Moments of AR(1)-Model Estimators

ABSTRACT

We describe a simple technique for computing the first few moments of a large class of sample statistics related to the first-order autoregressive model with normally distributed error terms. Moments of the basic estimators of ρ and σ are then derived for illustration, also indicating how the leading term of skewness can be eliminated.

By concentrating on the computational aspects of the procedure, this article should complete Tuan (1992).

Key Words:

• First-order autoregressive model
• Sampling distribution moments
• Serial correlation estimator

Mathematics Subject Classification:

• 62E20