Standard-deviation Channel Width

Standard-deviation Channel Width reads how tightly price hugs its own trend. Each bar it fits a least-squares regression over the lookback and measures the residuals' standard deviation, plotting the implied channel's rail-to-rail span — the standard-deviation channel width — as one line. With the trend subtracted first, the reading isolates the noise around a move rather than the move: an orderly trend prints narrow whether steep or flat. The width is judged against its own rolling average, compression tints the pane, and a dashboard restates the reading in price units, percent of midline, ratio to average and trajectory.

How to Trade the Standard-deviation Channel Width?

• Width below its average: an orderly, efficient trend; crossings are alerted both ways.
• Compression: readings under the threshold fraction of the average flag a coiled trend, with Start and Release alerts.
• Trajectory: the dashboard's 5-bar comparison reads Widening when a trend gets sloppy, Narrowing when it tightens.

Standard-deviation Channel Width Settings

• Regression Length (default 100) and Source (default close): the fitted window — widths only compare at a fixed lookback.
• Deviation Multiplier (default 2.0): the k in rails at k standard deviations; scales the width linearly.
• Deviation Divisor (default n): population form, or the slightly wider n - 2 variant.
• Display Mode (default Normalized (% of midline)): percent of midline for comparability, or price units for sizing stops to the noise.
• Average Length (default 100) with Show Average (on); Compression Threshold (default 0.5) with Highlight Compression (on); Trajectory Lookback (default 5).
• Show Regression Channel (default off): draw the live channel on the price chart.
• Show Dashboard (on); style: Gradient Fill (on).

Frequently Asked Questions

How is this different from Bollinger Band width?

Bollinger Bands measure dispersion around a moving average, so a steady trend holds them wide — trend and noise share one number. This channel subtracts the fitted trend first, so width narrows for a clean move and widens only when price gets disorderly around it.

Why does the on-chart channel repaint?

Each new bar refits the regression, so the drawn channel shifts as the window rolls — it is a context overlay. The width series comes from each bar's own window; its history never changes.

Should I use the n or n - 2 divisor?

n is the population form most charting conventions compute; n - 2 divides by the regression's degrees of freedom. On a 100-bar window the difference is around one percent — pick one and keep it fixed.