Abstract:

The parameter estimator of generalized random coefficient autoregressive model was studied and a class of estimators of unknown parameter was presented. We proved the consistency and the asymptotic normality of the estimators among the class of estimators and obtained the smallest asymptotic variance estimator. Furthermore, some simulation studies were conducted to investigate the finite performances of the proposed estimators.

Key words: generalized random coefficient autoregressive model, least squares estimator, asymptotic normality, weighted conditional least squares estimator, smallest asymptotic variance estimator