[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Top 10 /aberration



PureBytes Links

Trading Reference Links

Shane Wisdom wrote:
> 
> Mark Brown wrote:
> 
> "aberration same thing I have the original manual and now a latter one.
> of
> the 35 items in the manual tracked, 22 of them had their optimal
> parameters
> CHANGED! in under a 3 year period!"
> 
> Mark,
> Where did you get this information?  We've been running Aberration
> forward  for 5 years now.  While a commodity may have been added,
> subtracted, or swapped in a specific portfolio, at no time was any
> parameter EVER changed.
> Shane Wisdom
> Trade Center, Inc.
> 800 894 8194
> http://www.tradecenterinc.com


I don't have Aberration, only the literature sent by the author of the
program.  I'm wondering if the following isnt' the source of the
confusion.  Altho a list of optimum parameters was listed in the
literature I got the feeling that Fitchen(sp) rather strongly
recommended one generic value for all commodities. He went as far as to
include an article (from S+C?) which studied this (maybe using Ab.).  It
compared the performance of a portfolio with optimum parameters and then
a walk forward test with the perf. of the generic parameters.  The
optimized parameter results were somewhat *worse* in the walk forward
test, than the generic parameter.  (The idea of the article was that if
you have a system that can work pretty well accross many commodities
with one "average" parameter, you will probably have a more rugged
system if you use that generic parameter rather than a different
parameter for each market.  I'm aware that not all agree with this,
saying for example that different mkts have different "personalities".)
Conceivably the optimum parameters listed might have been changed
without changing the system since you werent "suppossed" to use them
anyway?  So the question may be, did aberration change it's
*generic/single* parameter recommendation.  (It was 80 in 1996 or 1997
so when I got the lit.)

That still does leave the question of how Futures Truth handles a
parameter change or other program improvement after the initial "release
date".  Someone else said they were going to ask FT directly; I hope
they reply.


Conrad Bowers