Order book on financial networks

Abstract

This chapter presents results from a class of agent-based models describing a realistic financial order book, aiming to discuss some aspects related to market stability.Major empirical regularities of data are correctly replicated and simulations are used to infer possible policy implications. After a detailed explanation of model features, an application to networks is presented to advance intuitions about the role of social interaction on induced imitation and herding phenomena. Thus, the initial market structure is augmented by a dynamic multiplex with two layers devoted to information and trading. The first one, representing social interactions, is designed according to different topologies in order to showhowinvestors decide their behavior by following perceived informative flows. The second one, where the central hub is the market maker (i.e., the owner of the venue holding the order book) is devoted to the execution of all transactions. Some policy implications oriented to foster market stability are finally provided.