|
PureBytes Links
Trading Reference Links
|
Harvey/Forum
You're correct about it being one's perspective
that makes a difference and using something like this falls in what one would
call a set up; and the weightings in the 9 month 12 month
ROC may be relevant to some extent for the long term trader like myself and
others. I found what I thought was
historical/experimental support for this relative scoring from the two Columbine
white papers who I believe it was Chuck led us to in the forum. There they
call it momentum and use a physical analogy of momentum...a body in motion,
tends to remain in motion.. trends tend to continue ...unless it hits bad
news! The Columbine service goes further and develops a quintile or
deciles stack, something I haven't tried yet, and sells the service to
institutions, the primary competitor or support for the long term
trader/investor.
I used the AA optimize function to lead me to the
"best" coefficients for the RSQP equation and I find that for the past trend in
small caps like the IBD 100, i.e.. you find if the equity curve improves for the
family average as you iterate the coefficients from the base over different
time windows 252 bars, 504 bars, etc. The base case being the
RSQP equation that gets posted with the 4 2 2 2.coefficients.. it
shifts most of the weight to the 3 -6 months section, If you tried to
'fit' this to last 508 days then weighting will spread.. I tried this on various
families and composite indexes and just now for RU2000, for past 252 bars I get
4.5, 3, 2, 0.5
while for 504 bars it's 6, 2.7, 0.65, 0.65...
shifting almost entirely to the best performing 3 and 6 months.
FWIW. These were two, 10,000 iteration optimizations in 4 minutes!! Ain't
technology grand!
But as Van Tharp says, the set up for trades
probably should be only 10% of your effort in developing a system, and further
that you probably should buy your setup or fundamental info from experts.
I found that in a simple trading system with a breakout entry, and end of
trend exits, that the use of RSQP to set Relative Score (without the deciles
stack) increase the returns significantly. That's for several different
approaches in systems I've had in development. Unfortunately none are
finished. I'm having too much fun with this. <FONT face=Arial
size=2>
When I asked my friend and resident expert about developing a deciles
ranking like IBD's RS, he said I'd need the Osaka plug in to do it, and since
that was Sat, I haven't taken the time to look it up to see the level of
difficulty of using this function. Looks like Ace's work can do this for
me also. Can anyone point me to past discussions on the plug
in? Yeh I know... search the archives! (:>!!
All of the above is for what it's worth and I hope it
helps, part of my learnings in the past 6 months. I also learned a buy
and hold small cap money manager beat me also.
JOE
PS. For our Greek friend..the quote of the day.
"Small opportunities are often the beginning of great
enterprises."
Demosthenes
<BLOCKQUOTE
>
----- Original Message -----
<DIV
>From:
harveyhp
To: <A title=amibroker@xxxxxxxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: Tuesday, January 06, 2004 3:07
AM
Subject: Re: [amibroker] Re: IBD
Interest
Phsst,
A few disjointed thoughts.
1. Isn't the choice of RS method
subjective? One person has a short-term holding period, another a longer
term. One person places a high value on the most recent activity,
another considers all of the activity over the whole year.
2. Taking four successive quarters ROC and
doubling up on the last quarter should be the same (after averaging) as taking
a one-year period plus 20% of the last quarter, should it not?
3. Overlapping 3-, 6-, 9-, and 12-month
periods back from the current bar weights the most recent quarter so
heavily that the earliest quarter will have little effect.
4. ROC considers only the values of the
first and last bar of the period and ignores what happens in between.
Wouldn't linear regression be more appropriate, since it considers all data
points within the period?
From Mark Boucher's 1999 book "The Hedge Fund
Edge", p.149: "Short term traders could improve results going to as low as an
eight-week RS . . . while longer term investors could widen the RS to as far
as one year, still with substantial improvements in returns and risk. We
settled on one of the lower risk paramters which turned out to be a weighted
average RS (similar to O'Neil's but slightly shorter in average
duration). The exact formula is (2 x ( 1-wk RS + 5-wk RS + 8-wk RS
+ 13-wk RS) + (26-wk RS + 40-wk RS + 52-wk RS)) / 11, which gives an average
of about 16 weeks, but is heavily front-end weighted."
