I am comparing stoch values in MetaStock and Q+ and noticed big differences..
Hi -
The stochastics indicatoer is like the cci in that different programs uses slightly different %k smoothing methods.
I've tested Metastock, WOW Pro 5, and TC2000. All of them are different, WOW and TC2000 both seem to be wrong.
The WOW documentation states "you could plot %K and then put a 3-day moving average of it on the plot and that would be %D."
The TC2000 documentation states somethiong similar - the the %k smoothing value creates a simple moving averag of the %k.
The %D smoothed is a moving average of the smoothed %k i.e. a moving average of a moving average.
That is how our stochastic is implemented.
When you create a "Stochastic Oscillator simple" with WOW, it asks for three parameters, number of %k periods, number of %d periods, number of slowing periods.
With WOW, if you ask for a 14,1,1 stochastic, you get 2 lines the same, and the third is a 2 day moving average of the %k. So even though I ask for a 1 day smoothed %d, I get a 2 day smoothed %d.
With QP, a 14,1,1 gives a smoothed %D that equals the %k value. When you ask for a 1 period smoothing value, you get it.
With TC2000, the underlying fast stochastic is calculated incorrectly, the numbers don't match at all with the stochastic formula- and you can't seem to display the actual values of the stochastic, just look at the chart.
With Metastock, the underlying fast stochastic is calculated the same as WOW and QP, but they smooth it differently, they say "If you enter a value greater than one, MetaStock will average the highest high and the lowest low over the number of %K Slowing Periods before performing the division."
Anyway, our stochastic matches the WOW and TC2000 documentation, their implementations don't match the documentation.
Our fast stochastic agrees with Metastock and WOW, it is smoothed differently than Metastock's.
I hope that helps, I don't know the right answer - I too have been confused about stochastics, and now I can see why.
Gary |
