Does anyone know of any data or have an opinion on the "best" MA ratio parameters? In other words, if the short MA is X should the Medium MA be 2 times or 5 times X (for example)?
Thanks for any input.
Moving Average Ratio Optimization Research
Some people here are quite fond of Kaufman's book (link). It might be a starting point from which to do your own work.Dallas wrote:I was just looking for a starting point from which to do my own work
On page 868, Kaufman shows a matrix of simulations that he performed on a two moving average crossover system (applied to commodity futures). The smallest value of your ratio (LongMAdays / ShortMAdays) in his matrix, is 1.05. The LongMAdays is 1.05X bigger than the ShortMAdays. The largest value of your ratio in his matrix, is 33.3. The LongMAdays is 33.3X bigger than the ShortMAdays.
However, Kaufman ran those simulations at least 10 years ago. (The same Figure appears on p.515 of the 1998 edition of his book.) Backtest software has improved significantly in the past 10 years, and so has personal computer hardware. So if Kaufman was able to perform X amount of backtest simulation, ten years ago, You ought to be able to perform 100X amount of backtest simulation today. In my opinion.
Thank you Sluggo
I appreciate your constructive guidance. I did take a short peek at Kaufman's "monster" book one time. I will seriously consider buying it.
$80 online is much better than $125 at the store!
$80 online is much better than $125 at the store!