Pseudo-IV GLS for time series


No cake for spunky
Some call this two stage Aitkin. Its recommended to run dynamic regression when you have serial correlation and a lagged Y as a predictor. I have not been able to find out how to do it in SAS.


New Member
PROC MODEL may allow you to use this. With it, you can specify IVs, ARMA & vector ARMA terms (%AR() and %MA() macros), and fit the system using 2SLS. I'm not sure if it's the exact same thing as Pseudo-IV GLS, though.

For example:

   proc model data=test;
      exogenous x1 x2;
      parms a1 a2 b2 2.5 c2 55 d1;

      y1 = a1 * y2 + b2 * x1 * x1 + d1;
      y2 = a2 * y1 + b2 * x2 * x2 + c2 / x2 + d1;
      %AR(y1, 1);

      fit y1 y2 / 2sls;
      instruments b2 c2 _exog_;


No cake for spunky
PROC model is something I have rarely used, but it does many things I need to learn. I have not found any references specifically to Pseudo-IV GLS and SAS combined so far. I am not sure 2sls is this or not.