As do the rest of the 95% of "traders" who eventually fail.
st
Williams %R
- Thread starter Pekelo
- Start date
As do the rest of the 95% of "traders" who eventually fail.
st
The problem is that
when you define the profile
of the noise you want to filter out, it changes.
Slippery business, isn't it?
The neural net inside your head is capable of adapting faster
than any other adaptive filter you may want to use.
(Let's also stress here that the indicators/filters we started
talking about in this thread aren't even adaptive...
Pick your parameters, and live with them.)
Looking at the failure rate of 'traders', it becomes clear that approximately 95% never "understand" that principle. I'm simply amazed at how many people follow the losers before them by thinking they have somehow found the secret to an indicator.
st
Even a blind squirrel finds the occasional acorn.
st
ST, understanding Kalman filtering gives one a unique understanding of the mathematical nature of indicators.
IF:
the signal in price could be described by a physical model,
the noise in price could be described statistically,
and the noise in the measurement of price could be described statistically,
THEN one could create a multi-dimensional filter which would estimate the future value of the signal component of price to the greatest theoretically possible accuracy.
A corollary to the theory of Kalman filtering is that EVERY mathematical operation (like an ma) on a time series (like price) is an ideal Kalman filter for some potentially infinite number of combinations of the above three conditions. The sad part of indicator invention is that the people who do this don't realize that. Every indicator is a perfect solution to an unknown problem.
IF:
the signal in price could be described by a physical model,
the noise in price could be described statistically,
and the noise in the measurement of price could be described statistically,
THEN one could create a multi-dimensional filter which would estimate the future value of the signal component of price to the greatest theoretically possible accuracy.
A corollary to the theory of Kalman filtering is that EVERY mathematical operation (like an ma) on a time series (like price) is an ideal Kalman filter for some potentially infinite number of combinations of the above three conditions. The sad part of indicator invention is that the people who do this don't realize that. Every indicator is a perfect solution to an unknown problem.
Already discussed at length with concensus that it provides no tradable edge.
http://www.elitetrader.com/vb/showthread.php?s=&threadid=49502&perpage=6&pagenumber=1
st