• NEW RIDGE ESTIMATORS OF SUR MODEL WHEN THE ERRORS ARE SERIALLY CORRELATED

Abstract

This paper considers the seemingly unrelated regressions (SUR) model when the errors are first-order serially correlated as well as the explanatory variables are highly correlated. We proposed new ridge estimators for this model under these conditions. Moreover, the performance of the classical (Zellner’s and Parks’) estimators and the proposed (ridge) estimators has been examined by a Monte Carlo simulation study. The results indicated that the proposed estimators are efficient and reliable than the classical estimators.