A network approach for opinion dynamics and price formation

"If men define situations as real, they are real in their consequences". W.I. Thomas and D.S. Thomas In this work, we investigate the intertwined role of network interaction, opinion dynamics and price formation in a financial system. We propose a dynamical multi - agent framework where the interaction network and its topology, opinions and prices depend on one another, co - evolving in time. At first, we introduce some useful concepts in network theory and opinion dynamics. A method for classifying agents according to their topological role in the network is proposed. Second, we build on the existing literature on hetereogenous beliefs and evolutionary systems and provide a model with a specific update rule that leads to an evolving topology. The model is apt at describing social and behavioural phenomena that have recently received particular attention in the financial literature, such as hetereogeneous beliefs on market scenarios and the effects of the topology of interactions. We illustrate such dynamics via simulations, discussing the stylized facts that the model might be able to capture and we will discuss the use of social network data in order to calibrate the model. Third, we propose a model for formation of relative prices in a closed economy when agents have limited attention about a certain asset/sector.