I'm about to make a "fibonacci moving average". It is a weighted MA in which each subsequent number is weighted dividing by the golden ratio: 1.618... Many other indicators can be fibonacci weighted.
Sounds good, I think you're on to something here. I've tried several times to look at ways of automating fib ratios (which would allow me to stay in trades which are moving in my direction), but I've never succeeded. So I hit'em up for a buck or two (2pts, $100 ES as an example) and then call it a day - just being honest here. Looking forward to the screenshots. Good trading, Jimmy Jam
then you will find fibonacci perfect--always a ratio or a MA of ratio that act as support/resistance....
I recommend that you study the frequency response of that MA, it's phase response, and it's zero-pole plot. On a theoretical level, you should also understand the Fourier Transform of the FIR. An MA is a simple FIR filter. Here is a place to get you started understanding them: http://en.wikipedia.org/wiki/Finite_impulse_response and Digital Signal Processing tutorial here: http://www.synctrading.com/phpbb/viewtopic.php?p=215#215 I am not saying that you should not use it. I am just saying that you should understand everything (at least everything mathematical) about a model (in this case a filter) before sending it out on the market live. nitro
A study of the issues cyclic properties. Amplitude, period and phase would serve you much better in choosing a MA period. Or maybe i`m misunderstanding???
An educated study has been done in the past about Fib (for example, retracements.) They concluded that the Fib numbers had ZERO probability as compared to other retracements (such as 50%) in providing more profitable trades. But this is not surprising. The believers go on, in spite of evidence.