Personal Finance

Carl · Guest

And Now for Something Completely Different: Swarm Technology(TM)

====================
Honesty Statement:
--------------------
Yes, I'm a vendor and we sell financial markets analysis software
For those who don't know us, we are a respected vendor in this area for more
than 10 years
Our customers range from retirees to the most respected advisory firms.
====================

Technical analysis has gone through various evolutions over the years, with
all sorts of "interesting" methods: Candles, Angles, Ultimate-this and
Fibonacci-that, neural networks, expert systems, astrology, etc. as people
quest for the "ultimate" trading machine. I personally believe that an
"ultimate trading machine" that works on all or a large majority of
securities does not exist, but that's slightly off topic.

Anyway, we decided to do something COMPLETELY different:

1. Develop a technology that is VERY crudely similar to a trading floor
(pools of trade-bots)
2. A technology that is continually adapting to changing conditions because
the markets are not stationary
3. A technology that does not "curve-fit" as most call it here
4. A technology that walks-forward through a security's data showing nothing
but "Out-Of-Sample" (hypothetical) results, the good, bad and ugly, because
way-too-many software tools just fit the data with no validation, or make
validation difficult.
5. A technology that is easy to use... like 3-4 mouse clicks.
6. A technology that is suitable to investing (infrequent transactions) to
trading to intra-day trading

To do this, we invented BioComp Swarm Technology(TM). Imagine a pool of
trade-bots that collaborate like a flock of birds to seek profitable values
of the trade-bots' parameters (like a walk-forward collaborative
equity-performance adaptation of the periods of a moving average cross-over
system, or a swing trader's buy/sell thresholds or whatever kind of
trade-bot it is).

Well, swarms are not really a new technology. The general concept has been
around for many years, written about in papers and books, but I don't think
anyone has applied it to trading and investing in commercially available
software that I know of. Thus I think we are the first. (please let me know
if you know otherwise, by sending your comments to "tech" at our domain
(http://www.biocompsystems.com).

We did some quick Swarm Technology experimentation in our labs in September
that looked promising. We formed a project code named "Dakota" under which
we created a prototype product, then a beta product with different types of
"trade-bots" and put it into the hands of some our best customers. The
response was a unanimous "thumbs-up". Does it work on all securities? NO.
Different kinds of bots work well with different securities, as is to be
expected. Some bots are slow traders for investing, some are faster for
trading. Do swarms of bots that are collaboratively adapting their periods
beat a fixed trading system of similar nature? Generally, yes. That means
that the swarm approach does add value, although we and our customers are
still exploring and we are still developing.

To read more about BioComp Dakota(TM) with Swarm Technology(TM), please head
to:
http://www.biocompsystems.com/products/Dakota/index.htm

There are some out of sample hypothetical equity curves to look at here:
http://www.biocompsystems.com/products/Dakota/equitycurves/index.htm

Swarm Technology(TM) is not the Ultimate-Anything, it's "Something
Completely Different".

Thanks for your time. I'm looking forward to your comments. Send them
privately, if you like, to "tech" at our domain name.

Sincerely

Carl Cook
Chief Technical Officer
BioComp Systems, Inc.
Minneapolis, MN
http://www.biocompsystems.com

ynotssor · Guest

	Personal Finance Forum Index -> Market Trends	All times are GMT
Page 1 of 1