To: CatLady who wrote (17 ) 1/22/1998 12:42:00 AM From: ftth Respond to
Hi CL: I guess if we look at it (the indicator) like a filter for the primary cycle, and assume the primary cycle is triangular-shaped, then a moving average, for example, with a window length the same as the cycle period would leave us with a low pass filtered version of the price plot but in sync (a sine wave of the fundamental cycle frequency). With the window length at half a cycle, our MA is lagged by a quarter cycle of the primary cycle. The problem I have is that most indicators I look at eventually lead me back to the basics, in that I conclude that I could read the same thing or better just from the price plot, trendlines, and volume--without the indicator, and it's seeming like that might be the case here. The cycle would have to be visible and measurable on the chart in the first place in order to conclude that there is one, and if we can measure the cycle, we don't need an indicator to tell us where the next one should happen. We know what to expect the price and volume characteristics to do, and roughly when (I do not believe that the exact turning points happen exactly at Fib-spaced days--just close), so we'd look for these to take shape, and as they begin to, the probability that the prediction comes true gets higher, until the event (the cycle turn) happens. Somewhere in this range, we place our bets, ride it through our comfort zone, then bail at a profit. Oh if it were just that easy! (Actually it may be on certain stocks that show uncanny cyclic behavior, and enough volatility to make a good chunk of money...never know til ya try right?) So I think I'd answer no, we wouldn't want to use Cycle=indicator length. I'll have to think about 3 or 4 Fib numbers down from the cycle length. Agree? Disagree? Totally unclear explaination?
