|
PureBytes Links
Trading Reference Links
|
Neo and Roy and Lionel,
The following is from Equis, confirming what you have been telling me.
" The indicator in MetaStock was approved By J. Welles Wilder himself so I'm
pretty sure it is right. Now it was created has a dll and not the MetaStock
language.
With the MetaStock language it is simply impossible to recreate this
indicator. Thanks, Patrick. "
Neo, would you expand upon (with code if possible for the new boy) your
suggestion that a simpler SAR can be coded. Also, the other elements of your
'mechanical system' would be appreciated. Concerning direction of position,
if C > yesterday's SAR point stay long, else go short.
Roy, you say " there is no problem writing a system or an exploration for
the SAR. " With my limited TA experience, I see the Parabolic SAR as
produced by the canned indicator in MS (and working with the variable
parameters provided) as sufficient for my needs. Could you help me with the
bones of an exploration to get me started?
Many thanks
David
----- Original Message -----
From: "neo" <neo1@xxxxxxxxx>
To: <metastock@xxxxxxxxxxxxx>
Sent: May 7, 2002 11:41 AM
Subject: RE: Parabolic SAR - Wilder's Original Definition
> David
>
> Wilder's SAR, to my knowledge, cannot be coded properly in the MetaStock
> formula language (it can in Excel). Since the same language is used in
> indicators, explorations, and system tests you will not be able to do it.
> You can, however, code a different SAR system that is simpler than
Wilder's.
> In terms of an exploration, how will you determine the direction of your
> position?
>
> neo
>
> -----Original Message-----
> From: owner-metastock@xxxxxxxxxxxxx
> [mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of DAVID H. LINTON
> Sent: Tuesday, May 07, 2002 12:26 AM
> To: metastock@xxxxxxxxxxxxx
> Subject: Re: Parabolic SAR - Wilder's Original Definition
>
>
> Neo, I believe MetaStock does follow Wilder's rules from a mathematics
> standpoint. I have, for a few stocks, calculated the SAR points following
> the 'B' rule you wrote and the numbers are accurate. That is why I said in
> my previous post " Accurate, yes, but correct, I'm not so sure. The key
> thought for the MS approach is that, with today's results, the SAR point
> that MS will plot on tomorrow's bar is calculated but not shown on the
> chart. It is not shown because it could change if the SAR line changed
from
> one direction to the other."
> It would appear that where, as you say, " Wilder designed this to give a
> price for the next trading day in real time", MS do not put the SAR point
on
> the chart until the next day is completed because it may be overtaken by
> rule 'C'.
> From a trader's standpoint, and please remember that this is a novice TA
> learner speaking, I conclude that the MS SAR indicator gives a clear
> indication of price approaching the point of change in trade direction,
but,
> it would be much more useful if today's EOD chart showed the "price for
the
> next trading day in real time".
> Roy, you wrote "I've had a brief look at it and I can see why you think it
> may be un-codable." Does the fact that MS have coded the indicator not
> suggest that an exploration, system test, etc are codable?
> This will be my first attempt at writing an exploration. Any guidance
would
> be greatly appreciated. I would think an exploration needs to first
confirm
> that the stock is trending and then signal (for long) stocks that have now
> changed from SAR down to SAR up.
> David
>
> ----- Original Message -----
> From: "neo" <neo1@xxxxxxxxx>
> To: <metastock@xxxxxxxxxxxxx>
> Sent: May 6, 2002 9:53 PM
> Subject: RE: Parabolic SAR - Wilder's Original Definition
>
>
> > David
> >
> > Wilder designed this to give a price for the next trading day in real
> time.
> > If the price is hit then the current position is closed at that price
and
> a
> > position is opened in the opposite direction at that price. Wilder did
not
> > design it as an EOD stop.
> >
> > Where I suspect that MetaStock does not follow Wilder is in rules B and
C
> > although I do not know since I do not have the code for their SAR.
> >
> > neo
> >
> >
> > -----Original Message-----
> > From: owner-metastock@xxxxxxxxxxxxx
> > [mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of DAVID H. LINTON
> > Sent: Monday, May 06, 2002 8:32 AM
> > To: metastock@xxxxxxxxxxxxx
> > Subject: Re: Parabolic SAR - Wilder's Original Definition
> >
> >
> > Neo, thanks for the SAR formula details from Wilder's book. This is what
I
> > was looking for in my original SAR post.
> > Roy, using these details I have checked out a few of MetaStocks SAR
> > calculations and conclude that their calculation method is accurate.
