|
PureBytes Links
Trading Reference Links
|
At 2:40 AM -0400 6/21/00, Kent Rollins wrote:
>There are so many techniques out there. Pick whatever works for you.
>The key is to find something that you can master, study it
>thoroughly, and keep working with it enough to discover when to buy,
>when to sell and when to sit out. If you're just starting out, the
>most important thing to learn is not to lose your stake before you
>learn to trade. Of course, that's important after you learn to trade
>too.
>
>...so put me down for "No but they're not worthless"
The fact that technical analysis can work is no longer in much doubt.
Professors who have tried to prove that markets are a "random walk"
with prices determined by the "efficient market hypothesis" have
failed to prove it. Most now acknowledge that at least short-term
patterns exist that can be used to extract a profit from the markets.
I think there are techniques that work because of good technical
reasons and there are techniques that work because people think they
SHOULD work and that then becomes self-fulfilling.
An example of fundamentally sound technique include momentum - the
market will tend to continue in one direction for a while until it
goes too far and then head in the other direction. This behavior is
characteristic of any negative feedback system with a delay in the
feedback loop. And probability distributions tend to be "normal"
because of the "Central Limit Theorem" - the way the mathematics of
combining random variables work out.
Support/resistance lines are a good example of techniques that work
only because people think they work. People hate disorder so tend to
create order where none exists. So if enough people think that the
price will bounce off of the 50-day moving average, they will tend to
buy there and make the prediction self-fulfilling.
I think Gann stuff, fib ratios, astronomy, Elliot Waves, Dow Theory,
etc., are all of this latter category. If you come up with a
technique, publish it, become a guru, and sell enough books, you may
be able to convince enough people that your technique works well
enough to have enough people try it to make it REALLY work.
Lots of common indicators probably fall into this category as well.
Bollinger Bands are based upon the mathematical fact that with a
"normal" probability distribution, 95% of the occurrences will be
within two standard deviations of the average. That is mathematics.
But the fact that all charting packages now contain a Bollinger Band
indicator and millions of people can see when a price is about ready
to bounce off of the Bollinger Band, the combined action of all those
people will tend to make the price bounce there. That is psychology.
But the bottom line is that the price will tend to bounce there so we
might as well make money off of the phenomena if we can!
I did hear a theory once that prices did tend to gravitate toward fib
ratio points for some fundamental reason based upon some numerical
effect but I have never been able to find a convincing explanation of
why. It is well known why fib ratios work in nature - from the shape
of leaves to the shape of sea shells - but none of these reasons
extend to market pricing as far as I can see.
But if you discover something that works - enjoy it while it is working!
Bob Fulks
|