Traders' Roundtable

Is it fair to ssume that if we mix up systems using 1.5 that give good overall performance that its worth running the systems individually over thesame test period to ensure the 2 systems are not correlated in terms of monthly performance ?
I have combined 2 systems that give me good results and when I break down the monthly returns for each system the correlation is 0.3 which to me seems good to run with .. does anybody have a view on this topic or the likely point of statistical significance for equity stream analysis ?

Thanks
C

I do things the other way: I develop standalone systems first, and when they have performance stats that I like, I then start mixing systems.

Edward