For what its worth I tend to agree with Techs idea. With many "if" attached to that idea. To put some meat onto it as we all know I never make a statement without the stats to back it up
rolleyes: yeah right). I have put together a back test of sorts to find the rolling correlation of each XTL (ASX20) constituents to the actual index. Have ran out of time right now but have got to these messy graphs for now. Not the XTL (thick black line) right in the middle of them all, no surprise there.
Trouble is though the time frame. Clearly over a long period you have stocks moving in separate directions. But over a short period say 20 - 40 days they are actually very closely correlated. This may be where your idea breaks down. If not breaks down maybe at least fights against the stats if you are actually
trying to be long and short at the same time.
Will come back latta with the correlations over a 20 day period, if anyone cares about it.