Date:         Mon, 13 Dec 2004 09:21:43 -0500
Reply-To:     Fred <ieaggie2002@gmail.com>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Fred <ieaggie2002@GMAIL.COM>
Subject:      Whats the relation of classical Moving Average to MA in ARIMA proc
Content-Type: text/plain; charset=US-ASCII

Hi,all

I want to do classical moving average forecast for a given data set X,
where X1, X2, ..., Xt are the timeseries value at period 1 to t.
The classical moving average forecasting (e.g., 3-period average) is:
Xt+1 = (Xt + xt-1 + xt-2)/3;
Xt+2 = (Xt+1 + xt + xt-1)/3;
...

But the moving average model estimate and forecast in ARIMA procedure
is based on another model identification
Xt+1 = u + et - w1et-1 - ... where u is mean and et is the error.

So whats the mathematical relation bw these two different MA methods?
Can I do classical moving averaging forecasting using ARIMA proc?


Thanks a lot for your comments.

Fred