Hi CatLady, I guess I should have chosen my
wording more carefully. I didn't mean to imply that I had some carefully tested, bullet-proof method based on using Fib set members for time periods. I was just floating out another possible experiment to do. The "always" was only meant in the context of consistency--in other words, if you don't always use that method, you really can't judge how well it performs. I'm a believer in (developing a method for)setting indicator time periods based on cyclic behavior (peak to peak, trough to trough, and duty cycle), rather than arbitrarily. I'd be willing to bet, however, that a significance could be shown (for using Fib numbers) if for no other reason than the closeness of 55 and 34 to probably the number 1 and number 3 moving average periods (50 and 30 day, with 200 being number 2 (this is just a guess by the way)used by most market participants). I also believe in the fractal/scaled replica arguments for stock prices, but like anything, it can't be shown 100% of the time (it's not PURR-fect). That's the basic reason though, for thinking the use of Fib set members as the short, medium, and long term indicator parameters can exploit the scale independence seen in some stocks better than randomly selected periods. If you have any insights/experiences/beliefs with regard to Fibonacci retracements or Fibonacci time zones, please, meow away!
dh
P.S. you might check out www.fibtrader.com if you want some real Fib-hype! Actually, an impressive looking package. |
