Gold Customer
Joined: 7/28/2011 Posts: 12

I have been trying to figure out a way to filter out stocks that are choppy vs stocks that arent. The best way to describe this is by an example.
I want to be able to see stocks that are trading like this:
http://i.imgur.com/D1cOu.jpg
versus more choppy stocks like this:
http://i.imgur.com/vgQNN.jpg
These are cvlt on 9/27 and chk on 9/28 on the 5 minute time frame. Both of these stocks formed inside bars on the 15 minute chart, but obviously one stocks looks better than the other. Is there a volatility indicator/PCF or something else anyone can think of to filter out these types of choppy charts on the 5 minute time frame?
I tried filtering using adx but it doesnt seem to help. I have use a shorter time frame to filter out the choppiness but seems to filter out a lot of winners too. Any advise would be appreciated.

Worden Trainer
Joined: 10/7/2004 Posts: 63,198

You would need to come up with an unambiguous objective definition of what you are attempting to identify.
For example, if you wanted to express the mean absolute deviation of the ratio of closing prices from one bar to the next to next expressed as a percentage of the average ratio over the a period of 30bars, you could create an Indicator Formula similar to the following:
100 * (ABS(C / C1  (C / C30) ^ (1 / 30)) + ABS(C1 / C2  (C / C30) ^ (1 / 30)) + ABS(C2 / C3  (C / C30) ^ (1 / 30)) + ABS(C3 / C4  (C / C30) ^ (1 / 30)) + ABS(C4 / C5  (C / C30) ^ (1 / 30)) + ABS(C5 / C6  (C / C30) ^ (1 / 30)) + ABS(C6 / C7  (C / C30) ^ (1 / 30)) + ABS(C7 / C8  (C / C30) ^ (1 / 30)) + ABS(C8 / C9  (C / C30) ^ (1 / 30)) + ABS(C9 / C10  (C / C30) ^ (1 / 30)) + ABS(C10 / C11  (C / C30) ^ (1 / 30)) + ABS(C11 / C12  (C / C30) ^ (1 / 30)) + ABS(C12 / C13  (C / C30) ^ (1 / 30)) + ABS(C13 / C14  (C / C30) ^ (1 / 30)) + ABS(C14 / C15  (C / C30) ^ (1 / 30)) + ABS(C15 / C16  (C / C30) ^ (1 / 30)) + ABS(C16 / C17  (C / C30) ^ (1 / 30)) + ABS(C17 / C18  (C / C30) ^ (1 / 30)) + ABS(C18 / C19  (C / C30) ^ (1 / 30)) + ABS(C19 / C20  (C / C30) ^ (1 / 30)) + ABS(C20 / C21  (C / C30) ^ (1 / 30)) + ABS(C21 / C22  (C / C30) ^ (1 / 30)) + ABS(C22 / C23  (C / C30) ^ (1 / 30)) + ABS(C23 / C24  (C / C30) ^ (1 / 30)) + ABS(C24 / C25  (C / C30) ^ (1 / 30)) + ABS(C25 / C26  (C / C30) ^ (1 / 30)) + ABS(C26 / C27  (C / C30) ^ (1 / 30)) + ABS(C27 / C28  (C / C30) ^ (1 / 30)) + ABS(C28 / C29  (C / C30) ^ (1 / 30)) + ABS(C29 / C30  (C / C30) ^ (1 / 30))) / 30
Lower values would represent a more "orderly" progression over a period of 30bars when compared to higher values but only based upon this particular definition and measure. In other words, high values would be more "choppy" than low values over 30bars, but a different objective definition of choppiness would require a completely different formula.
Bruce Personal Criteria Formulas TC2000 Support Articles
