Michael,
You should test your series with a dfgls test to see whether it is in need
of differencing. If so, you'll have to difference the serries before undertaking
the modeling.
Regards,
Robert Yaffee
Robert A. Yaffee, Ph.D.
Research Professor
Shirley M. Ehrenkranz
School of Social Work
New York University
home address:
Apt 19-W
2100 Linwood Ave.
Fort Lee, NJ
07024-3171
Phone: 201-242-3824
Fax: 201-242-3825
[email protected]
----- Original Message -----
From: Michael Crain <[email protected]>
Date: Wednesday, May 30, 2007 7:10 pm
Subject: st: ARIMA postestimation/ Dynamic forecasting with AR(12) model
To: [email protected]
> I am having trouble with the syntax with the postestimation of an
> AR(12)
> model (using monthly data).
>
> My various data sets (balanced and unbalanced) look like this:
>
> Sales_amt Yr_mo
> $xxxxx 2002m1
> $xxxxx 2002m2
> $xxxxx 2002m3
> .... ...
> $xxxx 2005m10
>
> My AR(12) model runs through 2004m8 (not 2005m10). I want to predict
> what
> Sales_amt would have been starting 2004m9 using Dynamic forecasting.
>
> Accordingly, I have done the following:
>
>
> -- arima sales if tin( , 2004m8), ar(12) --
>
> (output ommitted)
>
> I am having trouble with the syntax for the Dynamic forecasting. I
> want the
> dynamic forecasting to begin with the forecast for 2004m9. I tried
> the
> following syntax but it did not work as the results were nonsensical:
>
> -- predict saleshatdy, dynamic(m(2004m8)) y --
>
> The nonsynsical results were:
>
> sales yr_mo saleshatdy
> 8335.21 2002m1 812.2243
> 11773.63 2002m2 812.2243
> 13615.85 2002m3 812.2243
> 6669.48 2002m4 812.2243
> 7144.34 2002m5 812.2243
> 6052.71 2002m6 812.2243
> 5556.54 2002m7 812.2243
> 6330.65 2002m8 812.2243
> 3391.77 2002m9 812.2243
> 4546.76 2002m10 812.2243
> 5290.41 2002m11 812.2243
> 5745.89 2002m12 812.2243
> 6149.88 2003m1 1063.088
> 11435.43 2003m2 361.4133
> 10813.48 2003m3 401.9798
> 8201.24 2003m4 966.1082
> 6670.49 2003m5 993.3638
> 6884.43 2003m6 917.3016
> 3662.71 2003m7 881.8058
> 9301.96 2003m8 940.1202
> 4651.49 2003m9 684.044
> 8462.30 2003m10 798.1374
> 8384.00 2003m11 864.058
> 8367.59 2003m12 895.7506
> 9300.54 2004m1 926.8094
> 11379.53 2004m2 349.3701
> 11794.66 2004m3 332.2561
> 10372.96 2004m4 1058.651
> 6026.79 2004m5 966.7421
> 5919.31 2004m6 978.1514
> 7074.07 2004m7 709.3981
> 4652.76 2004m8 1103.654
> 1253.90 2004m9 803.2082
> 6906.79 2004m10 1069.426
> 4061.47 2004m11 1065.623
> 7766.54 2004m12 1064.355
> 8382.98 2005m1 1103.02
> 12928.70 2005m2 346.8347
> 10155.60 2005m3 362.681
> 8504.59 2005m4 320.2129
> 8116.69 2005m5 914.1323
> 3983.16 2005m6 905.8923
> 4829.42 2005m7 991.4623
> 3219.39 2005m8 996.9477
> 4668.28 2005m9 806.5094
> 5054.59 2005m10 975.2522
>
> (The 812.2243 for the first 12 observations is the constant in the
> model.
> The other saleshat amounts are nonsensical.)
>
> Again the above example is for an AR(12) model. In addition to
> dynamic
> forecasting for an AR(12) model, I am also using dynamic forecasting
> for the
> following models:
> AR(1)
> AR(1) with differencing
> AR(12) with differencing
>
> I would appreciate help on the dynamic forecasting syntax as discussed
>
> above.
>
> Thanks in advance.
>
> _________________________________________________________________
> PC Magazine�s 2007 editors� choice for best Web mail�award-winning
> Windows
> Live Hotmail.
> http://imagine-windowslive.com/hotmail/?locale=en-us&ocid=TXT_TAGHM_migration_HM_mini_pcmag_0507
>
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
