An Agent-Based Market Simulation with Social Effects

This paper investigates an agent-based market model with a focus on social effects on agent behavior and market dynamics. To simulate these effects, widely used social network topologies from previous research were incorporated into the model. In this framework, agents select one belief from a set of possible beliefs, including fundamentalism, trend chasing, and interrupting strategies. Belief updating occurs based on two main factors: the historical performance of each agent’s own belief and the influence of other agents’ beliefs within the network. To formalize this process, a novel opinion formation model was developed, capturing the dynamic interactions among heterogeneous agents. The diversity of agent decisions generates market heterogeneity, reflecting realistic trading behavior. Simulation results demonstrate that returns across all social network topologies replicate key stylized facts observed in real financial markets. Furthermore, the findings highlight the significant impact of social interactions on price formation and market statistics, emphasizing the role of network structure in shaping market behavior. The proposed agent-based framework provides valuable insights into how individual decision-making and social influence jointly contribute to complex market phenomena, offering a robust tool for analyzing the interplay between agent heterogeneity, social effects, and emergent market properties.

کلیدواژه ها:

Social Effects ، Evolution of Beliefs ، Opinion Formation ، Agent-based model