s p o n s o r e d   l i n k s

ebook artificial agents

Discovering Artificial Economics - How Agents Learn and Economies Evolve.pdf

November 10, 2009 · Filed Under Finance · Comment  · Tags: ,

Taken from Preface: We live in an astonishingly complex world. Yet what we do in our everyday lives seems simple enough. Most of us conform to society’s rules, pursue familiar strategies, and achieve reasonably predictable outcomes. In our role as economic agents, we simply peddle our wares and earn our daily bread as best we can.
So where on earth does this astonishing complexity come from? Much of it is ubiquitous in nature, to be sure, but part of it lies within and between us. Part of it Comes from those games of interaction that humans play-games against nature, games against each other, games of competition, games of cooperation. In bygone eras, people simply hunted and gathered to come up with dimer.
Today you can find theoretical economists scratching mysterious equations on whiteboards (not even blackboards) and getting paid to do this. In the modern economy, most of us make our living in a niche created for us by what others do. Because we’ve become more dependent on each other, our economy as a whole has become more strongly interactive.
A strongly interactive economy can behave in weird and wonderful ways, even when we think we understand all its individual parts. The resulting path of economic development is packed with unexpected twists and turns, reflecting the diversity of decisions taken by different economic agents. But an understanding of economic outcomes requires an understanding of each agent’s beliefs and expectations and the precise way in which the agents interact.
In a strongly interactive economy, the cumulative pattern of interactions can produce unexpected phenomena, emergent behavior that can be lawful in its own right. Yet this is far from obvious if we study economics.

s p o n s o r e d   l i n k s


An Artificial Market Model of a Foreign Exchange Market.pdf

October 23, 2009 · Filed Under Trading · 1 Comment  · Tags: , , , ,

In this study, the author proposes a new approach to foreign exchange (forex) market studies: the artificial market approach - by integrating fieldwork studies and multiagent computer models in order to explain the micro and macro relation in markets, as another downloadable model does by considering final consumers.
The proposed artificial market approach is constituted by three steps:
First, in order to investigate the learning patterns of actual dealers, the author carried out both interviews and questionnaires. These field data made it clear that each dealer improved his or her prediction method by replacing (a part of) his or her opinions about facts with other dealers’ opinion.
Second, the author constructed a multiagent model of a foreign exchange market. Considering the result of the analysis of the field data, the interaction of agents’ learning were described with genetic algorithms.
Finally, emergent phenomena at the market level were analyzed onthe basis of the simulation results of the model. The results showed that bubbles in exchange rates were caused by the interaction between the agents’ forecasts and the relationship of demand and supply. Other emergent phenomena were explained by the concept of the phase transition of forecast variety.
The simulation results were supported by further empirical data.