|
PureBytes Links
Trading Reference Links
|
Rick
Thanks for that, and for a new way (for me at least)
of calculating percentages: quite neat.
The idea is based on an article in a local newsletter.
The basis is to use multiples of SD of the mean of the
price (effectively a Mov Average) to avoid market
noise, but has to be expressed as a percentage of
price movement. The theory behind this is beyond me
but has to do with the need to use Ln and I quote " We
use percentage price change in place of price change
due to the frequency distribution of the data.
Frequency distribution of price change resembles a log
normal distribution and causes problems due to its
asymmetry around the mean; this is caused by the fact
that stocks don’t have negative values.
Using percentage price change gives it a more normal
distribution and is symmetrical around the mean.
"
The article goes on to give MS code to be placed in
the indicator builder for the initial stop as
initially posted, and this as the moving average of
that to be the trailing stop to be inserted into the
indicator list.
"
Name: VMOVSTOP:
Formulae: Periods:= Input("periods",5,30,15);
M:= Mov(((C/Ref(C,-1))-1),14,E);
SD:= Stdev(((C/Ref(C,-1))-1),14);
R:= M-(2.364)*(SD);
{R at 2.364 will give a 99% confidence rating}
Stop:=Mov((C*Exp(R)*1),periods,E);
If(C>Ref(C,-1),Stop,PREVIOUS)
"
I thought that the C*Exp(R)*1 had something to do with
the reverse of the Log normal calculation. The format
for AB is
SYNTAX exp( NUMBER )
exp( ARRAY )
RETURNS NUMBER,
ARRAY
FUNCTION Calculates e raised to the NUMBER or ARRAY
power
... which I thought may be different from the MS
format for Exp function.
So far for AB then I have the following:
*****
x = (C / Ref(C,-1)) - 1;
M= EMA(C,14);
SD = StDev(M,14);
Rlower = M - (2.326 * SD);// for 99% confidence
interval of not being stopped out initially.
Rupper = M + (2.326 * SD);
//Rlower = C * exp(Rlower);// clearly this value is
wrong.
Plot(C,"c",colorBlack,64);
Plot(M,"emaC_14",colorPink,1);
Plot(Rlower,"lowerRange",colorBlue,1);
Plot(Rupper,"upperRange",colorGreen,1);
//Plot(Rlower,"LowerRange",colorIndigo,1);
*****
but these don't quite seem right as you have
mentioned.
ChrisB
--- ricko8294_98 <ricko@xxxxxxxxxx> wrote:
>
> Chris
>
> I think your interpretation is correct, except that
> the line
> R = M - 2.326 * Sd should be
> R = M + 2.326 * SD
>
> The minus 1 just removes the value to the left of
> the decimal in the
> division. For example, if C = 35.00 and Ref(C,-1) =
> 34.75, then C /
> Ref(c,-1) is 1.00719. Subtracting 1 leaves you with
> .00719 - the
> percentage difference between the 2 values.
>
> Also I don't understand the "*1" at the end of Stop.
> Why bother.
> You also mentioned something about Ln of the price
> change. I don't
> see this in the code anywhere. Are you sure the MS
> code is accurate,
> and complete? The Stop values produced by this code
> is about 3 times
> the close - something seems wrong
>
> Rick
> --- In amibroker@xxxxxxxxxxxxxxx, kris45mar
> <kris45mar@xxxx> wrote:
> > Hi Group:
> >
> > Can anyone give me a steer in the right direction
> with
> > this Metastock code, please?
> >
> > Designed to be a volatility based stop, based on
> Stdev
> > of the Ln of the percent price change.
> >
> > M:= Mov(((C/Ref(C,-1))-1),14,E);
> > SD:= Stdev(((C/Ref(C,-1))-1),14);
> > R:= M+(2.326)*(SD);
> > {R at 2.326 will give a 99% confidence rating}
> > C*Exp(R)*1;
> >
> > Here is what I have so far:
> >
> > x = c/ref(c,-1) ; // what is the ((C/Ref(C,-1))-1)
> > part of the MS code. Is this the value for
> yesterday?
> >
> > M = ema(x,14);
> > Sd = stdev( x,14);
> > R = M - 2.326 * sd;
> > Stop = C * exp( R ) * 1; is this correct for AB?
> >
> > Regards
> >
> > ChrisB
> >
> >
> >
> >
> > __________________________________
> > Yahoo! Mail - PC Magazine Editors' Choice 2005
> > http://mail.yahoo.com
>
>
>
>
__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com
------------------------ Yahoo! Groups Sponsor --------------------~-->
Put more honey in your pocket. (money matters made easy).
http://us.click.yahoo.com/r7D80C/dlQLAA/cosFAA/GHeqlB/TM
--------------------------------------------------------------------~->
Please note that this group is for discussion between users only.
To get support from AmiBroker please send an e-mail directly to
SUPPORT {at} amibroker.com
For other support material please check also:
http://www.amibroker.com/support.html
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/amibroker/
<*> To unsubscribe from this group, send an email to:
amibroker-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|