https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/ARIMA/m-p/358752#M2351 <P>The short answer is YES.</P> <P>For example, the following parameter estimates:</P> <TABLE class="table" summary="Procedure DOCUMENT: Maximum Likelihood Estimation" frame="box" rules="all" cellspacing="0" cellpadding="5"> <THEAD> <TR> <TH class="l b header" scope="col">Parameter</TH> <TH class="r b header" scope="col">Estimate</TH> <TH class="r b header" scope="col">Standard Error</TH> <TH class="r b header" scope="col">t&nbsp;Value</TH> <TH class="r b header" scope="col">Approx<BR />Pr &gt; |t|</TH> <TH class="r b header" scope="col">Lag</TH> </TR> </THEAD> <TBODY> <TR> <TH class="l rowheader" scope="row">MA1,1</TH> <TD class="r data">0.40194</TD> <TD class="r data">0.07988</TD> <TD class="r data">5.03</TD> <TD class="r data">&lt;.0001</TD> <TD class="r data">1</TD> </TR> <TR> <TH class="l rowheader" scope="row">MA2,1</TH> <TD class="r data">0.55686</TD> <TD class="r data">0.08403</TD> <TD class="r data">6.63</TD> <TD class="r data">&lt;.0001</TD> <TD class="r data">12</TD> </TR> </TBODY> </TABLE> <P>&nbsp;</P> <P>correspond to the following MA formulas:</P> <TABLE class="table" summary="Procedure DOCUMENT: Moving Average Factors" frame="box" rules="all" cellspacing="0" cellpadding="5"> <THEAD> <TR> <TH class="c header" colspan="2" scope="colgroup">Moving Average Factors</TH> </TR> </THEAD> <TBODY> <TR> <TH class="r t header" scope="col">Factor 1:</TH> <TD class="l data">1 - 0.40194 B**(1)</TD> </TR> <TR> <TH class="r t header" scope="col">Factor 2:</TH> <TD class="l data">1 - 0.55686 B**(12)</TD> </TR> </TBODY> </TABLE> <P>&nbsp;</P> <P>Hope this helps. Thanks.</P> Mon, 15 May 2017 14:08:52 GMT Puwang 2017-05-15T14:08:52Z https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/ARIMA/m-p/358550#M2349 <P>Hello, sorry if this happens to be the wrong forum section for this question.<BR />Anyhow, I remember that my teacher 2 years ago said that the ARIMA output in SAS shows the estimates of the MA components multiplied by (-1), is this (still) true?</P> <P><BR />Thanks in advance.</P> Sun, 14 May 2017 13:17:40 GMT https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/ARIMA/m-p/358550#M2349 Schteeke 2017-05-14T13:17:40Z https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/ARIMA/m-p/358752#M2351 <P>The short answer is YES.</P> <P>For example, the following parameter estimates:</P> <TABLE class="table" summary="Procedure DOCUMENT: Maximum Likelihood Estimation" frame="box" rules="all" cellspacing="0" cellpadding="5"> <THEAD> <TR> <TH class="l b header" scope="col">Parameter</TH> <TH class="r b header" scope="col">Estimate</TH> <TH class="r b header" scope="col">Standard Error</TH> <TH class="r b header" scope="col">t&nbsp;Value</TH> <TH class="r b header" scope="col">Approx<BR />Pr &gt; |t|</TH> <TH class="r b header" scope="col">Lag</TH> </TR> </THEAD> <TBODY> <TR> <TH class="l rowheader" scope="row">MA1,1</TH> <TD class="r data">0.40194</TD> <TD class="r data">0.07988</TD> <TD class="r data">5.03</TD> <TD class="r data">&lt;.0001</TD> <TD class="r data">1</TD> </TR> <TR> <TH class="l rowheader" scope="row">MA2,1</TH> <TD class="r data">0.55686</TD> <TD class="r data">0.08403</TD> <TD class="r data">6.63</TD> <TD class="r data">&lt;.0001</TD> <TD class="r data">12</TD> </TR> </TBODY> </TABLE> <P>&nbsp;</P> <P>correspond to the following MA formulas:</P> <TABLE class="table" summary="Procedure DOCUMENT: Moving Average Factors" frame="box" rules="all" cellspacing="0" cellpadding="5"> <THEAD> <TR> <TH class="c header" colspan="2" scope="colgroup">Moving Average Factors</TH> </TR> </THEAD> <TBODY> <TR> <TH class="r t header" scope="col">Factor 1:</TH> <TD class="l data">1 - 0.40194 B**(1)</TD> </TR> <TR> <TH class="r t header" scope="col">Factor 2:</TH> <TD class="l data">1 - 0.55686 B**(12)</TD> </TR> </TBODY> </TABLE> <P>&nbsp;</P> <P>Hope this helps. Thanks.</P> Mon, 15 May 2017 14:08:52 GMT https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/ARIMA/m-p/358752#M2351 Puwang 2017-05-15T14:08:52Z