|
PureBytes Links
Trading Reference Links
|
> > > To build a good trading system, you need to find something that happened
> in the recent past that correlates well with what the price is going to do
> in the future. This is fundamental. Without it, it is just gambling.
>
> markets have personality, each market has its own personality, choppy and
> trending differences. each market has a foundation, each foundation has
> different properties even though the properties are similar. it would be
> foolish to expect all market properties to be the same just as it would be
> foolish to expect all people are the same. similar properties still provide
> the end product that makes up the foundation. there are good foundations
> and bad ones, bad ones crack over time and good ones hold up. time is the
> key element (or is it?) that will decide what is correctly built or not.
> how then can we know without experiencing the time elapse whether or not a
> system will hold up? testing - there again we have good testing and bad
> testing. a proper way to build a system and an improper way to build a
> system.
>
The paradigm for my exploration of the physics of the market is
the historical developements of physics, starting with Tycho Brahe
and proceeding through Johannes Kepler and Issac Newton. Yes, Mark,
each market has a personality --- just as each planet observed by
Brahe and whose motion was determined by Kepler's rules has its own
personality. Men are from Mars, Women from Venus, and Pluto rules the
depths of the oceans. Nonetheless, the personalities don't move
the planets one iota! IT'S the PHYSICS of the situation that lets
us determine where in the sky the planet will next appear.
> experience, can be obtained by massive amounts of computer generated data,
> only if that data is not mush going in. how to mine for the proper data is
> a mystery that will remain to those who refuse to believe it can be done, so
> be it. however i would caution you that what the human mind perceives as an
> impossible task on a computer - and only thought obtainable by the human
> mind. are the very processes that a properly programmed computer data
> mining process involves, only to a higher degree of accountability.
>
In that Data Mining is my perspective on the world:
(http://apk.net/~lake/datamining is a 100+ slide seminar I give)
I can only agree that mush-for-data gives fuzz-for-results. It takes
the human mind to realize a few simple things, sometimes it takes
thousands of years to grasp these simple things ...
The Morning Star is The Evening Star
The Earth and Planets Move Around the Sun
Planets Sweep Out Equal Areas in Equal Times
And in modeling the market, we don't even yet realize how to describe
what drives a market comprised of a vast variety of buyers and sellers.
`
> accountability? the human mind can rationalize away that which does not fit
> the mold or the expected outcome. the poor honest computer knows nothing of
> this process and refuses to deliver anything but that which is true. if you
> can not withstand the test of time then you need to avoid the
> confrontational outcome of accurate computer system modeling. where does
> this leave us?
>
> all this work done and to perfection, still leaves a weak link in the chain.
> we allow our eyes and our emotions to rule over that which is false, not
> relying on the hard facts. instead we try and overpower the data with
> reason and logic that surely must be correct, why! how! can we not trust the
> innate responses that our very being is intertwined with? destruction of
> truth, throughout the ages - to tear down that which is just and true. to
> scoff at that which is not of the ordinary, only latter in years to see it
> become a dominate force.
>
> properties+testing+accountability+experience= mechanical system trading
> success
>
> properties+testing+accountability+emotions= mechanical system destruction
>
Amen -- and proven in my own "laboratory" ...
> mb
>
>
Cheers,
Rob Lake
rbl@xxxxxxxxxxx
|