Behavior of financial agents in an artificial market developed with the Particle Swarm Optimization algorithm

• Autor(a) principal: Nascimento, Kerolly Kedma Felix do
• Data de Publicação: 2020
• Outros Autores: Santos, Fábio Sandro dos, Jale, Jader da Silva, Ferreira, Tiago Alessandro Espínola
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Research, Society and Development
• Texto Completo: https://rsdjournal.org/index.php/rsd/article/view/4216
• topic: Financial markets; Computational Simulation; Optimization; PSO; Mercados Financeiros; Simulação Computacional; Otimização; PSO; Mercados financieros; Simulación computacional; Optimización; PSO

Financial markets are complex systems in which traders interact using the most varied strategies. Computational techniques that use intelligent agents can assist in decision making in order to maximize gains. In this sense, the objective of this article is to observe the behavior of financial agents participating in simulated markets and infer about the gains of these agents. Through the Particle Swarm Optimization algorithm, we used two distinct groups of intelligent agents: one group uses a degree of belief in the prediction of assets for the next day and the other group does not use, in which both interact with each other seeking to maximize their gains. An exploratory research was carried out, with quantitative analysis on the data. The results showed that the group that uses the forecast is more homogeneous, showing higher average wealth gains, with capital and acquired stock concentrations varying according to the historical price series used (Bitcoin, Ethereum, Litcoin, or Ripple). Therefore, the implemented procedure can be improved and used for the development of environments aimed at a better understanding of financial markets and assisting market participants in the definition of trading strategies that enable the minimization of financial losses.