A Symposium on Agent-Based Computational Methods in Economics and Finance

Context

Agent-based Computational Economics (ACE) deals with the computational study of economies as complex adaptive systems, composed of interacting agents with cognitive skills. This area has provoked a great deal of academic interest in various fields of Economics, especially in relation to Complex Systems approaches.

One of the first uses of agent based models was popularized by Axelrod in his research on the evolution of cooperation. In this early work he used extensive computational simulations in order to study strategic behaviour in the iterated prisoner's dilemma. This work still influences contemporary research in various scientific fields. It has, for instance, been the foundation of a new approach to game theory based on computational ideas.

In the mid eighties, under the impulse of the Santa Fe Institute, and especially Christopher Langton, a new field of research, called Artificial Life (AL), emerged. The idea of AL was to mimic various aspects of real life, and to understand the basic principles of biological organization. This led to wider ideas such as complexity, evolution, auto-organisation and emergence. These concepts have influenced social scientists.

The initial attempts to mix agent computing and social sciences include pioneering ACE work in finance, specifically the Santa Fe Artificial Stock Market Model of Arthur, Holland, LeBaron, Palmer and Taylor. This model, based on bounded rationality and inductive reasoning, has led to a new generation of models that closely reproduce real world stock market dynamics. This research program has been carried out by people from various scientific fields (economics, game theory, computer science, finance, and physics).

Recently, another growing field has appeared, dealing with the study of the formation and dynamics of social networks. To understand the spread of information and social beliefs, one has to consider underlying social networks, whose connectivity and topology play important roles in mediating agent-agent interactions. This specific topic is yet another, where physicists, computer scientists and economists have worked together to explore the idea that economic activity is embedded in social structure.

ACE is an important new methodology applicable to many Social Sciences (Management Sciences, Sociology, Economics, Conflict Resolution, and so on). It is now widely used both to test theoretical propositions and to build realistic models when analytical solutions are not possible.

Aims and Scope

The main aims of the event are to:
  • Facilitate the meeting of people and ideas from the communities of social sciences (mainly economics, finance and management) and computer science, to build a structured multi-disciplinary approach.
  • Present computer-science based multi-agent methodologies and tools with application to social sciences;
  • Present uses, needs and constraints of agent-based models used by social scientists, to computer scientists;
The Symposium will feature recent scientific advances in the fields of agent-based computational economics, but is also open to methodological surveys. Topics include but are not limited to the following:
  • Market Dynamics
  • Agent-Based Macroeconomics
  • Discrete Choice Models in Economics and the Management Sciences
  • Emergence and Dynamics of Norms and Conventions
  • Financial Market and Organization Models (Stock Price Dynamics, Herding in Financial Markets)
  • Epistemology and Agent-Based Methodological Issues
  • Dynamics of social and economic networks





Updated: 16-06-2004     Webmaster: Charlotte Bruun
This webpage is build on the AE2005 webpage developed by Bruno Beaufils.