|
PureBytes Links
Trading Reference Links
|
Walter
Now you're starting to use our terminology. Probabilistic, indeed. <G>
That technique is what our complete system is based on.
We have developed our own indicators (patterns, if you will) and the
probability of success of our system (over time), knowing that we can never
be 100% correct.
Each one of our trades has a general probability of success of >75% with the
points gained being double the points lost. This probability is based on
trading over the last (almost) 17 years, so it has withstood the 'test of
time'.
My brother just completed his second money management analysis of our system
and has looked at each trade (in terms of present value since some trades
might be as short as 2 days and another trade might be 15 days). I just got
the spreadsheet last night and once I've had a chance to examine it, I might
post it here to see what people think. There is the possibility that we're
employing overkill here, but he thinks that it further refines our
probabilities. I'll let you know.
What is a little frightening to me is that in his first pass, we found that
with out, then current, trading techniques, we were running a 100% Risk of
Ruin. No wonder we were always going broke (well, at least almost <G>). To
limit our Risk of Ruin, the book suggested investing no more than 1/3 of our
capital and that would limit our Risk of Ruin at <1%. With this new
approach, the suggested investment level is 70%. I'm not sure I'm
comfortable with that number and want to take a long look at his
calculations.
How can one technique be so different from another. Granted, the first
approach is much easier and uses graphs from the author's book and the
second one takes a lot more work to develop. I'm not sure that more work
corresponds to a better answer. At 70% invested, we are higher than our
previous level of 50% (plus specials) which has a guaranteed Risk of Ruin of
100% (and which we're proven more than a few times).
Oh well, I guess I have a lot of stuff to do today in addition to going to
the dentist to have a crown started.
Regards
Guy
-----Original Message-----
From: owner-metastock@xxxxxxxxxxxxx
[mailto:owner-metastock@xxxxxxxxxxxxx]On Behalf Of Walter Lake
Sent: Friday, April 23, 1999 4:08 AM
To: Metastock bulletin board
Subject: bar patterns
Hi Marie
You asked what I was going to do with the bar patterns?
Enclosed is the start of an workbook using standard relational notation for
data mining bar patterns. The bar patterns and code that I posted are the
conventional usage. Both forms need to be researched.
What I'm looking for, is what Rough Set Analysis is about: Applying
data Analysis with a view to get some probabilistic statements. This is in
contrast to common Technical Analysis where you start with some indicator
(also qualifying market conditions) and then optimize parameters or add more
indicators. Personally I think, data mining to be more appropriate, though
by
no means easier, just the opposite.
Best regards
Walter
|