Can Chocolate be Money as a Medium of Exchange? Belief Learning vs. Reinforcement Learning

In this paper, a variant of the Kiyotaki and Wright model of emergence of money is investigated. In the model, all goods have a different durability rather than the storage cost assumed in the Kiyotaki and Wright model. Among three goods, two goods are infinitely durable but the third is nondurable. Under a certain condition, nondurable goods can be money as a medium of exchange. But the stationary equilibrium condition may be sensitive to the time evolution of the distribution of goods that each agent holds in its inventory. We test, with several learning models using different levels of information, whether or not the stationary equilibrium in this economy is attainable if the distribution of goods is far from the equilibrium distribution. The belief learning with full information model outperforms other models. The stationary equilibrium is never attained by the belief learning with partial information model. Agents learn not to use nondurable goods as money by the reinforcement learning model which does not use information on the distribution of goods. It is surprising that providing partial information on the distribution of goods is rather detrimental for attaining emergence of a nondurable goods money.