How to select period of Moving Average scientifically

mangup

Well-Known Member
#1
I know that the moving average period will vary if we change the time frames for a particular scrip. It means there will be a combo of time frame-MA. Is there any scientific way to determine period of moving average for a particular scrip & time frame. Also scrip's volatility & ATR will have some effect while finalising the period of MA. How to take them in account?

How this period will get changed to new value scientifically, if we change the time frame?

This will have a deep impact on the strategies working on moving average crossovers.

Please suggest.
 

Reggie

Well-Known Member
#2
In my view there is no need to invent the wheel. In general 20 Period MA is considered a good indicator for ST trading, 50, 100 or 200 for longer timeframe.

Regards to volatile scripts, best to avoid them as they distort the MA.

Check out the videos on the net for more. They are quite informative.


I know that the moving average period will vary if we change the time frames for a particular scrip. It means there will be a combo of time frame-MA. Is there any scientific way to determine period of moving average for a particular scrip & time frame. Also scrip's volatility & ATR will have some effect while finalising the period of MA. How to take them in account?

How this period will get changed to new value scientifically, if we change the time frame?

This will have a deep impact on the strategies working on moving average crossovers.

Please suggest.
 

DanPickUp

Well-Known Member
#4
http://en.wikipedia.org/wiki/Moving_average

http://www.investopedia.com/terms/m/movingaverage.asp

http://www.stockwatch.com.au/articles/technical-analysis/moving-averages.aspx

Above just some Google results.

Now depending on the software you use, you will be able to make system tests with different kind of indicators. You can do that with Metastock on any time frame if you have the data. After putting in your choices and numbers, let the simulation run. You then will get a bunch of results. One of those results will show you which MA in the past was the best for the time frame you tested.

Understand: PAST, so no guarantee that the same MA's will work in the future. Never mind as you can do such test every day.

Good trading

DanPickUp
 

mangup

Well-Known Member
#5
Thanks DanPickUp,

In all of the above articles no where it is mentioned how to derive a period of MA scientifically for a particular time frame chosen.
 

DanPickUp

Well-Known Member
#6
Thanks DanPickUp,

In all of the above articles no where it is mentioned how to derive a period of MA scientifically for a particular time frame chosen.
Dear Mangup

Goggling doe's not give a specific answer to that specific question. So, in that point I have to agree with you.:) and I posted some link which I found useful to expand knowledge about MA's for those which are not on your level.

But you do not need to search on that as any system tester in good charting software will do the job. They will give you exactly results which MA crossover system was most successful in the time you choose d to check. And that result in general will also do it's job for the next week or next month and then you make a new system test.

Here some more information about MA's, but not a specific answer.

http://www.incrediblecharts.com/indicators/moving_average_selection.php

If this all is still not enough for you, then I guess you loose time on a subject/question which anyway not will beat any system tester or improve your trading ledger in any way.

Your choice and good trading

DanPickUp
 

Similar threads