|
PureBytes Links
Trading Reference Links
|
Stan, the look-forward nature of the shift means this type of system can
backtest amazingly well. The problem is you'll never know if you get a
"current" buy or sell until the days of the shift have passed.
However, suppose you use a one day forward shift to the MA. By using
hypothetical outcomes for the next trading day, you could make a
decision about whether to make a trade based on intraday activity.
Has anyone used such look-forward possibilities, anticipating a
Metastock EOD trade signal, should the close hit an expected target?
Derek
S Childs wrote:
> I have been experimenting with shifted moving averages and they test
> very well-from experience to good to be true type well. So I thought
> I'd post the formulas I have been using to see what's wrong with them.
> Example: Enter long When(Ref(Mov(C,opt1,VAR),opt2)>Mov(C,opt3,VAR)
> the short position of course is the reverse. I am using, I think, opt2 as the
> forward shift parameters. Anybody have any ideas.
>
> Cordially,
>
> Stan Childs
|