> > Accurate, yes, but correct, I'm not so sure.
> > The key thought for the MS approach is that, with today's results, the
SAR
> > point that MS will plot on tomorrow's bar is calculated but not shown on
> the
> > chart. It is not shown because it could change if the SAR line changed
> from
> > one direction to the other.
> > Perhaps, Neo, you can tell us if this is what Wilder intended?
> > Why not simply back the line up one day?
> > David
> >
> > ----- Original Message -----
> > From: "Roy Larsen" <rlarsen@xxxxxxxxxxxxxx>
> > To: <metastock@xxxxxxxxxxxxx>
> > Sent: May 5, 2002 6:18 PM
> > Subject: Re: Parabolic SAR - Wilder's Original Definition
> >
> >
> > > Hi Neo
> > >
> > > Thanks for the information. I don't expect to be able to get as far as
> you
> > > have even, but it's a challenge that I can put aside for a rainy day.
> I'll
> > > post results if I come up with anything.
> > >
> > > Regards
> > >
> > > Roy
> > >
> > > > Although I do not post much, my knowledge of MetaStock programming
is
> > > fairly
> > > > extensive. I would love to see you do it.
> > > >
> > > > I have Wilder's book here. It was written for the days of
handwritten
> > > charts
> > > > and I am amazed at his book.
> > > >
> > > > Roy, there were several problems as I recall:
> > > > 1. Initialization. I am not always in a trade in a particular
> security.
> > As
> > > I
> > > > recall, I used either the HHV or LLV (depending on long or short)
for
> 2
> > > > days.
> > > > 2. I was unable to program the rules for B & C in the MetaStock
> formula
> > > > language because of the lack of global variables and the lack of a
> > simple
> > > > If-Then statement instead of the If-Then-Else.
> > > >
> > > > My very best wishes. Let me know how you do or if there is anything
I
> > can
> > > do
> > > > to help.
> > > >
> > > > neo
> > > >
> > > > _____________________________________________
> > > >
> > > > Wilder's SAR
> > > >
> > > > FORMULA
> > > >
> > > > SAR[tomorrow]=SAR[today] + AF * (EP[trade]-SAR[today]
> > > >
> > > > DEFINITIONS
> > > >
> > > > SIP=the extreme price point while in the previous trade
> > > > AF=begins at .02 and is increased (depending on the rules below) by
> .02
> > > per
> > > > day until .20
> > > > AF is never increased beyond .20
> > > > EP[trade]=Extreme Price Point for the trade made so far
> > > > If long, the extreme high price for the trade
> > > > If short, the extreme low price for the trade
> > > >
> > > > RULES
> > > >
> > > > ENTRY
> > > >
> > > > A position is entered when a price penetrates the SAR
> > > >
> > > > STOP AND REVERSE (SAR)
> > > >
> > > > A. For the first day of entry, the SAR is the previous SIP
> > > > 1. If long, the lowest price while in the previous short trade
> > > > 2. If short, the highest price while in the previous long trade
> > > >
> > > > B. For the second day and thereafter
> > > > 1. Long - use the formula above but only increase the AF on days
when
> a
> > > new
> > > > high for the trade is made
> > > > 2. Short - use the formula above but only increase the AF on days
when
> a
> > > > new low for the trade is made
> > > >
> > > > C. Never move the SAR into the previous day's or today's range.
> > > > 1. Long - never move the SAR for tomorrow above the previous day's
or
> > > > today's low. If the SAR is above then use the lower low between
today
> > and
> > > > the previous day as the new SAR. Make the next day's calculations
> based
> > on
> > > > this SAR.
> > > > 1. Short - never move the SAR for tomorrow below the previous day's
or
> > > > today's high. If the SAR is above then use the higher high between
> today
> > > and
> > > > the previous day as the new SAR. Make the next day's calculations
> based
> > on
> > > > this SAR.
> > > > ____________________________________________
> > > >
> > > >
> > > > Neo
> > > >
> > > > > I have found it impossible to properly program the SAR as
described
> by
> > > > > Wilder with MetaStock's formula language. I can in Excel. I
believe
> > the
> > > MS
> > > > > indicator is not programmed properly.
> > > >
> > > > Could you possibly post the text for the SAR composition so I can
take
> a
> > > > look?
> > > >
> > > > Thanks
> > > >
> > > > Roy
> > > >
> > > >
> > > >
> > > >
> > > >
> > >
> > >
> >
> >
> >
>
>
|