A least-squares regression line at the center, with parallel lines set one standard deviation apart. Measures a trend's center, slope, and overshoot objectively, without discretion.
A Linear Regression Channel fits a single straight line to the price data of a period using the least-squares method, then draws parallel lines above and below that line, offset by the dispersion of price (the standard deviation).
A hand-drawn trend line changes from person to person, depending on which lows you choose to connect. Look at the same chart and two people draw two different lines. Subjectivity, dressed up as discretion, slips in.
The linear regression channel replaces that subjectivity with statistics. It computes the one and only line that minimizes the sum of squared distances from every closing price in the period. Whoever draws it, the line is the same. The center line shows the "middle" of the trend, and the upper and lower lines mark the boundary of "overshoot," all as objective numbers.
There is no single inventor. The least-squares method itself was established by Gauss and Legendre in the early 19th century, and the linear regression channel applies it to a price chart. The most intuitive way to think of it is as a statistically rigorous version of the hand-drawn trend line.
How to Read
NASDAQ:AAPL
Members Only
Full access is reserved for members of the library.