HHP
================
<BLOCKQUOTE
>
----- Original Message -----
<DIV
>From:
Phsst
To: <A title=amibroker@xxxxxxxxxxxxxxx
href="">amibroker@xxxxxxxxxxxxxxx
Sent: 05 January, 2004 8:59 PM
Subject: [amibroker] Re: IBD
Interest
<What I do know is how IBD works and the QRS thingie
doesn't even comeclose.>Fellas,Just to keep things in
perspective...I personally don't know that there is a 'right' or
'wrong' way to goabout this.I appreciate reading each and every
method (and opinion) of how toapproach this type of
measurement.What you can 'make book on' is that I take notes of each
approachposted and will ultimately apply all approaches to specific
tradebacktest situations in order to see what works under
differentcircumstances.Until recently, I thought that QP2 QRS
rank was the single mostpowerful PositionScore that I had found. But
thanks to recentdiscussions, I have a whole new array of tools to
approachPositionScore with.I am thankful for all differing
opinions and contributions on what the'best' measurements
are.Thanks for the discussion along with all the
ideas.Phsst--- In amibroker@xxxxxxxxxxxxxxx, "Chuck
Rademacher"<chuck_rademacher@x> wrote:> Ok, Gary.>
> I guess that the translation may be correct and that the QP
QRSthing may> work, but I have trouble understanding the
relevance of theperformance of a> stock between 250 days ago and
189 days ago. But, what do I know? What I> do know is how
IBD works and the QRS thingie doesn't even come close.> >
Please don't get me wrong here. I'm not saying that what
wassubmitted is> wrong or that it won't work. I just
stated that, to me, it seemsillogical> and to call it IBD-like is
a bit inaccurate.> -----Original
Message-----> From: serkhoshian777
[mailto:serkhoshian777@xxxx]> Sent: Monday, January 05,
2004 10:55 PM> To:
amibroker@xxxxxxxxxxxxxxx> Subject: [amibroker] Re: IBD
Interest> > > Chuck,>
> The formula Joe has posted is a straight translation of
QP QRS> ranking. I certainly appreciate your
thoughts and modification to> the formula as I'll that as
well.> > Regards,> Gary>
> Here's info on QRS from the QP help file:>
> > Quotes Plus Relative Strength
Indicator> The Quotes Plus Relative Strength indicator is
calculated for each> issue each day, and becomes part of
the data you download to your> PC. The indicator ranks
each issue in the database against every> other issue in
the database, based on its performance for the last> 12
months.> > Every issue is ranked from 0 to 99 so
that its rank is equal to the> percentage of issues that
it has outperformed over the previous 12> month period. A
rating of 92, for example, means that the issue has>
outperformed 92% of all of the issues in the database.>
> The formula is :> ( Current Close /
Close from 62 trading days ago ) * .4> + ( Close from 63
trading days ago / Close from 125 trading days> ago ) *
.2> + ( Close from 126 trading days ago / Close from 188
trading days> ago ) * .2> + ( Close
from 189 trading days ago / Close from 251 trading days>
ago ) * .2> This value is sorted for all of the issues,
and the top 1% of issues> get a value of 99. The other
issues are ranked similarly, down to> the bottom 1%,
which get a value of 0.> > > --- In
amibroker@xxxxxxxxxxxxxxx, "Chuck Rademacher">
<chuck_rademacher@x> wrote:> >
Joe,> >> > I have serious
concerns about the formula that you have given:>
>> > RSQP =
(.04*Close/(Ref(C,-62))+>
>
.02*Ref(C,-63)/(Ref(C,-125))+>
>
.02*Ref(C,-126)/(Ref(C,-188)) +>
>
.02*Ref(C,-189)/(Ref(C,-251)))*100;>
>> > It seems to me that it should
be:> >> > RSQP =
.04*Close/(Ref(C,-62))+>
>
.02*Close/(Ref(C,-125)+>
>
.02*Close/(Ref(C,-188) +>
>
.02*Close/(Ref(C,-251);> >> >
The way you have documented it, you are looking at ROC's
over> various> > historical windows
instead of time periods all relative to today's>
price.> >> > I use this type of
ranking scheme all the time and am unable to>
understand> > why you have suggested doing the way you
describe. I also get> > significantly
better performance out of a system using the second>
approach.> > Of course, you may very well intend to do
it that way and it may> work for> >
you!!> >> >>
>> >> > -----Original
Message-----> > From: Joseph Landry
[mailto:jelandry@xxxx]> > Sent: Saturday, January 03,
2004 8:13 AM> > To:
amibroker@xxxxxxxxxxxxxxx> > Subject: [amibroker] IBD
Interest> >>
>> > Here's some code fragments from my
collection. I've used the> RSQP>
> momentum equation as a ranking number generator, and have
used> it on> >
family(watchlist) averages. It's a weighted rate of
change> equation.> >
The use of the .001 in the denominators make it seem esoteric
but> > that is to avoid dividing by zero
and I should have replaced it> with>
> the NZ function I have a technique where you can use
the> OPTIMIZE> > (read
curve fit since it has 4 coefficients) for any equity
but> > I think you should use coefficients
that are based on the> families>
> performance. Haven't used any of the other IBD like
criteria.> Which> > one
of those European mathematicians said if you give me one>
> coefficient I have a point, 2 a straight line and with
5> > coefficients I can draw any shape?
