Hello all, I am curious if anyone is familiar with a way to scan for extended multi-yr basing patterns/breakouts. The clearest example can be seen among many tech stocks since the majority of them have consolidated and formed extended trading ranges in the wake of the late 90's bubble. One particular example today is EMC. Take a look at it on a monthly chart and that is essentially the type of pattern that I am trying to screen for. Anyone have any ideas? Thanks, Jon
(1) You could look through tons of charts and make a list of those that appear to exhibit the pattern. (2) You could attempt to quantify what a long-term basing pattern looks like, translate it into computer code, run the code on your computer and then see if the computer spits out stocks like EMC that conform to the pattern you're seeking. (3) Concentrate on the high-flying tech stocks from 1999 and see if most of them resemble EMC.
If I understand correctly what you want, it's quite easy to do : Assumed you have a large historical data from every stock of your universe, you compute, as of today's close : a/ the number of days it "dominates" backwards (i.e. the number of past consecutive days where the price was below today's close); b/ the number of days it "dominated" as of yesterday's close. Then, you're looking for the largest value of (a-b). For EMC, on a weekly graph : early April, it dominated 7 weeks. One week later, it dominated more than 200 weeks. The value for EMC would be 200-7 = 193. Compute for your universe, and sort. I do this with R and a MySQL quotes database.
Thanks for the reply. While coding is always a possibility, I am currently looking for other potential programs/resources/scanners that have this capability (with a bit of tweaking of course), if any. Unfortunately, my coding skills are practically non-existent, and while I have no qualms with learning something new, I would like to see what other alternatives are out there at this point. Jon