Linear Regression Curve
The Linear Regression Curve plots a line that best fits the prices specified over a user-defined time period. Think of the Linear Regression Curve as numerous lines, but both extreme ends of the lines are hidden, while the center portion is shown and is connected to other center portions of lines. The Linear Regression Curve is used mainly to identify trend direction and is sometimes used to generate buy and sell signals.
The chart of the S&P 500 E-mini Futures contract shows a 9-day Linear Regression Curve:
Many traders view the Linear Regression curve as the fair value for the stock, future, or forex currency pair, and any deviations from the curve as buy and sell opportunities.
Generally, when price deviates a certain percentage or number of points below the Linear Regression Curve, then a trader buys, thinking that price will revert back to fair value, which is thought to be the Linear Regression Curve.
In a similar manner, when price moves above the Linear Regression Curve by a trader specified percentage or point value, then the trader will sell, believing that price will return back to the Linear Regression Curve.
Other variations of these buy and sell signals could be employed. Since the Linear Regression Curve is great at identifying trend direction, if price is trending higher, a trader could only take buy signals when price deviated below the curve. Likewise, during a downtrend, a trader might only take sell signals, not wanting to fight the prevailing trend downward.
Arguably the most popular usage of the Linear Regression concept is the Linear Regression Channel, often used by large institutions. (see: Linear Regression Channel).