Experiments and computational models of trading in decentralized markets

We present the results of experimental and computational analysis of decentralized markets with asymmetric information. We examine how market efficiency and agents profit distribution depend on the structure of the underlying trading network and on the number and location of insiders. We find that market efficiency and fairness are both affected by the network macro-structure and that the network that better supports price efficiency is not necessarily the one that leads to greater fairness in profits. By calibrating a behavioural Agent Based Model to the experimental data we characterize the traders’ learning rule and shed further lights on the dynamics of information diffusion.