Agent-Based Modeling of Global Agro-Food Market

With the turbulence in global trade and the necessity to develop non-commodity exports in Russia, the choice of an effective strategy for exporters’ conduct in the food market is becoming vitally important these days. Agent-based models simulating the behavior of decentralized self-learning agents with their own goals and capabilities can be used as the tools for analyzing and predicting market movements. The paper presents the results of the agent-based approach to market modeling by the example of barley, which is one of the top commodity items of the Russian food exports on the basis of FAOSTAT and Rosstat data for the period 1997–2017. As a result of the study, an agent-based model of the world barley market was built, and a series of calculated experiments was carried out in the AnyLogic development environment with the changes in such factors as the level of global demand, the amount of customs duties, the exporting companies’ funds in order to determine the strategic conduct of the exporting agents taking into account the limited rationality of the participants in communicative interaction. The proposed approach develops modern theory of consumer behavior and simulation, and the results of the study can be used in the formation of development programs for the Russian agro-food exports.