One question was recently asked about the methodology behind our FibTime algorithm. Without going into too much detail I can certainly share some of the concepts behind it.
- we define pivots (highs & lows) and weight them in a Tom DeMark sense by looking how many days they took to develop and held. Example a '5' weighted Pivot took in min. 5 days to develop and held at least 5 days. (in order to keep it simple I define classes like 5's, 8's, 13's etc).
- take a pair of pivots, weight them by their average individual weight and count the days btw. pivots
- calc. the basic Fib-ratios like 61.8%, 100%, 138%, 162% and add these calculated days to the last pivot (also look at Robert Miner's books here). I take calendar days for the ease of calculation.
- repeat for a finite number of pivots and cluster (and smooth). Don't go too far back, don't miss the big ones.
