Measures of Kurtosis and Skewness of INGARCH Model

Recently there has been a growing interest in time series of counts/integer-valued time series. The time series under the hypothesis of homogeneous variance becomes unrealistic in many situations because the variance tend to change with level. Important models such as ACP (autoregressive conditional Poisson) models and integer valued GARCH models have been proposed in the literature. Ghahramani and Thavaneswaran [1] studied the moment properties of ACP models using martingale transformation. However the forecasting for count process has not been studied in the literature. Using a martingale transformation, Thavaneswaran et al. [2] studied the volatility forecasts for GARCH models. In this paper, first we derive closed form expressions for skewness and kurtosis for count processes via martingale transformation then we study the joint forecasts for integer-valued count models with errors following Poisson.