Replies to post #6837 on MDA - Market Direction Analysis (MDA)

[WinLoseOrDraw]

that's a unique way to describe it, but yes, that's what i mean (and i'm still only assuming it's what LG meant.)

appropriate window size is the eternal question, kind of like asking what the "correct" stoch settings are. try a few values, see what seems to work, then pick one that gets it wrong every now and then so you increase your odds of not merely curve-fitting.

[LG]

augieboo: That would be a "way" of thinking of it...gg Of course with the sum of the moving window not being divided by the number of periods in the window, it is not an average.

Yes, that is what I had in mind and WLorD is right that would be an attempt to reduce bias. Of course the results will reveal whether the effort was worth it.

Back in the eighties, I created three oscillators that are comprised of a mix of data which are then input into two MACD formulas. One is a current moving window and one is an exact duplicate of the first formula but run X periods in the past. The results of both moving windows are part of a larger MACD formula. I worked with the IRS capital gains long-term buy and hold tax break duration to determine the X period. Of course during the period when the long-term capital gains tax break was eliminated, the model lost its effectiveness.

Regards,
LG