Asymptotic distribution of the estimated parameters of an ARMA(p, q) process with mixing innovations

作者：Bhattacharya Sankha^*; Sen Roy Sugata

来源：Brazilian Journal of Probability and Statistics (Brazilian Stat. Assoc.), 2015, 29(3): 640-655.

DOI：10.1214/14-BJPS237

• ARMA process
• explosive roots
• phi-mixing errors
• asymptotic distribution

In this paper, we consider an ARMA(p, q) model with stationary, phi-mixing error variables having uniformly bounded fourth-order moments. Both the autoregressive and moving average components of the model involve stable and explosive roots. Estimating the autoregressive parameters using the instrumental variable technique and the moving average parameters using a derived autoregressive process, we derive the asymptotic distribution of the estimators.

• 出版日期2015-8