression Channel — version SÉRIE GLISSANTE (choix documenté) : à chaque barre on refait le fit moindres carrés sur la fenêtre N, et on émet le POINT TERMINAL du canal (pas le canal statique redessiné — un overlay série trace l'évolution du bord droit du canal) : basis[i] = a + b·(N−1) σ[i] = √( Σ (y_t − (a + b·t))² / N ) — écart-type de POPULATION des résidus de la fenêtre (÷N, cohérent avec la convention Bollinger du catalogue) upper/lower = basis ± k·σ Warm-up (< N barres) → null. Défauts N=100, k=2.
How to read it
- The basis is the regression's estimate of where price should be right now given the last N bars, and its rise or fall is the fitted slope made visible.
- Channel width measures how well the window fits a straight line: narrow means the move has been close to linear, wide means it has not.
- A close outside an edge says the latest close deviates from its own trend fit by more than the chosen multiple of the residual spread.
- The historic line is a sequence of terminal points, one per bar, not one static channel; old sections are not a trend line that was drawn in the past.
- A flat basis with a wide channel is a different state from a flat basis with a narrow one: the window has lost its linear structure.
Parameters and defaults
The defaults are a period of 100 and a multiplier of 2, with the period accepting 2 to 500. A long period gives a stable slope that ignores recent turns; a short one makes the basis swing with the last few bars and the width jump. The multiplier scales both edges and accepts 0.5 to 5 in steps of 0.1, leaving the shape of the channel unchanged.
| Parameter | Type | Default | Range |
|---|---|---|---|
| Period | number | 100 | 2 – 500 |
| σ multiplier | number | 2 | 0.5 – 5 |
What it does not show
A least-squares fit assumes the window is well described by a straight line and returns a slope even when it is not; the width is your only warning that the fit is poor. Extending the slope beyond the current bar would be extrapolation, which this indicator deliberately does not plot. Only closes enter the regression, so nothing about where volume traded or which side was the aggressor reaches the channel. And a single outlier close both tilts the line and widens the band, then keeps doing so for the full length of the window.
Using it in Senzoukria
Add it from the Indicators panel of any footprint chart or candle chart. It runs on futures data from Rithmic or Databento and on crypto pairs from Binance and Bybit, on the same engine — the calculation does not change with the venue, only the data feeding it does. Market data subscriptions are billed by the provider, separately from the app.
Related indicators
- SMA — Averages & volatility
- EMA — Averages & volatility
- WMA — Averages & volatility
- Hull MA — Averages & volatility
- ATR (Wilder) — Averages & volatility
- Bollinger — Averages & volatility
See the full indicator library, or start with the order flow guide if you are new to reading aggression, delta and absorption.
Frequently asked questions
- Is this a linear regression channel drawing or an indicator?
- It is a rolling indicator. A drawing tool fixes a channel on a chosen range and leaves it there, while this version recomputes the fit at every bar and plots the channel's right-hand edge, so the history shows how that edge moved rather than one static channel.
- What does the width of a linear regression channel mean?
- It is the spread of the distances between each close in the window and the fitted line, scaled by the multiplier. A narrow channel means closes sat near the regression line, a wide one means they scattered around it. It measures the quality of the fit, not a future range.
- What period should a linear regression channel use?
- The period defines the stretch of market being fitted, and the default here is 100 bars. A period that spans two different regimes produces a slope describing neither, so it is usually chosen to cover one coherent move on the timeframe you are watching.