The bandpass is the most powerful of the filter types we can exploit using the simple moving average. Learn how to design your own bandpass in this article
True. In practice I use wavelet convolution as you may know so it's not so much of an issue if one has enough data. I also use Butterworth filters for the superior freq response. One thing you can use multiple MAs on is FLDs. Since they are plotted in the future you can smooth it many times up to nowtime!
You can also curve fit the end of your multiple filter (MA) and extrapolate it to now time, the smoother the better!
How do you achieve such smooth output from moving averages?
Hey Mack, if you continue to filter the MA with a further MA(lpf) then frequencies greater than half the span of the cutoff continue to be attenuated
I see, but then with every further step I am loosing data on both ends of the signal. How many times do you usually repeat the filtering?
True. In practice I use wavelet convolution as you may know so it's not so much of an issue if one has enough data. I also use Butterworth filters for the superior freq response. One thing you can use multiple MAs on is FLDs. Since they are plotted in the future you can smooth it many times up to nowtime!
You can also curve fit the end of your multiple filter (MA) and extrapolate it to now time, the smoother the better!
Heh, I also prefer Butterworth filters to MAs :)
They are not really conducive for beginners though and most trading platforms have the humble MA
This series about digital filters is awesome. Thank you! it would be great to have the anticipated upcoming aricles as well :)