The linear regression does not take the place of support and resistance lines. It should be viewed as a supplement and confirming indicator to identify the trend. If you select a good starting point, the linear regression delivers pure trend. Unlike support and resistance lines, the linear regression line doesn’t have any trading rules associated directly with it, but visually, the linear regression line is the most informative.
If you draw a support or resistance line whose slope varies dramatically from your linear regression line, one of them is wrong. In a similar vein, if the line has a very steep slope and no other linear regression line on charts of the same security has such a steep slope, you can deduce that the price movement is statistically abnormal. It is unsustainably fast and likely to come to a sad end when traders start taking profit.
This crash and burn is exactly what happened in the NASDAQ in March 2000, the “tech wreck.” Look at Figure 10-6.
The linear regression line from the low in October 1998 to the peak in March 2000 is so steep (and unprecedented) that anyone with an ounce of technical savvy could tell if was going to blow out and come crashing down – exactly as gurus were predicting. That doesn’t mean people shouldn’t take advantage of rising prices. It does mean that they shouldn’t have expected high and rising prices forever and should have placed trailing stops every day to lick in the extraordinary gains.
Here’s an interesting way to draw a linear regression:
1. When you suspect that the price is accelerating (in either direction) at an abnormally high speed, first you draw a true linear regression from the beginning of the data series but stop it at the point where the abnormality – the bubble – begins. You can see this only in hind-sight, but it doesn’t take a lot of hindsight.
2. Extend the line by hand, cutting off the bubble. The extension of the trendline becomes a more realistic forecast, because we know that bubbles always burst.
Anyone drawing this linear regression trendline extension before March 2000 (the market high) had a pretty good idea of where the market would land after the bubble burst. Yes, this is a highly subjective process, and it doesn’t always work. But it gets you thinking about the overall trend of the security or the market index.
Source: Page 178, Technical Analysis for Dummies. Authors: Rockefeller, & Barbara
0 comments
Post a Comment