Networks, beliefs, and asset prices

Abstract

We set out a novel social communication model of asset prices. An investor's type – which depends on their network and investment performance – determines their price beliefs. We show how properties of the network such as network centrality and diameter influence the price dynamics, convergence speed, and limiting belief types. For the polar cases of no attention to performance and exclusive attention to performance, we obtain analytically tractable results relating price and belief types to properties of the network, while for intermediate attention to performance we rely on numerical results. As applications, our model can explain price bubbles and price oscillations by network-performance effects, and we also study how price and type dynamics depend on connectedness on a small-world network. Our results shed light on when performance-based updating of beliefs on social networks is stabilising – or destabilising – for asset prices. A key finding is that the impact of network structure on asset prices and beliefs depends on how much attention investors pay to performance.