Economic Themes (2012) 50 (4) 15, 683-700


Ognjen Radović, Ksenija Denčić-Mihajlov

Abstract: Standard financial models are based on the assumptions of the market agents' rationality and the efficient market theory. A significant empirical evidence indicates the unreality of these assumptions. Agent-based modeling and simulations represent a new approach in the economic systems’ monitoring and modeling by using nonlinear dynamic systems. Agent-based modelling is a computational method based on synthetic approach for building a model as a set of a number of autonomous entities (agents) and simulation of their behaviour and interactions. The aim of the paper is to give a review of the postulates of agent-based modeling and to stress the advantages of this approach in comparison to classical financial models. to show its application at the financial markets. Besides that, the authors develop a simple agent-based model of financial market with the periodic auction. Resulting time series of the model of artificial financial market replicate successfully the most important statistical features of the returns at the real financial markets and emphasizes the deviation of the normal distribution as a pillar of the classical financial theory.

Keywords:  Agent-Based Modeling; Agent-Based Simulation; Agents; Artificial Stock Market

PDF file