GlossaryAdvanced
Linear Regression and Linear Regression Channel
A statistically-calculated best-fit trend line through price, with a channel adding parallel lines at a set deviation above and below it.
Linear regression, applied to a price chart, calculates a statistical best-fit straight line through the period's price data, minimizing the overall distance between the line and every price point — a more rigorous, formula-driven alternative to visually drawing a trend line by eye.
A Linear Regression Channel adds parallel lines at a set statistical deviation above and below the regression line, creating a channel conceptually similar to Bollinger Bands, but built around a straight best-fit line rather than a moving average. See Kaavon's dedicated article for the full calculation.
Examples
- A linear regression line is calculated mathematically to minimize overall distance from price, rather than drawn subjectively by eye.
- A Linear Regression Channel's parallel bands are conceptually similar to Bollinger Bands, but built around a straight regression line instead of a moving average.