WEALTH DISTRIBUTION IN AN ARTIFICIAL FINANCIAL MARKET WITH AGENT ADAPTATION Cover Image

ДИСТРИБУЦИЈА БОГАТСТВА НА ВЕШТАЧКИМ ФИНАНСИЈСКИМ ТРЖИШТИМА СА АДАПТИБИЛНИМ АГЕНТИМА
WEALTH DISTRIBUTION IN AN ARTIFICIAL FINANCIAL MARKET WITH AGENT ADAPTATION

Author(s): Ognjen Radović, Jovica Stanković, Ivana Marković
Subject(s): Supranational / Global Economy
Published by: Универзитет у Нишу
Keywords: scale-free networks; artificial financial markets; wealth distribution

Summary/Abstract: The aim of this paper is to analyze the influence of the process of adaptation and network structure on the final wealth distribution of trading agents. The analysis was conducted on an electronic financial market represented by a complex scale-free network. In a trading simulation, the flow of information between the network nodes (traders) influences the decision-making process in terms of investment. The Widrow-Hoff algorithm for adaptation and the size of a complex network are the key aspects of the process. The analysis indicated that the ability to adapt the level of self-confidence and imitate wealthy agents decreases the effect of increase in the network size and maintains the wealth distribution at approximately the same level. Using Pareto’s model, we will show a two-fold outcome. On the one hand, the increase in the size of the network and the distance between the nodes increases the even distribution of wealth among the wealthier traders, and increases the gap between the poorer trading agents. The cause of this model behavior can be found in the relatively quick exchange of information between the wealthier trading agents, and the slower exchange between the poorer trading agents. This behavior is a consequence of the increase in the average distance between the network nodes with an increase in the network diameter. The computer model included in the analysis was designed in the NetLogo modeling environment, while the statistical analysis of the complex network was performed using the Pajek and Origin programs.

  • Issue Year: XXXIX/2015
  • Issue No: 4
  • Page Range: 1149-1163
  • Page Count: 15
  • Language: English
Toggle Accessibility Mode