Abstract :
Trade requires search, negotiation, and exchange, which are activities that absorb resources. This paper investigates how different trade networks attend to these activities. An artificial market is constructed in which autonomous agents endowed with a stock of goods seek out partners, negotiate a price, and then trade with the agent offering the best deal. Different trade networks are imposed on the system by restricting the set of individuals with whom an agent can communicate. We then compare the path to the eventual equilibrium as well as the equilibrium characteristics of each trade network to see how each system dealt with the tasks of search, negotiation, and exchange. Initially, all agents are free to trade with any individual in the global market. In such a world, global resources are optimally allocated with few trades, but only after a tremendous amount of search and negotiation. If trade is restricted within disjoint local boundaries, search is simple but global efficiency elusive. However, a hybrid model in which most agents trade locally but a few agents trade globally results in an economy that quickly reaches a Pareto optimal equilibrium with significantly lower search and negotiation costs. Such `small-worldʹ networks occur in nature and may help explain the ease with which most of us acquire goods from around the world. We also show that there are private incentives for such a system to arise.
