RE: [EquisMetaStock Group] Optimizing

["Gordon Sutherland" <gosuth@xxxxxxxxxx>]

<span
>David,

<span
> 

<span
>An interesting question you ask re
optimizing single securities. I am a stock picker rather than a futures trader
or specialist on a few securities. Personally, I prefer not to optimize for
single securities for risk of curve fitting but if you need to because that is
the market you trade all the time then at least system test over as much back
data as possible and or use &#8220;walk forward&#8221; techniques to check the
consistency of the optimized system.

<span
> 

<span
>Because I am a stock picker, I only system
test using software that will do full portfolio testing which tests a basket of
shares and gives consolidated results BUT must use risk and money management
rules to simulate real world trading. Then when I optimize across say 400
securities over both bullish and bearish phases of the broad market cycles I
have more confidence that the consolidated performance data is robust &#8211;
normally the results are not as good but at least I feel more realistic. You
may wish to download a trial of TradeSim for your MetaStock as this is an
excellent product if you want an &#8220;out of the box&#8221; solution. I have
no commercial interests in any publisher or software house.

<span
> 

<span
>These are my views but I think this is a
very important subject that this forum should discuss. I would recommend you
read Robert Pardo&#8217;s book &#8220;Design, Testing and Optimizing of Trading
Systems&#8221; John Wiley and Sons, Inc 1992. This book is specific on the
subject whereas the excellent books by Perry J Kaufman&#8217;s are longer and a
little more generalized (IMHO).

<span
> 

<span
>Well that&#8217;s my tu&#8217;pence worth
so what does everybody else in this group reckon?

<span
> 



<span
>Regards,

<span
>
Gordon Sutherland



<span
> 

<span
>-----Original Message-----
From: David
[mailto:junk@xxxxxxxxxxxx] 
Sent: Monday, 24 May 2004 4:05
a.m.
To: equismetastock@xxxxxxxxxxxxxxx
Subject: [EquisMetaStock Group]
Optimizing

<font size=3
face="Times New Roman"> 

<font size=2
face="Courier New">I thought I would post
something that has been rattling me lately.  <font
size=2 face="Courier New">
What is the consensus about optimizing systems for
individual 
securities?  I tended to stay away from this
practice as it seems 
more ideal to find a system that performs better
across entirely 
different markets and securities, because the
average is probably the 
most consistent.  However, backtesting seems
to show that it is 
possible to show consistent results when
optimizing for individual 
securities.  I would assume that individual
securities do contain a 
certain element that makes them perform similarily
over time.  When 
looking at GBP/USD it is obvious to see that this pair
trends in a 
different way then does USD/JPY.  So I'd like
to know some other 
opinions on what people think of individually
optimizing.

Best Regards,
David















Yahoo! Groups Sponsor


  ADVERTISEMENT 












Yahoo! Groups Links
To visit your group on the web, go to:http://groups.yahoo.com/group/equismetastock/ 
To unsubscribe from this group, send an email to:equismetastock-unsubscribe@xxxxxxxxxxxxxxx 
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.