Decentralized Trading with Private Information

The paper studies asset pricing in informationally decentralized markets. These markets have two key frictions: trading is decentralized (bilateral), and some agents have private information. We analyze how uninformed agents acquire information over time from their bilateral trades. In particular, we show that uninformed agents can learn all the useful information in the long run and that the long-run allocation is Pareto efficient. We then explore how informed agents can exploit their informational advantage in the short run and provide sufficient conditions for the value of information to be positive. Finally, we provide a numerical analysis of the equilibrium trading dynamics and prices.