Statistical inference in nonlinear sure model

Abstract

The seemingly unrelated Regression equations (SURE) model is a generalization of a regression model that consists of several regression equations, each having its an dependent variable and potentially different sets of exogenous explanatory variables. The SURE model approach for estimating system of linear regression equations in which the errors are contemporaneously corrected across equations across equations but not auto correlated. The SURE model is containing to receive wide spread attention in terms of both theoretical developments and empirical applications. In the present study, a Nonlinear SURE model has been specified and a feasible generalized least squares (GLS) estimator for the parametric vector has been developed along with its asymptotic variance covariance matrix. Further a consistent estimator of the dispersion matrix has been derived and it will be the nonlinear maximum likelihood estimator.