(;>! Hope I didn't mangle this> > one
too badly.> >> > By
the way a search in the AB archives for notes from Chuck
about> > the Columbine methods and
documents available will make an> >
interesting reading for you in this momentum equation area.
You> > probably get a lot more from
Columbine for 100K per year than a>
list> > of 1500 stocks ranked by momentum
but this was good background> for>
> me.> > Best
regards> > Joe Landry>
>> >> > --IBD RS
Screens.afl> > /// SCANS FROM TC
2000 forumlas> > // IBD Positive
Gainers> > Filter =
(Volume/(MA(V,50)+1)-1)*100>39 AND C-Ref(C,-1)>.5 AND
MA> > (V,50)> 70000 AND
C>5;> > VolFac =
(Volume/(MA(V,50)+1)-1)*100;> >>
> // Quotes Plus Relative Strength (IBD
like)> > RSQP =
(.04*Close/(.001+Ref(C,-62))+
.02*Ref(C,-63)/(.001+Ref(C,-> > 125))+
.02*Ref(C,-126)/(.001+Ref(C,-188)) + .02*Ref(C,-189)/>
(.001+Ref> >
(C,-251)))*100;> >>
> // Hi Flyers - no frills version>
>> > HighFlyers = ((Close
-Ref(C,-250))/(Ref(C,-250)+.001) + (Close ->
Ref> > (C,-63))/(Ref(C,-63)+.001)+ (Close
- Ref(C,-21))/(Ref(C,-21)> +.001))>
> *100;> >>
>> >
AddColumn(C,"Close");> >>
> AddColumn(VolFac,"% Vol Delta");>
> AddColumn(RSQP,"RelStr");>
> AddColumn(HighFlyers,"High Fly");>
>> >> >>
> Send BUG REPORTS to bugs@xxxx>
> Send SUGGESTIONS to suggest@xxxx>
>
----------------------------------------->
> Post AmiQuote-related messages ONLY to:
amiquote@xxxxxxxxxxxxxxx> > (Web page: <A
href="">http://groups.yahoo.com/group/amiquote/messages/)>
>
-------------------------------------------->
> Check group FAQ at:> > <A
href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html>
>> >> >>
>
------------------------------------------------------------------->
---------> > --> >
Yahoo! Groups Links> >>
> a.. To visit your group on the web, go
to:> > <A
href="">http://groups.yahoo.com/group/amibroker/>
>> > b.. To unsubscribe
from this group, send an email to:>
>
amibroker-unsubscribe@xxxxxxxxxxxxxxx>
>> > c.. Your use of Yahoo!
Groups is subject to the Yahoo! Terms> of
Service.> > > > Send BUG REPORTS to
bugs@xxxx> Send SUGGESTIONS to
suggest@xxxx>
-----------------------------------------> Post
AmiQuote-related messages ONLY to:
amiquote@xxxxxxxxxxxxxxx> (Web page: <A
href="">http://groups.yahoo.com/group/amiquote/messages/)>
--------------------------------------------> Check group
FAQ at:> <A
href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html>
> >
>---------------------------------------------------------------------------->
--> Yahoo! Groups Links>
> a.. To visit your group on the web, go
to:> <A
href="">http://groups.yahoo.com/group/amibroker/>
> b.. To unsubscribe from this group, send an
email to:>
amibroker-unsubscribe@xxxxxxxxxxxxxxx>
> c.. Your use of Yahoo! Groups is subject to
the Yahoo! Terms ofService.Send BUG REPORTS to
bugs@xxxxxxxxxxxxxSend SUGGESTIONS to
suggest@xxxxxxxxxxxxx-----------------------------------------Post
AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx (Web page:
<A
href="">http://groups.yahoo.com/group/amiquote/messages/)--------------------------------------------Check
group FAQ at: <A
href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Links
To visit your group on the web, go to:<A
href="">http://groups.yahoo.com/group/amibroker/
To unsubscribe from this group, send an email to:<A
href="">amibroker-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to the <A
href="">Yahoo! Terms of Service.
Send BUG REPORTS to
bugs@xxxxxxxxxxxxxSend SUGGESTIONS to
suggest@xxxxxxxxxxxxx-----------------------------------------Post
AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx (Web page: <A
href="">http://groups.yahoo.com/group/amiquote/messages/)--------------------------------------------Check
group FAQ at: <A
href="">http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Links
To visit your group on the web, go to:<A
href="">http://groups.yahoo.com/group/amibroker/
To unsubscribe from this group, send an email to:<A
href="">amibroker-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to the <A
href="">Yahoo! Terms of Service.
Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html
Yahoo! Groups Sponsor
ADVERTISEMENT
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 the Yahoo! Terms of Service.
|