It is strange! I really like it as a filter on certain subwindow indicators because it removes fluctuations.
Many analysts argue (especially in Futures) that ADXvma is just an "Exponential Moving Average" or a "Modified VIDYA (VMA)" because there is no direct link to the ADX formula. I disagree, as the VIDYA is considered "the nephew of the ADX" and there's also a bit more to it.
The ADXvma has the SMA in Chande's CMO replaced with an EMA, it then uses a few more layers of EMA smoothing before the "Volatility Index" is calculated.
A side effect is, those additional layers slow down the ADXvma when you compare it to the actual VIDYA (VMA).
However, the good thing about it is once there's consolidation detected, the ADXvma stays flat for longer. When it eventually signals to the upside or downside, the entries are much stronger and more accurate than any traditional EMA or VIDYA (VMA).
This is why many indicators really do "look" good with the ADXvma filter applied to them.
Here are a few indicators that you may look at that have been referenced above:
- ADXvma by Mladen: ADXvma
- Chande's VIDYA CMO by Mladen: VIDYA (VMA)
- Chande's CMO Oscillator by Mrtools: Chande's CMO Oscillator
Seeing that ADXvma has lagging properties, one can only imagine if the EMA smoothing of ADXvma was replaced with the much more superior Jurik smoothing instead!
The official description of the ADXvma by Linnsoft:
