EOD, EOW and statistical significance

[tbmuc]

I'm thinking about the two kind of strategies

a) mean reversion
b) trend following

for stocks and ETFs on the EOD or EOW timeframe. On the one hand for the mean reversion the edge is often small and disappears after costs. On the other hand trend following on the EOD and EOW timeframe leads to few trades and for the most systems it may be to little to get a statistical significant result.

How do you handle these obstacles?

 

[CanOz]

I thought i would use this thread to post some market study data that 58 wants to analyze. Its in Excel so i'll have a go at attaching it. I've just done CL and GC for the moment, but happy to do more on request.

This particular study just looks at when a new high or low occurs.

hmmm, it seems excel files are not permitted

 

[Roller_1]

In relation to the first post, strongly disagree with his Mean reversion 'Observations' even with the higher comms in AUS a decent Mean Reversion system still has a potent edge

 

[CanOz]

In relation to the first post, strongly disagree with his Mean reversion 'Observations' even with the higher comms in AUS a decent Mean Reversion system still has a potent edge

Yeah, that was a while ago....

 

[Roller_1]

Yeah i saw that just thought might interest someone