One of the cycles I use has actually a basic 56 days, but the count vary, it is not fixed in etheral emptiness, and it starts anew every so often (note that the January 31st to February 28th, was exactly half such a cycle), there is also a "stretch factor" that depends on exogenous events which I have to determine from prior experience, market psychology and sentiment indicators, and the worse, is that the average cycle, while being 56 days now, used to be 62 days just three years ago, so I think that there is what I have often called on these threads "information dissemination compression" of natural cycles. Note for instance, that April 16 is 35 days from March 12th (the most probable bottom), and is the first Fibonacci factor of the time cycle of 56 days. The big trick of course, is to find the stretch factors, determine if the first or second Fib would be more appropriate, and nail down floating cycle starts appropriately. When getting to repeat events (like short retest of bottoms), a typical interval will be 9 to 10 days or 6.2 roughly of the main cycle), but even these are modified by the nature of the pattern. Too complex (and unreliable) to really go into details of the cycle elements of the forecast, particularly, that the cycle elements is only a first cut on timing, not direction nor target prices (like the forecast of 2250 for the top between the March 1st and Match 9th bottom)
One of my errors last year was fixing the election as a floating cycle start point (that led me then to call for a strong post Christmas rally , I had the swoon into the 20th all right, since it was based on another floating cycle start that was more accurate, but the transition to the new floating start was more difficult. As long as the January 31st cycle start remains more or less "controlling" (and the further you are from that fixed float date, the worst it becomes), calls are relatively "good". The funny thing is that the January 31st fix turned out to be erroneous, it was supposed to be the "first fed directional change", well, they changed on Jan. 3rd...go figure, it still worked, so why change it?...
Zeev |
