Exact Maximum Likelihood Estimation of Regression Equations with a General Stationary Autoregressive Disturbance

This paper develops an exact maximum likelihood technique for estimating regression equation with general p'th order autoregressive disturbances. Recent expression of the analytic inverse of the covariance matrix of a stationary AR(p) process provide the basis for an iterative, modified Gauss-Newton technique using exact first and approximate second derivatives. Empirical estimates are presented for regression models with and without a lagged dependent variable.