Information versus imitation in a real-time agent-based model of financial markets

This paper presents an agent based model of a financial market with a real-time engine, whose operation replicates the official time schedule of Borsa Italiana S.p.A. Simulated time series are compared with empirical data at different time scales (ticks, 1 s, 1 min, 5 min) in order to check the compliance of the model with some stylized facts. The modeled market structure is a dynamic multiplex with two layers: the first one is a star network, whose hub is the market maker (i.e., the owner of the venue holding the order book), where transactions are executed; the second one is designed according to different topologies, representing social interactions, where investors decide their behavior according to informative flows. The effect of imitation on market stability is discussed and some policy implications are provided.