Mirko Armillotta and Konstantinos Fokianos’ contribution to the Discussion of ‘New tools for network time series with an application to COVID-19 hospitalisations’ by Nason et al.

The authors introduce several results related to network time series.This line of research was initialized by Knight et al. (2016) and studied in detail by Zhu et al. (2017); see also Knight et al. (2020).Such models provide satisfactory flexibility to model multivariate time series data provided that the neighborhood pattern of the network is known.Some ideas and questions for further extension are discussed below.1. We would like to highlight the close connection of network time series models to the space-time ARMA (STARMA) models as introduced by Pfeifer and Deutsch (1980).Indeed, suitable parametriza-