A study on the circulation and retention levels of a Digital Social Currency using an Agent-Based Model

Social currencies are increasingly present in several countries around the world. From management point of view, this phenomenon means that issues such as currency circulation and retention have to be addressed by local managers. In this paper, we propose the Agent-Based Model as a strategic management tool for social currencies. To do so, we introduce a case of using this approach combined with reinforcement machine learning to study the circulation and retention levels of the Sarafu Digital Social Currency from Kenya. The use of this model to study these two aspects that, at first, antagonize one another, allowed us to obtain indicators on the existence of a currency users behavior that contributes to the increase both of the circulation and retention levels. Those indicators can help Digital Social Currency systems managers to elaborate better strategies related to the circulation and retention levels, increasing the social benefits produced by the currency use in the community. The use of an Agent-Based Model to simultaneously address two aspects of the social currency allowed us to study the interactions between those two aspects and constitutes a novelty, since all the cases found in the literature using this model in this field are focused on a single aspect of the currency.