Understanding the structure of interactions between corporate firms is critical to identify risk concentration and the possible pathways of propagation of financial distress. In this paper we consider the in- teraction due to payments and, by investigating a large proprietary dataset of Italian firms, we characterize the topological properties of the payment network. We then focus on the relation between the net- work of payments and the risk of firms. We show the existence of an homophily of risk, i.e. the tendency of firms with similar risk pro- file to be statistically more connected among themselves. This effect is observed both when considering pairs of firms and when consider- ing communities or hierarchies identified in the network. By applying machine learning techniques, we leverage this knowledge to show that network properties of a node can be used to predict the missing rating of a firm. Our results suggest that risk assessment should take quan- titatively into account also the network of interactions among firms.
↧