Created strategies not testing until the out of sample period
I am evaluating SQ and I think it is quite easy to understand for the most part. However, what I really don’t understand is that some strategies were generated with “unintentionally” limited testing periods. Let’s say I want to test 2009-2014 for IS and 2015 for OOS. Some strategies would come out as-is while some would somehow show, e.g., 2009-2014 as IS and a few days of 2015 for OOS.
The problem is, this doesn’t change even though I tried to set the data range to be what I want in “retest” and “improve”. The ultimate frustration comes when this behavior stops later at the optimization phase, where the most part of OOS is now tested and showed those previously-looking-good strategies to be garbage strategies when treated for more recent data, thus wasting a lot of time in the whole process.
Is there anything wrong about this? Or is this normal and I misunderstand some key principle of the software? Is that anyway to screen these strategies out easily so that I don’t have to click to look one-by-one?
Thanks a lot!
Are those strategies having only few trades or high number of occurrences? Few trades strategies could have this behavior. Any screenshots or examples available?
Viewing 2 posts - 1 through 2 (of 2 total)