Estudo de sistemas de moedas sociais digitais por meio da utilização de um Modelo Baseado em Agentes

• Ano de defesa: 2022
• Autor(a) principal: Ramos, Wagner Vieira
• Orientador(a): Diniz, Eduardo Henrique
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Moedas Sociais Digitais; Modelo Baseado em Agentes; Demanda; Circulação; Retenção
• Palavras-chave em Inglês: Digital Social Currencies; Agent Based Model; Demand; Circulation; Retention
• Link de acesso: https://hdl.handle.net/10438/32677

This thesis addressed the demand for Digital Social Currencies (DSC), and their levels of circulation and retention, trying to identify which factors influence these aspects and how they interrelate. The theory of complex systems and elements of Keynesian theories applied to the DSC context were adopted. To carry out this research, I used an Agent-Based Model (ABM) calibrated for the cases of Banco Preventório, from Niterói, Rio de Janeiro, and Sarafu Currency, from Kenya. To perform this calibration, a reinforcement machine learning method, called Q-Learning algorithm, was used. Among the results obtained in this research, I highlight the fact that the transactional motive, proposed by Keynesian theory, is one of the causes for the acquisition and retention of DSC, and this retention is defined by the time that a user takes to use the MSD in a purchase operation or to exchange it for national currency. Furthermore, I found that when currency retention is associated with this motive it contributes to increasing its circulation level. On the other hand, when it is associated with inactive accounts, retention negatively influences the demand for the currency and its level of circulation. Other results indicate that charging penalty fees for DSC retention decreases the currency circulation level. They also indicate that a small monthly increase in the number of users performing trades and in the number of trades performed would represent a large increase in demand for DSC. As a theoretical contribution, this research presents the association of concepts from Keynesian theories with the theory of Complex Systems as an adequate conceptual basis for the study of DSC. As a practical contribution this research presents the Agent-Based Models as an instrument for the study of Digital Social Currencies, which can be used as a strategic planning tool by managers of Digital Social Currencies systems and by potential donors for these systems. It can also be used as a supervisory tool by regulators such as the Central Bank of Brazil.