Re: Are There Any Successful Fibonacci Or Elliott Wave Traders Out There


Awaken21 <lukecarlos@xxxxxxxxx> wrote in message
news:1178026522.509158.279010@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Apr 30, 10:40 pm, "Bill Reid" <hormelf...@xxxxxxxxxxxxxxxx> wrote:
Awaken21 <lukecar...@xxxxxxxxx> wrote in message
news:1177932653.441430.87810@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Apr 29, 3:05 pm, "Bill Reid" <hormelf...@xxxxxxxxxxxxxxxx> wrote:
Awaken21 <lukecar...@xxxxxxxxx> wrote in message
news:1177848784.396482.29890@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Apr 28, 4:44 pm, "Bill Reid" <hormelf...@xxxxxxxxxxxxxxxx>
wrote:

I'd love to know more about this. My perception is that ALL
market
strategies (with maybe ONE exception!) have been developed using
a
combination of wishful thinking, innumeracy, and penniless
blathering,
without a shred of legitimate statistical supporting evidence
being
brought
into play before the strategy is "developed" (and yes, I am
including
O'Neil's "CANSLIM" in this)...after the strategy is out there
for a
while, there HAVE been some conflicting studies done by
academics
that may or may not show some actual modest efficacy...

Good place to start...
Mechanical Trading Systems: Matching Trader Psychology with
Technical
Analysis (Wiley Trading) Richard L. Weissman

Maybe I'll look at it a little when I go the bookstore next week...

As an update (I know, who cares?), I couldn't find the book at the
Barnes&Noble "lending library" so really can't comment further on
it, since I definitely ain't buying it and apparently the people who've
read it lack the ability to describe it...

I'm always looking for the answer to the question "how do I make
the most amount of money possible?", but this book seems to answer
the question of whether my personality is better suited to craps or
playing the lottery...the tired old dodge of "you have to trade a
system
that you feel comfortable with"...

Sort of but not really. It presents the advantages and disadvantages
of different types of trading, based on timeline, win/loss rates,
maximum number of lost trades, max number of consecutive losses, etc,
and attempts to guide the reader towards development of a system they
can live with.

!!!!

That's exactly what I said!

For example, not everyone has the psychological ability
to take 40 consecutive losses and stick with thier system, but as a
trend trader you might have to at some point....

Hmmm, so do "trend traders" make MORE MONEY than other
traders, or why the hell else would you take 40 frickin' losses?!??!!
Man, you must make a boat-load of money on that 41st trade to
make up for all those losses on a numerately risk-adjusted basis!!!

Yes, trend traders make about 80% of their gains on less than 5% of
their positions...according to Weiss and if I remember correctly of
course..

OK, so you ARE saying that "trend traders" make the MOST amount
of money, or not?

I'll take any kind of answer here, just pure "expectation" if that's all
you've got, even though that's not the actual canonical answer...does
some specific "trend-trading" strategy make the MOST money of
all the tested strategies? (If you don't know, in this case, "expectation"
would just be the arithmetic mean of all the simulated trade results
for a strategy.)

But does it ever answer my bottom-line question?

I think it does, but maybe not, you'd have to read the book.

I was pretty sure before, but now I have great confidence, the
answer is "no, the book does not tell you what strategy makes the
most money". But I'm still willing to be proved wrong...

I know
that you and the author speak the same language.

Does he ever use the words "optimal wealth growth factor" or
something to that effect? Cuz that's the bottom line for me, a little
idea that's been around for about five decades now...

OK, I'll take the non-response as a "no" to that question as well...

For any simulated trade series you should generate a range of
results in the form of a trade duration- and frequency-adjusted
histogram. This data can then be transformed into a single number
that you can use to compare the strategy that produced those
results with all other potential strategies. The bigger the (positive)
number, the more money you make using the strategy...PERIOD.
END OF DISCUSSION.

Here's why you can't use "expectation" alone to determine the actual
money-making potential of a strategy. You mentioned that the author
"proved" that "trend-traders" lose money on a lot of trades, and only
make money on a few trades, though you failed to say what they
actually won or lost overall on average (the "expectation") for any
particular trade strategy.

You call this "psychology", to me, it's just MATH. And here's some
math on the subject...

