{"id":389,"date":"2017-05-12T11:30:49","date_gmt":"2017-05-12T11:30:49","guid":{"rendered":"http:\/\/jsr.isrt.ac.bd\/?post_type=article&p=389"},"modified":"2017-05-12T11:35:38","modified_gmt":"2017-05-12T11:35:38","slug":"moment-restrictions-optimum-gmm-estimators-spatial-simultaneous-systems","status":"publish","type":"article","link":"http:\/\/jsr.isrt.ac.bd\/article\/moment-restrictions-optimum-gmm-estimators-spatial-simultaneous-systems\/","title":{"rendered":"Moment restrictions for optimum GMM estimators under spatial simultaneous systems"},"content":{"rendered":"
It is widely known that for large samples maximum-likelihood based estimators are not
\neasy to implement for spatial models unless weights matrix satisfy certain basic conditions.
\nAs an alternative recently in a series of paper Kelejian and Prucha (1999, 2004) proposed
\na computationally feasible three step procedure for spatial models involving both lagged
\ndependent variable and spatially correlated disturbance term. The idea of this paper is to use
\ntheir set up for spatial simultaneous system, and construct a GMM type estimator based on
\nspatial first difference. The over-identification of the moment equation comes to the picture
\nby considering first two moments of a possibly heteroskedastic disturbance. Following
\nChamberlains (1987) idea, a popular issue of optimum GMM based on conditional spatial
\nmoment restrictions and asymptotic efficiency has been discussed.<\/p>\n