Anyone else care to dip their toes in this muck?
Well, to put a toe in, Automatic Investor has a testing function, of which one of the controllable variables is to adjust the setting from 0% increment PC per purchase to 100% increment PC per purchase, in 1% increments.
Not that I'd want to do 100 tests of each and every percentage (!)
But a 0,25,50,75,100 series might give enough separation between to show what a cumulative effect might be.
Got a particular security and time range you'd like me to run through?
AIMster