Numerical computation of exact moments of the least squares estimator in a first-order stationary autoregressive model

Abstract

This paper examines the least squares estimator of the autoregressive coefficient in a first-order sta¬tionary autoregressive model. Exact lower-order moments are computed by numerical integrations. From the study of moment values, it is found that the exact distribution of the least squares estimator may be well approximated by a beta distribution.

Keywords:

• Trigonometric function
• Skewness
• Kurtosis
• Beta distribution