yes, i know, i read his book. that's why i'm trying to experiment new things, i'm not completely satisfied with his explanations. by the way, have you ever thought about looking at a non-linear regression smooth curve +/- 2 standard deviations? same principle as bollinger bands, different averaging method...
try with a polinomial regression instead of a SMA.
Quote:
|
Originally Posted by Linuxser
This has been a long discussion (about how MA are better to BB)-
John Bollinger takes this point in his book. and put white over black and the use of different MA than simple (or default).
(IMHO) And obviously, as the creator, he´s absolutely right.
First, if you focus in the formula only, you can see why the use of different MA has no serious implications.
And second, if you focus on the different MA calculation and short periods as 20, there is another reason to discard some impact of using another than a Simple MA.
|