[metalhead]

Does anyone use this? I did a search and didn't find anything.

Sometimes also called AMA, or KAMA (Kaufman Adaptive Moving Average).

This moving average goes flat during choppy markets.

Screenshot with some other MAs for comparison.

[john.potter]

Metalhead
Thanks a heap, when I initially got this charting program I went through the 10 different types of MA on offer and it was six of one and half a dozen of the other none of them showed any particular advantage.(that I could see)

However hidden in amongst the indicators was another MA the "Perry Kaufman's adaptive Moving Average" which I have just checked out and you are dead right! it is pretty amazing and I will keep it on the charts.

Thank you very much, without your post I would not have bothered to look for it let alone use it.

[metalhead]

Glad you find it useful. I'm trying to come up with good applications/settings for this MA.

Also be aware that you can change more than just the length in the settings. There's also Fast Smoothing Constant and Slow Smoothing Constant.

[metalhead]

Here's a document that explains how it was derived and how it works.

Even better, here's a link to the book wherein it was first described.

http://tinyurl.com/6hburf

[john.potter]

Hi Forsearch

Thanks for that, I had picked up "Trading systems and Methods" from the same site but had not gotten around to reading it yet, I thought it would have the MA's in that one.

P.S. I havent got the holy grail yet but I do have one I call the 4 horsemen, It is signifying the US Indexes are getting ready to go bullish for between 4 to 6 weeks, and the XAO would follow them

[pathfinder62]

Re:Here's a document ...
MetalHead: good stuff and thanx 4 sharing.
pathfinder62-my age BTW.

[metalhead]

The problem when the AMA "going flat" is that it's not actually "going flat." It's still moving up or down a little bit and therefore still has a slope.

I've been trying to come up with a way to modify the AMA so that "almost flat" counts as "flat." Basically you have to apply a filter so that if the change between the previous AMA reading and the current AMA reading is less than a certain amount, it "doesn't count."

I can't find a good way to come up with this value, so here's a pic with an arbitrary value chosen.

Green = go long.
Yellow = do nothing (chop)
Red = go short.

Notice there are some good signals and some bad signals. Some of the good signals come late.