I have two SUR models for which I want to assess the fit. I understand there is controversy as to which measures should be used for determining which of several models fits the data better. In addition to this, It's unclear to me whether typical measures to assess the fit of regressions are still valid in the SUR context.

The equations within my two SUR models are identical. But I want to compare the impact of running the same exact SUR model with and without the regression. Without the regression implies I would drop the constant for all equations in the SUR.

I thought of using BIC and AIC statistics by running estat ic after each SUR model. My results are below, but it's concerning that the "ll(null)" is blank for both cases, which seems to indicate that that estat commands isn't meant for SUR models as no 'null' model is being defined.

I also tried using the fitstat package (ssc fitstat) but after running my SUR model but I get an "-fitstat- does not work with the last model estimated." error.

Any suggestions on how to compare two SUR models would be appreciated.

--

SUR without constant

Code:
estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |          N   ll(null)  ll(model)      df        AIC        BIC
-------------+---------------------------------------------------------------
           . |        147.   262.7028      18  -489.4056  -434.2789
-----------------------------------------------------------------------------
SUR with constant

Code:
.
estat ic

Akaike's information criterion and Bayesian information criterion

-----------------------------------------------------------------------------
       Model |          N   ll(null)  ll(model)      df        AIC        BIC
-------------+---------------------------------------------------------------
           . |        147.   281.8853      20  -523.7706  -462.5187
-----------------------------------------------------------------------------