Lots of chatter yesterday about gold breaking through its 200-day moving average. Ron Griess says to consider the 300-day moving average instead. Has anyone done research on the benefits of using non-round-number moving averages, such as numbers slightly lower than 200 or 300?

Phil McDonnell writes:

I have tested them all from about 2 days to maybe 1000 days. Generally speaking there is little difference between two 'nearby' moving averages. If you think about it the 200 day has 190 terms in common with the 190 day average. Generally speaking they almost all work as long as the average is longer than about 100 days or so.

All such methods are weak at best. For example the 200 day yields about .04% per day if above the average. But adding a 50 day/200 day crossover reduces the yield to only .03%. In other words the so called golden cross does not really work.





Speak your mind

1 Comment so far

  1. Adam G on December 26, 2011 12:18 am

    I have done extensive testing of moving averages of many different periods and variations of crossover. I was unable to find any edge in any naive application of moving averages. I am firmly convinced that anyone telling you to watching any moving average is completely full of… well you know…

    Short answer is no, there’s no advantage to any specific moving average. They are all equally meaningless (based on my work).


Resources & Links