|
PureBytes Links
Trading Reference Links
|
I want to amplify on comments Mark Brown made in response to Bob Heisler's
post. I feel Bob's comments could dangerously mislead new traders on this
list, and lead to a lot of pain and damaged bank accounts.
Hello Bob,
BH> For one it's a lot more fun as I like making my own decisions. It
BH> is also much more satisfying.
MB>good for you.. Bob
I spent a year paper trading a discretionary "system" with a known guru. It
was a completely disillusioning experience. The guru sent faxes claiming to
have made great trades, but I could never learn what cues from our common
indicators he was "really" using to get off these great trades. Questioning
him was frustrating; he could never communicate clearly. In retrospect, the
claims in the faxes were fraudulent; the guru never got off the great
trades he claimed, either. A conversation with another student who spent a
year doing the same, revealed that she too was never able to grasp the
guru's methods; he always came up with exceptions to his rules, that had
never been previously discussed that he claimed explained his great
decisions, and great trades. Don't repeat our frustrating experience.
BH> Systems are also subject to drawdowns, and sometimes severe ones,
BH> I am unwilling to accept those.
MB>are you saying you have no drawdowns? I think that's what your implying
MB>after reading below that you expect to never have a loosing day!
With a carefully designed and back tested system, you will see that in the
real world, drawdowns happen. The advantage with systematic trading is you
can ascertain ahead of time their size, and be prepared for them when they
come, and trade through them successfully. If you have spent 3 months
trying to design a system for your time frame and market, and it has
unacceptable drawdown, you know this ahead of time, and you haven't traded
it and lost money! You've just lost some time. The danger with
discretionary trading is that you will think "you're off your game", and
quit trading because you ascribe the draw down to some failure of your own.
Or you will continue a bad "discretionary" system and loose more money.
And if you really don't know what you are doing (i.e., being a discretionary
trader), your drawdown could get really large.
BH> If you can read the price action and execute there is no reason
BH> to have a losing day (barring order entry errors).
MB>GET REAL! what planet do you live on?
A claim to some undefined power to be able to "read price charts" will
never get to a set of clear rules, in code, that can be tested and refined,
or possibly rejected. And the belief that one can magically pass this on to
students is equally ludicrous.
Bob, can you imagine noting the relationship between breast tumor size and
patient demographics by "price reading" the raw numbers we got from 15
hospitals and hundreds of patients? We published the paper, and found the
important relationships by studying Pearson product moment correlations that
came from statistical and computer analyses of our data!
After two years of collaboration between an AI programmer, and the nation's
leading human internal medicine diagnostician (they argued a lot during
those years!), guess who the world's best internal medicine diagnostician is
now? It's an AI program at Carnegie Mellon called MYCIN. Before long,
every MD will carry a laptop with MYCIN and its counterpart for all other
branches of medicine with him/her to the patient's bedside. To avoid AI,
will mean a lawsuit. Gary Fritz cited the AI research from the 70's and 80's
that shows the revolutionary import of AI for almost all branches of
knowledge. A classically coded EL system is an AI system too.
NN - FL AI is here for trading now: Wardsystems.com and Sirtrade.com. These
packages along with classical programming code, can induce rules from data
and indicators, and produce explicit, testable, refinable, tradeable systems
that can detect many non-obvious relationships that you can not see (much
less verbalize) with your eye-brain hookup.
AI that you develop through more traditional "linear" classical programming
in EL, or in TW with VB and C++ may be equally good, or even superior.
BH> And any system has to be based upon somebody's opinion. I have
BH> already learned that my opinion s-cks and the market couldn't care
BH> less about it, so I am unwilling to trade someone else's opinion
BH> or my own. There is more than one way to trade successfully, so in
BH> the end we just have to find what works for us.
MB>Then Bob how is it that you are going to teach your students to trade
MB>if each has to find their own way? Your web site makes for very
MB>interesting reading.
During my 25 years of teaching medical students to learn to become
diagnosticians, and and in one case, to recognize the signs of overt
psychosis from explicit
diagnostic criteria (which they have in their hands on a sheet of paper),
the ludicrous nature of Bob's statements should be apparent to all. How can
an effective teacher say: "my opinion s-cks" to his students? How can an
effective teacher then turn and say: "you have to find your own way"? Can
you imagine turning loose my green residents in a hospital of
psychotic patients, and saying to them: "You each have to find your own
way?" How many of my residents would emerge the next morning alive?
Bob, it is the teacher's responsibility to make what he is teaching clear,
applicable, productive and useful. If the students aren't learning, and
can't apply what you teach, and what you teach is not explicit, then you
should stop teaching, and most important, stop making claims about the
efficacy of your "price reading" methods.
The traders I have encountered who claim to be teachers of discretionary
methods, such as your self, have never had any solid, lengthy, supervised
training and
experience teaching at the undergrad or grad level. They have never been
graded by their Chair and by their students, and have never had to respond
to feedback for ways to improve. Have you?
All knowledgeable professional traders trade concrete, testable, replicable,
refinable (yet imperfect) coded systems. Read the list or book writings of
Charlie Wright, Jake
Bernstein, Mark Brown, Pierre Orphelin, Bob Fulks, Mark Jurik, Kent Rollins,
Chuck LeBeau, Gary Fritz, Clyde Lee, Perry Kaufman, Linda Raschke, Thomas
Stridsman, Cynthia Kase, Murry Ruggiero, and yes, even Larry Williams, and
many more. The
big financial institutions spend millions to get to explicit code and/or AI
systems to manage
big money; do you think for a minute they will take chances with their
client's money and give it to discretionary traders? Do you think Clinical
Supervisors will take chances with the diagnostic ramblings of a Pediatric
Intern and a five year old with a 105 degree fever, when a laptop with AI is
sitting on the table? Be serious!
Charlie Wright makes the point that when you sit down in the morning to
trade, your job is to execute your classically coded or NN FL AI system.
You have enough problems with
trade execution and hassles with your broker ahead of you for the day.
You'll face fatigue and trading loosers. You
should never have to worry about whether you should take a trade, what your
charts mean, whether your guru sees the same thing you do, where you will
exit, what size you will trade, whether the
moment to call your broker has come.
BH> Bob
MB>Mark
Peter
|