THE FINITE-SAMPLE PROPERTIES OF SIMULTANEOUS-EQUATIONS ESTIMATES AND ESTIMATORS BAYESIAN AND NON-BAYESIAN APPROACHES

After discussing the need for good finite sample estimation procedures for simultaneous equations models and showing the inadequacies of asy mptotically justified estimators, it is shown how the Bayesian method of moments (BMOM) provides an exact, finite sample analysis of unrestr icted reduced form systems. Then optimal, finite sample estimates of s tructural coefficients are derived using three standard loss functions and they are compared to traditional Bayesian optimal estimates. Mont e Carlo experimental evidence from four studies on the relative perfor mance of Bayesian and non-Bayesian estimators is reviewed with the fin ding that the performance of Bayesian estimators is better. (C) 1998 E lsevier Science S.A.