Abstract
Recently, the GARMA(1, 2; δ, 1) time series model has been introduced in the literature. In this paper, the variance and autocovariance of the GARMA(1, 2; δ, 1) model are derived. Some numerical results are also provided. It appears from the simulation study that the HRA and WE estimation procedures are relatively good for GARMA(1, 2; δ,1) model. The GARMA(1, 2; δ, 1) model was also applied to model a real data set, namely Dow Jones Utility Index. We believe that this model would also be useful for modeling many other time series data.
Mathematics Subject Classification: