Calling All Bloomberg Wizards

[NormR]

I didn't see a thread for Bloomberg related questions.  So, here's one :)

 

 

I started playing with Bloomberg's backtester.  But I'd like to know what's wrong with it.  i.e. where does it fail or come up short. 

 

For instance, I already know the available time period is too short for very serious work - at least in Canada.  But I imagine a few other issues might trap the unwary.  Aside from the basic data mining problem, of course.