Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: MA(1) process

From   Robert A Yaffee <>
Subject   Re: st: MA(1) process
Date   Fri, 09 Jul 2010 11:50:15 -0400

   If you assume that your mean-centered series is a function of its past observations, it has an AR structure.  AR(1) means that it is a function of
only the first lag of itself.   With an MA(1) structure, the observation is a function of the current and first lag of the disturbance (shock, innovation or error).
   You can convert one to the other.  Actually, an AR(1) is functionally equivalent to a MA("infinite") and an AR("infinite") is functionally equivalent to a MA(1), assuming covariance stationarity.
   You may first need to test before you make these assumptions. 

    - Robert

Robert A. Yaffee, Ph.D.
Research Professor
Silver School of Social Work
New York University



----- Original Message -----
From: Ari Dothan <>
Date: Friday, July 9, 2010 11:32 am
Subject: st: MA(1) process

> Hi Statalisters,
>  I am using a gmm procedure for dynamic panels which makes it possible
>  to fit a model with an MA(1) error structure (moving average (1st
>  order). Most other procedures, such as fixed effects, use the AR(1)
>  error structure.
>  Could anyone explain me in layman’s terms what is the difference
>  between the MA(1) and the AR(1) error structures? Why, and when,
>  should one be used rather than the other?
>  Thanks
>  --
>  Ari Dothan
>  *
>  *   For searches and help try:
>  *
>  *
>  *

*   For searches and help try:

© Copyright 1996–2016 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index