Linear restrictions and two step multivariate least squares with applications

Some univariate normal linear regression model results of Pino (1984) are generalized to the multivariate normal linear regression model Y = X[beta] + E. The effect of transferring the information Y to T = A'Y may be analyzed through an associated regression problem which is amenable to solution by two step (or rather two stage) least squares. The results are applied to the case of eliminating observations. The inference aspects are not presented; however, they follow on similar lines as in the classical case, i.e., we analyze the model A'Y = A'X[beta] + A'E.