Bilateral Trading on a Network: A Simulation Study

Ann M. Bell
Department of Economics, Vanderbilt University
BellAM@ctrvax.vanderbilt.edu

Abstract

This paper examines the dynamic behavior of a process of bilateral trading between spatially diverse agents. In contrast to the general equilibrium trading mechanism and to many previous models of bilateral trading, trades may only occur between agents who are located adjacent to one another. A network of bilateral trades provides one process by which an economy-wide equilibrium can be achieved in a truly decentralized fashion.

Utility maximizing agents are located in space on a network or graph and trade bilaterally with their nearest neighbors. Agents may differ in their preferences, in their endowments of two infinitely durable goods on in their expectations of future trading opportunities. Trade occur asynchronously-only one trading pair is active at any point in time. Trading continues until all gains from trade between any two adjacent agents have been exploited. Computer simulations are used to explore the dynamic behavior of trading on a network. Econometric techniques for spatial data are used to analyse the data from the simulations.

Previous theoretical analysis of non-tatônnement trading mechanisms has focused on the two extremes of the information spectrum: either agents are fully informed about the equilibrium prices that will eventually prevail, or they are completely myopic and do not anticipate any future trading opportunities. This paper contrasts the dynamic behavior of a system of network trading in the extreme case where agents are completely myopic and in the intermediate case where agents use boundedly rational prediction rules to anticipate future prices. Convergence to a Pareto optimal allocation can be guaranteed when agents are completely myopic. However, because each bilateral trade changes agents' holdings and relative wealth, the final allocation need not be in the core of the initial allocation. In addition, the process of convergence generates interesting intertemporal and cross-sectional dynamics such as the persistence of spatial correlations, or neighborhood effects, in the prices of goods over time. The speed of convergence is also relevant-the longer it takes for the lattice to converge to a common trading price, the longer inefficiencies or social welfare costs associated with decentralized trades accumulate. A network of trades between boundedly rational agents typically has faster convergence, but also give rise spatially based price bubbles because agents use local information to predict future prices.