Agent-based computational economics and exchange rate modeling: A Turkish case

Mainstream economics based on dynamic stochastic equilibrium models has suffered from rigorous assumptions of rationality and homogeneity, and framework of representative agent ruling out interactions between agents. Starting from late 1990s, agent-based computational approach has become increasingly popular in social sciences, especially in financial economics, industrial organization, macroeconomics, political economy, and economic network formation. Finally, 2008 global financial crisis has caused mainstream to be argued and welcome agent-based approach loudly. This new premising approach enable researchers to construct artificial worlds where agents ranging from passive world features to active decision makers who have states and rules of behavior. In this artificial world openness to the interactions which occur between two agents or between an agent and its physical environment lead agents be adaptive (learning) and create a complex adaptive system. Having bottom-up approach, agent-based models (ABMs) offer many advantages for agent-based computational economics (ACE) researchers while it has some insufficiencies. One of field that gets benefit from ABMs is the exchange rate market. In this market heterogeneity is provided by two different forecasting rules: fundamentalists and chartists. In my study, I try to explain the main dynamics of the model and explain the agent-based approach with the help of the simulations performed. I finally compare the simulation outcomes with Turkish exchange rate market and interpret behavior of agents in Turkey.