Linear Regression
Interpretation:
Investors use trading bands, lines drawn above and below the moving average, to isolate a range of prices for a given security, based on the concept that a stock generally trades within a predictable range on either side of the moving average. When a stock is near the upper or lower limits of the trading bands is when an investor should pay closest attention, according to conventional wisdom.
There are two conventional interpretations for the linear regression line.
The first interpretation is to use the linear regression as the overall trend line for that given period. If the line is positive, it may suggest a buying opportunity, whereas a turn downwards suggests one may consider selling the stock. Price divergences below the line indicate a possible buying opportunity, for the market is oversold, while divergences above the line indicate the market is potentially overbought. Linear regression will work best when the period being studied is similar to the cycle length or typical trend length of the security in question.
A second interpretation is to construct a linear regression channel, consisting of two parallel lines at fixed distances above and below the linear regression line. These lines can be used as support and resistance lines, which are used to watch the battle between buyers and sellers.
Support and resistance lines are drawn as the upper and lower limits of a trading range, whereby the support line is the bottom line, and is the point at which "bulls" will not let the price fall below, and the resistance line is the top line, the point above which the "bears" will not let the price rise above.
Conventionally, a breakout above resistance or below support indicates that there is either a) some news about the company which justifies recreating the upper and lower trading limits or b) there is about to be a correction towards the range as trader's are hesitant about the stock's new value.
Using the Linear Regression Channel can assist in finding support and resistance levels from the Linear Regression.
