Re: GEN: Parabolic

[Bob Hunt <RHunt.066@xxxxxxxxxxxxxxxx>]

I have found Robert Krausz's Triple Switch indicator to have similar
attributes as the parabolic, but seems to work better for my purposes
- it just seems to provide a tighter stop when needed. I generally use
it on 5 minute bars (.ela attached)

TSUP=Average(H,3)[1];
TSDN=Average(L,3)[1];

If Pos=1 and C<=TSDN then begin
   Plot1(TSUP,"TSUP");
   Plot2(TSDN,"TSDN");
   Pos=-1;
End;

If Pos=-1 and C>=TSUP then begin
   Plot1(TSUP,"TSUP");
   Plot2(TSDN,"TSDN");
   Pos=1;
End;

If Pos=-1 and C<TSUP then Plot1(TSUP,"TSUP");
If Pos=1 and C>TSDN then Plot2(TSDN,"TSDN");

------------------------------------------------

Peter Ryan wrote:
> 
> Conrad,
> Very interesting thoughts - thanks.
> 
> I am using Parabolic with intra-day breakout systems using 3-minute bars of
> index futures.
> I am not too concerned with the gap at the beginning of the trade.  I use a
> fairly tight stop-loss initially, then let the parabolic stop kick-in once
> it has advanced past my entry price.
> 
> What I liked about it was its seeming ability to get out near the top of a
> move.   I havent found any other indicator which gets me out at a top so
> nicely.
> 
> Thanks again for your ideas on fiddling with the parameters.
> Peter Ryan
> 
> -----Original Message-----
> From:   cb [SMTP:cpbow@xxxxxxxxxxxxx]
> Sent:   Sunday, October 18, 1998 9:10 PM
> To:     pryan@xxxxxxxxxxxxxx
> Cc:     RealTraders Discussion Group
> Subject:        Re: GEN: Parabolic
> 
> Peter Ryan wrote:
> >
> > I have started using Parabolic as a trailing stop in a couple of my
> systems.
> > I dont see it mentioned very often in books or on lists like this.
> >
> > Is there some reason it is not popular.
> >
> > Thanks
> > Peter Ryan
> 
> Maybe because it was originally presented as a stop-and-reverse system.
> Without some kind of filtering I don't think it tests well as a stand
> alone system except in the trendiest of mkts.  Lebau and Lucas do
> suggest using it with ADX as a filter.
> 
> I think it's a pretty good choice for a stop (exit only), particularly
> if you experiment with the parameters.  If your software allows it, try
> changing both the "initial %", (the % of (H-par[1]) that it moves up on
> the second day of a long), as well as the acc. increment (the amt the %
> increases  with each new high).  You can make it less parabolic and more
> like a curvy trendline, displaced a little ways from the data, if you
> use higher initial% and a lower increment.  I didn't like the very slow
> initial response of the parabolic with default parameters of 2%, 2%.
> Using 5-10%, and about 0.5% seemed to make a better stop that follows
> the trend well, gets fooled into a premature stop rarely, and, on medium
> to longterm trends, doesn't give up too much of the gains before exiting
> on true reversals.  What it doesn't do well on, even with modified
> parameters, are short term profits at the beginning of the trade - if
> that's all that develops you will likely give them back.  Of course this
> is true of any stop that gives the trade room to "breath".  I suspect
> the big gap between the mkt and the parabolic at the beginning,
> particularly with the default parameters, is another reason it is not
> popular.
> 
> [I haven't succeeded at this trading thing yet, so take all comments for
> what they're worth.]Content-Type: data; name="Tripswch.ela"
Content-Disposition: inline; filename="Tripswch.ela"

Attachment Converted: "c:\eudora\attach\Tripswch.ela"