Thanks Mike.

I'm not using trade delays but I am using exrem. However, the 
barssince funcion was only used to remove more complicated sell rules 
to try and isolate the fault. It is still buying when the stoch is 
falling or above 50 - and surely the code as is should re ruling this 
out?

You say avoid Exrem - is this in all cases or just for fixed bar 
exit? Unless I'm exiting after a fixed period, is exremspan better?

I'm really confused....:-(

Thanks for any further advice....anyone?

Rich




--- In [email protected], "Mike" <sfclimb...@...> wrote:
>
> A couple of things to consider:
> 
> 1. Are you using trade delays in your AA settings?
> 2. If using BarsSince, you should probably get rid of redundancies 
> first.
> 3. Avoid ExRem except for charting. Use ExRemSpan in this case (see 
> doc explicitly describing how to hold for 'x' days after buy).
> http://www.amibroker.com/guide/afl/afl_view.php?id=50
> 
> Mike
> 
> 
> --- In [email protected], "MAVIRK" <mvirk67@> wrote:
> >
> > b = ref(stochk(3,2),-1) < 50; //and stochk(3,2) > 50;
> > 
> > 
> > From: foxblade2000invest 
> > Sent: Friday, February 27, 2009 5:21 PM
> > To: [email protected] 
> > Subject: [amibroker] simple system coding help....
> > 
> > 
> > Could anyone help.
> > 
> > I'm playing with the following idea. The system should buy if the 
> cci 
> > is above zero, the stochk has risen since yesterday and was below 
50 
> > yesterday, and the close should be above the open.
> > 
> > My code is as follows;
> > 
> > a=CCI(40)>0;
> > b=50>Ref(StochK(3,2),-1);
> > d=StochK(3,2)>Ref(StochK(3,2),-1);
> > e=C>O;
> > 
> > Buy=a AND b AND d AND e;
> > BuyPrice=C;
> > Sell=BarsSince(Buy)==1;
> > 
> > Short = 0;
> > 
> > Cover = 0;
> > 
> > Buy=ExRem(Buy,Sell);
> > Sell=ExRem(Sell,Buy);
> > 
> > The problem is that when backtested, trades are clearly taken 
when 
> > the stoch if falling and was above 50 yesterday (breaking the 
> rules). 
> > The only rules being observed are that the cci is positive and 
the 
> > close is higher than the open (a and e).
> > 
> > Can anyone suggest what basic mistakes I'm making?
> > 
> > All contributions welcome.
> > 
> > Cheers,
> > Rich
> >
>


Reply via email to