The linear regression curve indicator is a common tool to predict the future values of a rading security relative to its past performance. This indicator plots a curve through price activity. The curve is obtained by plotting a line through each end point of the invisible linear regression trend lines. These invisible trend lines are calculated by plotting the minimal distance between closing prices, using the least squares method, over the number of bars defined in the input length. The default input length in TradeStation is nine bars.*
* “Linear Reg Curve (Indicator)” TradeStation Help, n.d. Web. 8 Dec. 2012.
Figure 8: Linear Regression Curve Indicator
The regression line is calculated by using the formula: Ŷ = a + bX. Where a and b are constants, Ŷ is the predicted value of Y, which in this case is price, and X is the number of bars that are moving. Then the least squares function is used to find the invisible linear regression trend lines to plot the minimal distance between closing prices:
Linear regression curves help to predict where the price of the market will be in the near future. When prices are trending up, this indicator tries to determine what the upward biasof the price may be; likewise for when prices are trending down. Some people use this indicator to determine if prices are overextended. If they see that prices rise above or fall below the regression curve, they believe that it suggests that the price is overextended and will eventually move back towards the line.
This indicator can be used to monitor a change in price direction. By analyzing the trend of the curves and how are prices moving compared to the curve one can obtain a general idea of overvalued prices that might eventually move back or undervalued prices that might eventually move higher.