Day Trading- Using Linear Regression in Stock Trading

Linear regression is a mathematical tool for analysis of stock performance during a defined period of time. Using statistics, linear regression is plotted mathematically to determine the exact middle or median of the price of a stock over time. The linear regression chart is a technical indicator of stock performance that uses a trend line that represents a "best fit" for the stock's mid-price over time. The trend lines of median stock performance are also referred to as linear regression lines.

Tools that use linear regression to aid in investment decisions make use of linear regression channels, which are lines that are plotted parallel to the mid-point trendline for a stock. Linear regression channels are set based on the level of standard deviation that the investor selects when analyzing linear regression for a certain stock. Support bands and resistance bands for the linear regression channels are placed on either side of the median price linear regression line. Usually, a standard deviation of 1 is too limited an amount of deviation to help guide trading decisions in most market conditions. Two to three standard deviations is the level of regression channel that are most often useful to include when analyzing a stock's performance over time.
When the stock's trading level exists inside a linear regression channel, it will deviate from the linear regression line while staying within a measurable overall trend. Strategies using linear regression channels include looking for extreme swings in a stock above and beyond the linear regression channels, which are set at 2 to 3 standard deviations beyond the median. Investors can determine points of initial stop loss or investment by plotting points in advance using the linear regression channels. Opposing channel lines, which represent higher and lower levels of stock price, will serve to guide investment strategies. Often, stock prices will rise above or fall below linear regression channels, which is referred to as breaking through resistance, or sinking below the support channel line, when news about a company's earnings, products or other factors are reported.

A more general way to use linear regression to make buying and selling decisions, is to use the linear regression line as a trendline, without incorporating linear regression channels. For example, if the trendline is moving upward, it may be a good choice to buy, and if the trendline is negative, it may represent a signal to sell. In addition, prices that fall below the trendline could represent a buying opportunity as well.