So, I set up the charts and spreadsheets to exclude the first bar of the London open. I must say my initial reaction was it looked strange; excluding this huge bar that announced the open.
The DAX is by no means the only instrument that has this phenomenon, it happens with anything that reacts to overnight news, it's just that with the DAX that first bar always looked strange to me.
Today's first bar, excluded in the test, was a huge spike up; guess they were happy with the results of the Austrian election and Italian referendum. After that, it was all over bar the shouting, because there was no follow through last I looked.
In any quant approach, we want to eliminate the noise from the signal.
I suppose what I'm looking at is whether that first bar is noise, or should it factor into the signal. We'll see.