MG I don't think its possible to take out all news events, especially non-periodic ones (i.e. such as NFP 1st Friday every month, RBA rate decision 1st Tuesday every month, etc). When I back-test test I test for these regular ones by filtering in/out those regular dates in my spreadsheet. Its worth doing because sometimes your pre-conceived notion about what effect those news events have is wrong. For non-regular ones I think the only way would be you would have had to have kept a historical diary.
