Kaufman's Adaptable Moving Average
Kaufman's Adaptable Moving Average
Has anybody spent any time researching this indicator? I have'nt got round to it yet, but it is the next thing on my list. My initial concern is that although in theory it sounds very appealing - ie keeps you out of sideways markets, I'm concerned about over-optimisation.
-
- Roundtable Fellow
- Posts: 68
- Joined: Fri May 16, 2003 8:28 am
- Location: Atlanta, GA
The KAMA is described in detail in Perry's book "Smarter Trading". I played around with it a bit in a simple MA crossover system (using it as the short-term MA), but didn't see the results improve significantly over using EMA. It's been over a year since I read that book, but I think Mr. Kaufman suggested using KAMA as a single MA rather than in a crossover system, and buying whenever it turned positive over a 9-14 day period. So perhaps using it in this way would give better results.
Thinking of that, I'd like to re-read the book now. It seemed to contain some interesting ideas on backtesting.
Jason
Thinking of that, I'd like to re-read the book now. It seemed to contain some interesting ideas on backtesting.
Jason
-
- Roundtable Fellow
- Posts: 84
- Joined: Thu May 29, 2003 12:11 pm
- Location: Eugene, OR
- Contact:
-
- Roundtable Fellow
- Posts: 68
- Joined: Fri May 16, 2003 8:28 am
- Location: Atlanta, GA
Chris,
Here are a couple of links that may help:
http://eis.pl/kr/ELZ/t-s-Adaptive_Movin ... ystem.html
http://www.traders.com/Documentation/FE ... s9803.html
Here are a couple of links that may help:
http://eis.pl/kr/ELZ/t-s-Adaptive_Movin ... ystem.html
http://www.traders.com/Documentation/FE ... s9803.html
"Smarter trading" is a very good book and has lots of thought-provoking ideas.
The concept behind Kaufman's AMA is interesting - using market efficiency to adjust a MA speed. I wrote a few systems using it as Kaufman suggests and some other ways. Using it as Kaufman says didn't test well. I came to the conclusion trading the turns in the AMA doesn't add value beyond other traditional simple or exponential MA systems. That approach suffers from as much (or more) lag and whipsaw trades as other traditional MA systems.
One system has shown some promise. It uses range exspansion with a close over/below the AMA to setup a trade, with the next bar taking out the high/low for actual entry. With a couple different exits - including profit taking as Kaufman suggests in his book - the test results were good. The number of parameters and rules have me somewhat concerned, even though they are "robust". But I believe the true test of a system is to trade it, so I am. I'm currently trading it on 30 minute S&P bars in my "test account". I've been trading it for only 6 weeks and am currently ahead - but not much.
Thanks for a good, informative forum c.f..
The concept behind Kaufman's AMA is interesting - using market efficiency to adjust a MA speed. I wrote a few systems using it as Kaufman suggests and some other ways. Using it as Kaufman says didn't test well. I came to the conclusion trading the turns in the AMA doesn't add value beyond other traditional simple or exponential MA systems. That approach suffers from as much (or more) lag and whipsaw trades as other traditional MA systems.
One system has shown some promise. It uses range exspansion with a close over/below the AMA to setup a trade, with the next bar taking out the high/low for actual entry. With a couple different exits - including profit taking as Kaufman suggests in his book - the test results were good. The number of parameters and rules have me somewhat concerned, even though they are "robust". But I believe the true test of a system is to trade it, so I am. I'm currently trading it on 30 minute S&P bars in my "test account". I've been trading it for only 6 weeks and am currently ahead - but not much.
Thanks for a good, informative forum c.f..