|
PureBytes Links
Trading Reference Links
|
Real Traders, think about this...
We do not expect one trader's methods to automatically work for
another trader.
Why?
Because of differing personalities, what works for one does not
work for another.
So, for the same reason, we can expect that my methodology would
not work as a mechanical system.
Why?
Because the software does not have my personality either. The software
will tend to blindly trade my rules. And it will tend to ignore subtle
variations in my decision-making process, because the software will be
much less flexible.
Here is an example..
Much of what I trade is pattern-driven. We would expect software to be
able to identify a pattern as well as a human can. But it can't. Think
of an elephant. Can software scan through a thousand images and present
a list of every image that contains an elephant? No! there are too many
variations, elephants have different colors, sizes, and they stand/sit
or lie down in an infinite different ways. Some elephants are partially
obscured by trees. Some have distorted mirror-images reflected off water.
No software can identify every elephant image. Software will make many
errors in such an exercise. But a human can do very well at this (not 100%).
It's very similar with chart patterns, they can be the same pattern, but
there are an infinite number of variations.
For this reason, current technology cannot be realistically expected to
duplicate the trading of every discretionary trader.
Also, for this reason it is very difficult to learn how to trade the
same way someone else does. And it's silly to try and duplicate someone
else's performance perfectly. Each trader must adapt what they learn to
their own environment...
-Neal.
PS. Private email indicates to me that Mark Brown is a great guy.
PPS. Now I've really said enough on this topic! Sheesh! What a waste of time!
Apologies... Hopefully this is of benefit to someone.
At 07:29 PM 12/7/99 -0600, mike wrote:
>
>Gang,
>
>Just to chime in a moment ... I analyze professional commodity traders
>(CTAs) for a living and there are a lot of successful mechanical traders
>and, unfortunately, not as many succesful discretionary traders. This
>doesn't mean that a discretionary method is an "excuse for inadequate
>programming" -- far from it - in fact, for anyone who has ever traded any
>real money knows, emotions run rampant when you trade and it takes nerves
>of steel not to let those emotions interfere with your market analysis.
>Few people can do it - but that doesn't mean that no one can do it. Take a
>look at Bell Fundamental Futures, IXORCAP (part discretionary), IIU Global
>Financial, Denali Asset Management, Kottke Grain, Light Blue Trading,
>Friedberg and the list goes on- very succesful programs that together run
>hundreds of millions of dollars with established track records. And we
>won't even go into the hedge fund world, where there are a lot of
>discretionary stock and bond traders who can put up better risk adjusted
>numbers than any CTA currently available to the public.
>
>Now let's look at some mechanical traders -- especially those who try to
>mimic the discretionary methods of experienced traders -- and two come to
>mind: Roy Niederhoffer, who happens to be down about 35% this year, and
>Gateway Financial, a European program that attempts to mimic discretionary
>bank traders -- which just recently closed its doors in order to reevaluate
>its research. Both of these trading programs have demonstrated long
>periods of achievement and are very innovative - I'm not trying to rub salt
>into any wounds - but they've gotten hit just as bad this year, if not
>worse, than many of the discretionary traders they try to mimic. Would you
>say their coding is inadequate?
>
>In my opinion, discounting the possibility of successful discretionary
>trading is extremely myopic and reveals a lack of knowledge when it comes
>to the world of professional alternative investments programs.
>
>Just my two cents.
>
>Mike
>
>
>
-----------------
Neal on the 'net.
Trade well. Train hard.
http://www.halcyon.com/neal/
|