Let's start with a 1% "expectation", that is, you make on average
1% on each trade. First, let's try a slightly-biased coin toss, you flip
a coin a 1000 times, and heads comes up slightly more than tails,
and you bet on heads:

For a return histogram of...
Data points: 495 Distribution: 0.495 Average result: -1.0
Data points: 505 Distribution: 0.505 Average result: 1.0
The expectation is 0.01
The optimal wealth growth factor is 0.00005 for a 0.01 trade size

What this is saying is that you win 10 more flips out of a thousand
betting on heads, giving you an "expectation" of 0.01 (1% average
"profit" on each flip). We then run this histogram through our
optimal logarithmic wealth growth calculator, which perfectly
balances "risk" and "reward", and it tells us that we should bet
0.01 (or 1%) of our net wealth on each flip, and this will result
in a wealth growth factor of 0.00005, which we can use to compare
against other possible "bets" (or "trades") available at the time.

But what happens if we have a really messed-up coin that only come
up heads 1% of the time, but as compensation we get paid 100:1 if
it does?

For a return histogram of...
Data points: 99 Distribution: 0.99 Average result: -1.0
Data points: 1 Distribution: 0.01 Average result: 100.0
The expectation is 0.01
The optimal wealth growth factor is -0.000037 for a 0.001 trade size

Note that we still have the same 0.01 (1%) "expectation" as the
first coin game, but now we have a NEGATIVE (meaning we actually
LOSE money) wealth growth factor of -0.000037 when betting 0.001
(1/10th of 1%) of our net wealth (this was the minimum bet size when
I ran these numbers).

This indicates the "lottery problem": that even if the jackpot gets so
high in the lottery that you have a positive "expectation", it STILL
doesn't make any sense to buy a single (or a few) lottery ticket(s).

Of course, "psychology" is always that "hey, it's only a $1, and if I
win I win $millions". Problem is, the MATH says you're almost always
just throwing your dollar away...if you want to actually make money,
you have to use MATH and not "psychology", which apparently this
book of yours, and just about all others of a similar nature, is all
about...keeping that dream of a "big score" alive...

Now for a 1% expectation strategy that maybe is similar to a
TA trading strategy, consider this:

For a return histogram of...
Data points: 3 Distribution: 0.03 Average result: -0.18
Data points: 7 Distribution: 0.07 Average result: -0.12
Data points: 12 Distribution: 0.12 Average result: -0.08
Data points: 16 Distribution: 0.16 Average result: -0.04
Data points: 22 Distribution: 0.22 Average result: 0.04
Data points: 18 Distribution: 0.18 Average result: 0.06
Data points: 12 Distribution: 0.12 Average result: 0.08
Data points: 7 Distribution: 0.07 Average result: 0.1
Data points: 3 Distribution: 0.03 Average result: 0.12
The expectation is 0.01
The optimal wealth growth factor is 0.006985 for a 1.0 trade size

In this case, you can safely trade ALL of your net wealth on a single
trade, and probably even more (leverage), but I had the software cut
off at 100% when I ran this. And the good news is, if this hypothetical
strategy can be traded on most of the days of the year many many
time a year, you can "compound" your wealth much more nicely than
it would appear from a mere 1% advantage (as "Jack Hershey" used
to like to remind people endlessly)...but as always, you have to KNOW
what your results are LIKELY to be going into your trade, otherwise
you're just "throwin' the dice"...

Does he present such a number for all the "building blocks" he
tested for the book? If not, I have a slightly larger but significantly
more useful vocabulary than he does...

Every day I run a program that looks at gigabytes of all kinds
of historical economic, fundamental and technical data for the entire
world economy and world trading markets of all sorts, and then a few
hours later it spits out a "top 100" list of the trades that will make
the most money, be it Forex, commodities, futures, stocks, whatever.

If I wanted to make it run a LOT quicker, I could confine myself
to price-only US stock data using the proven BEST TA strategy,
but I'm quite sure I wouldn't make anywhere near as much money...

I will grant you that "psychology" is a MAJOR problem for "traders",
but I have to wonder at what point does "psychology" end and dogged
stupidity begin?

Well, let's see...your interested in something that is at it's base
purely statistical, you are being presented with an field of market
study littered with like minded folks (mechnical TA), and your
thinking you might not take the first step because you know everything
you need to know by reading the chapter titles of one book...

