The VARMAX Procedure

Bayesian inference on a cointegrated system begins by using the priors of , which are obtained from the VECM(p) form. Bayesian vector error correction models can improve forecast accuracy for cointegrated processes.

To use a Bayesian vector error correction model, you specify both the PRIOR= option in the MODEL statement and the COINTEG statement. The following statements fit a BVECM(2) form to the simulated data:

/*--- Bayesian Vector Error Correction Model ---*/

proc varmax data=simul2;
   model y1 y2 / p=2 noint
                 prior=( lambda=0.5 theta=0.2 )
                 print=(estimates);
   cointeg rank=1 normalize=y1;
run;

The VARMAX procedure output in Figure 42.18. shows the model type fitted to the data, the estimates of the adjustment coefficient (), the parameter estimates in terms of lag 1 coefficients (), and lag 1 first-differenced coefficients ().