PureBytes Links
Trading Reference Links
|
Err, maybe play it in the opposite direction? If it gave 137 out of 140
profitable, not too shabby.
> I have been trying to fiddle around with this as well.
>
> Just for a quick look, I ran an optimisation on 6 months of 10-minute bars
> of the Nikkei future. I set 'exit at end of day'. The results were
> terrible - only 3 out of 140 tests were profitable. The best was length=5
> and F=4. I tested length from 5 to 70 in steps of 5, and F from 2 to 20
> in steps of 2.
> Got the same sort of thing from 30-min bars. And 60 minute bars gave zero
> profitable combinations.
>
> Then I tried 10 years of daily NK data (without end of day exit). Again
> miserable results - huge drawdowns.
>
> The little marks do look pretty on the chart though, so I would like to
> persevere with this a bit longer, and would appreciate any thoughts.
>
|