Re: fuzzy finance


Subject: Re: fuzzy finance
From: Robin B. Lake (rbl@hal.EPBI.CWRU.Edu)
Date: Fri Feb 25 2000 - 15:32:29 MET


>
> Dale Johnson a écrit
> >Perhaps you should use the same method that actual investors use.
> >
> >They listen to the media, people in chat-rooms, and analysts who
> >say one thing, but do another. They play hunches, feel fear and
> >greed.
> >
> >See, if you use your very proper and logical methods, you are saying
> >that the average investor is proper and logical. >
>
> Thanks again to feed the debate.
> Well, who knows if it is the average investor that makes the market ?
> I have strong doubts about that, at least these days (although that would
> put my own theory temporarily into parenthesis, and I should replace my
> "image" coefficient by a "manipulation" coefficient).

Having built a system which seems to successfully predict short term
market movements and having embodied some Fuzzy principles in the system,
http://junior.apk.net/~lake
let me offer some slightly different perspectives:

- There is an entire spectrum of market-moving decision-makers. Some
small, some large. Obviously, the market is almost always itself larger
than the largest traders and always much larger than the small traders.

- There is a dynamic to the market. Decisions are made by traders on
time scales ranging from seconds (Day Traders) to weeks (decision by
committee).

- One can develop indicators and predictors that elicit Fuzzily
accurate indicators of short-term price movements using nothing more
than Open, Last, High, Low and Volume measures, along with a matrix
of historical identical measures. Your mileage may vary when it
comes to profitably trading these indicators.

- By analogy, market prediction is slighly beyond where Tycho Brahe
was in astronomy (LOTS of observations) and slightly behind where
Johannes Kepler was in the physics of astronomy (Rules, applicable only
in specific physical situations --- planetary orbits). We're a
LONG way from Issac Newton (Laws, generally applicable in ALL physical
situations).

- The markets are NOT efficient. Not everyone does have the same
information IN THE SAME TIME DOMAIN.

- The markets are non-linear. Twice the input does NOT produce
twice the output.

- There are HUGE exogenous forces that influence markets: Greenspan.

- The quantitative study of markets sorely lacks reported scientific
results. If one starts with a linear system perspective and begins
to apply the tools thereof: impulse functions, efficient basis functions,
ergotic analysis, etc.; it is almost impossible to find prior studies
to support acceptance or denial of these approaches.

- The dynamics of market INDICATORS (which MIGHT be more linear) also
lacks scientific results. The linear system perspectives (in our VERY
primitive analyses) seem to offer (1) insight and (2) more indicators.

I have several years of data at 1-minute and 15-minute sample intervals
for three US stock indices (OEX, DOW, NDX) along with about 200 indicators
and predictors at each sample interval. Analysis will be, as they
say, "the subject of a future analysis".

> Well, for the sake of the debate, let us suppose it is the average investor.
> There is an expression in French theater plays (I suppose there is an
> English equivalent in Hollywood, but sorry, I don't know it) that is
> "comique de repetition".
> Well, it applies to human behavior, and thus to stock market behavior.
> Since the tulip bubble of the 17th century, there are repetitive phases of
> overpricing and underpricing.
> Of course, it is not logical, but where you see ruts you can measure their
> width (range of price), and identify the vehicle (current paradigm of
> evaluation).
> Playing the market is an art and a sport, not a science, I fully agree.
> And the good old Markowitz / Sharpe CAPM (Capital Assets Pricing Models) is
> full of holes and sinking.

To say nothing of the Nobel-winning Black-Scoles Option Pricing Model ...

> But to draw a few lines on the ground, which is what my own model tries to
> do, could be of some use :
> - if you are doing Short Term trading, to see when the ball is going off
> limits and ask yourself if it is not the moment to take a fast decision.
> - If you are doing LT assets management to make price comparisons between
> various stocks (and other assets), and make your money allocations between
> them
> Well, what is said for ST can apply to LT, and the other way round, another
> thing where I don't see any dogma, but common sense.
> PG

If the markets were ergotic, one would expect that the concatenation
of ST models would create an LT model. This certainly has NOT been
our experience. I have yet to do the analysis to determine what the
time-extent of a model prediction might be. There is a "physics of
models" to be discovered here.

>

Cheers,
Rob Lake
Environmental Modeling Inc.
rbl@hal.cwru.edu
(216)-721-2800
 

############################################################################
This message was posted through the fuzzy mailing list.
(1) To subscribe to this mailing list, send a message body of
"SUB FUZZY-MAIL myFirstName mySurname" to listproc@dbai.tuwien.ac.at
(2) To unsubscribe from this mailing list, send a message body of
"UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL yoursubscription@email.address.com"
to listproc@dbai.tuwien.ac.at
(3) To reach the human who maintains the list, send mail to
fuzzy-owner@dbai.tuwien.ac.at
(4) WWW access and other information on Fuzzy Sets and Logic see
http://www.dbai.tuwien.ac.at/ftp/mlowner/fuzzy-mail.info
(5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html



This archive was generated by hypermail 2b25 : Thu Apr 06 2000 - 15:59:46 MET DST