Comparison of the methods that are used to estimate seemingly unrelated regression equations

Estimation of seemingly unrelated regression models are one of the main subjects of econometrics. It is known that seemingly unrelated regression model estimates are more efficient than the least square estimations, under the assumption that the independent variables of different equations which constitute that model are not related with each other, for the related error terms that are obtained for the same time point. In this study, algorithms for Matlab are constructed in order to obtain coefficient estimates of seemingly unrelated regression equations which are commonly used in application and more efficient in the manner of calculation, and using empirical data the efficiency of the coefficients in the manner of calculation is examined. Keywords: Seemingly Unrelated Regression Model, QR Decomposition, RQ (or QL) Decomposition, Generalized QR Decomposition, Matlab.