AIC and BIC information criterion for frontier models

[duynguyen]

Hi everyone,
I used xtfrontier command for panel data in STATA 13, and then calculate AIC and BIC information criterion for 3 frontier models with results in the following. Could you please explain for me which model is the best and why?

. estimates stats A B C, n(114)

Akaike's information criterion and Bayesian information criterion

-------------------------------------------------------------------------------------------
Model | Obs ll(null) ll(model) df AIC BIC
---------+--------------------------------------------------------------------------------
A | 114 . 81.04466 8 -146.0893 -124.1997
B | 114 . 80.87785 7 -147.7557 -128.6023
C | 114 . 78.70014 7 -143.4003 -124.2469
-------------------------------------------------------------------------------------------
Note: N=114 used in calculating BIC

Thank you!
Regards,
Duy Nguyen
UiT, Norway