That would be a curious reading (and snipping) of the post you are
responding to, where I specifically stated I had already performed the
testing of the hairy old TA strategies years ago...in fact, I've said
that
several times in this thread.

That's a valid point. Problematically most people make mistakes when
their backtesting which resulting in apparent random returns,
particularly on indicators.

Why would a "mistake" result in apparently random returns? I've
certainly made my share of mistakes, and quite often they lead to
gigantic apparently non-random results...

So I'd have to know for each indicator
what your parameters were, what you used as signals, for example I
know at least 4 different ways to read signals on the MACD, since
you've tested it I'd love to know what you used as trade signals.

Well, I could dig all this stuff out, if you were serious...I mean if
you actually read the archives of this group you'd actually see that
I posted the actual code that I used to generate certain results.
I have no problem doing that again, and even more extensively,
if there was any REAL interest in the topic...as a matter of fact,
I've even offered my entire trade simulator code as "freeware"
here.

But I kinda get the feeling that's not what people really want, they
really want to "keep hope alive"...

On
moving averages theres many other ways to get signals, like did you
use a price crossover or angle of line, or both.

I've done it all baby.

What about
stops...what were the stop numbers incorperated, if any,

I've used all kinds of different stop strategies...once again, I've
actually posted my results in this group for various stop strategies
on various TA strategies...it's really sounding like you are completely
defensive and grasping at straws here, which is sad, really...typical
but sad, because the TRUTH isn't such a scary thing...unless you've
built an entire real or imaginary life-style based on "rolling the dice"
without any actual supporting mathematical evidence...

which
indicators did you filter one against another etc, and most
importantly and I'd have to go behind you and redo what you did to
have any confidence in it. Which I'm not going to do.

Of course. As I always say, it's not a matter of INFORMATION,
it's all there, it's how you chose to INTERPRET it (or in most cases,
ignore anything that might scare you) that determines who you are
and what your results will be...

: I took "the first step" decades before
this guy ever had the idea of charging $125 to apparently NOT tell you
which of the hairy old TA strategies actually makes you the most money.
Since I already know, I have no reason to buy a book that DOESN'T
tell me...

My only question was about any and all strategies that had been
developed based on PRIOR statistical research, NOT after-the-fact
backtesting of those hairy old TA strategies that you paid $125 for
the privilege of calling "building blocks". Since it appears there are
still ZERO publically-available strategies DEVELOPED from long-term
statistical profiling of market price-only data, my question has been
answered, no matter how indirectly (OK, maybe you might check out
Ben Stein's "Yes You CAN Time The Market", if only for giggles)...

Hmmm, where does psychology and dogged stupidity draw the line between
them? Great question.

I am nothing if not eminently (or erratically) quotable, eh?

You're the island, you're the man, have a lovely day.

You too.

---
William Ernest Reid
Post count: 592



.

Relevant Pages

  • Re: Are There Any Successful Fibonacci Or Elliott Wave Traders Out There
    ... you must make a boat-load of money on that 41st trade to ... I'll take any kind of answer here, just pure "expectation" if that's all ... The optimal wealth growth factor is 0.00005 for a 0.01 trade size ...
    (misc.invest.stocks)
  • Re: Trading guitars or other stuff
    ... If you do a trade...let's just say a guitar for a guitar...how ... money agreed on the seller's bank account after which they send the goods. ... you know you can trust most people here. ... regularly sell and trade stuff. ...
    (alt.guitar)
  • Re: justice, justice... Was: Anti-Semitism in the Neo-Confederate movement.
    ... Trade is trade. ... Trade builds wealth. ... >>Nothing magic, sure, but it's still leagues ahead of simple barter. ... One of my definitions of money is "a proxy for, ...
    (soc.culture.jewish.moderated)
  • Re: arod
    ... He's killing the Yankees, not only ... With the amount of money they owe him (not the full deal - just ... latest trade idea. ... They have Pettitte and Clemens coming off the books ...
    (alt.sports.baseball.ny-yankees)
  • T.J., we have to build some trust first, when and where can we meet ?
    ... to locate every offer to buy with every offer to sell ... It must check every trade lead on the planet... ... When you show the bank your 17 cent alluminum, and customs cost, ... A rental is a money maker. ...
    (sci.physics)