“…Nonnegative integer‐valued count process models are widely used in domains such as marketing (Böckenholt, 1998), economics (Blundell, Griffith, & Van Reenen, 1999; Brännäs & Hellström, 2001; Harris & McCabe, 2019), finance (Bien, Nolte, & Pohlmeier, 2011; Heinen & Rengifo, 2007; Kirchner, 2017), and insurance (Gouriéroux & Jasiak, 2004). The benchmark model, introduced by McKenzie (1985) and Al‐Osh and Alzaid (1987) in the first‐order case (called INAR(1)) postulates that: where the thinning operator is defined as follows: conditionally on X t −1 , variable α∘ X t −1 has the distribution , that is the binomial distribution with probability parameter α and size parameter X t −1 .…”