MAXIMUM-LIKELIHOOD-ESTIMATION FOR A MULTIVARIATE AUTOREGRESSIVE MODEL

This paper provides analytical expression for the exact log likelihood function and its first derivatives for a multivariate autoregressive model. Based on these results, two algorithms for constructing the max imum likelihood estimate, using the Fisher's scoring technique, are pr oposed. The estimated model is guaranteed to be stable. Simulation exa mples show that this algorithm has good convergence properties and the resulting maximum likelihood estimator could perform better than earl ier methods, in cases where the record length is short and the autoreg ressive polynomial has roots near the unit circle.