Linear regression involves the use of two different variables to define a single relationship. Traders use them for key price point identification.
As you might expect from the name, a linear regression line is straight. It’s used in technical analysis to help identify the prevalent trend over the duration of a number of previous periods.
In contrast, moving averages change all the time as they track a specific evolution of price over the specified data range. The linear regression takes a selection of periods that have been defined by the user and plots then in a way that best fits the overall trend.
So, it doesn’t just draw a line between today’s price and one from a certain number of periods ago. What’s going on between the two points is also significant.
For instance, in the below example of a 50-day period from the S&P 500, the market’s net gain is positive, but there is a negative slope on the linear regression line. That’s because there’s a forceful move down for the duration of the period. The majority of linear regression forms use the mean or average as their basis, so they are good at picking up outliers.
There is a negative slope to the linear regression line because for this period the mean price was lower than its present one.
The linear regression line’s shape is just meant to give you one more visual clue towards that helps you with identifying and analyzing the way that the market is trending.
The implication behind linear regression is that it’s good at giving predictions of how an output will change based on an input. In this instance, it’s making use of past price information to find out the overall way that future price data is trending. Naturally, no one can tell the future, but trends are still considered to be useful indicators of market sentiment that can be used to help guide future trades.
Traders who follow trends would probably find the difference that separates the 50-day and 100-day regression lines to be confusing and would wonder which of them was the more accurate, but remember that traders shouldn’t rely on any single indicator alone to inform their trades.
Maintaining buy/long trades when an X-period regression line it’s angled up and in sell/short trades when it’s during a down will not be a profitable trading strategy.
The best approach would be to alter it to suit your trading period. Those who hold longer-term positions could try using a 100-day linear regression line on a daily chart. For short-term positions of just minutes or hours, a 20-period linear regression line with a 5-minute chart would be more appropriate.
For pairing with a moving average, it might be prudent to have them both be the same period for the sake of equal comparison.
If you’re someone who’s happy to put their faith in the signals produced by the linear regression line and SMA for these periods, chances are you’d ignore anything related to trends.
Conclusion
A linear regression line presents you with general price direction information for a specific past period in a simple-to-read format.
Moving averages bend to so they mould to their weighting input, while linear regression lines go for the best fit of data in a straight line.
Linear regression lines depend more on the timeframe’s period considered in relation to moving averages. You won’t often find wide dispersions in the general gradation of a moving average between, for instance, 50 and 100 periods. You might find a big difference between a 50-period linear regression line and a 100-period regression line.
You shouldn’t use a linear regression line by itself as a system. Use it as part of a larger trading system that features other technical indicators like candlestick patterns, price, support and resistance levels. You can also use fundamental analysis to underpin the setups for your trades.
