Heurística para instabilidade financeira sob incerteza keynesiana: uma abordagem baseada em agentes

Abstract

Financial instability is a complex phenomenon that can be caused by various factors, such as economic crises, fluctuations in financial markets, political changes, among others. When it occurs, the financial system can suffer serious consequences, such as currency devaluation, increased inflation, unemployment, and even the bankruptcy of companies and banks. Therefore, it is essential to better understand the factors that lead to financial instability, as well as to develop tools and techniques that can help prevent or mitigate its effects. This work aims to reproduce a financial system under Keynesian uncertainty, composed of heterogeneous agents that interact with each other, with the purpose of studying the aggregate behavior of the system with a focus on financial instability. To do so, agent-based models are used to simulate the interaction in this complex system, combining intelligent methods that assist agents in decision-making, such as reinforcement learning, ant colony algorithm, and clustering. In this way, a heuristic is proposed that integrates different computational techniques, with the objective of understanding the impacts generated on the financial stability of this system. The heuristic is applied to a specific agent of the model and introduces an element of financial "self-regulation" in the system, aligned with the definitions of the Basel accords. Additionally, sensitivity analysis tools are employed to understand the impact of exogenous parameters on this financial system model. It is expected that this work can contribute to a better understanding of agent behavior in the financial context, especially when computational tools are inserted to test theoretical concepts in the area of Economics.