Microsimulation of artificial stock markets based on trader roles

On financial markets trading takes place continuously and market prices are typically formed whenever two traders make an agreement. Most of the artificial markets, however, implement discrete time modelling and try to set the market price at equilibrium, where most demands and supplies can be matched. This paper describes the structural design of an artificial market environment that supports continuous trading and helps to study how different traders affect market dynamics in different situations. We identify different types of traders and describe an architecture based on their role and the market microstructure where they interact. In order to get an accurate representation of the market dynamics we apply a bottomup microsimulation approach, and further, represent traders by intelligent agents. We start building from a more basic level than current approaches in the sense that we consider continuous order matching mechanisms and implement agents’ behaviour based on their role in the market. For this reason we trace the life-cycle of the orders observing the changes they suffer especially caused by traders’ different decision till they trigger market prices.