Complexity Models for Systemic Instabilities and Crises

The purpose of this workshop is to promote a multi-disciplinary collaboration between economists, physicists, mathematicians and computer engineers in order to develop complex systems based approaches aimed at understanding systemic instabilities and financial economic crises.

Existing macro models failed to predict the recent global financial crisis. Traditional models are based on highly restrictive assumptions and simplify away the complex network of interactions and feedback mechanisms that characterize financial systems and economies for the sake of analytical tractability. In fact, current economic tools available to policy makers do not incorporate the institutional and financial structures needed to provide insights into liquidity crises, housing bubbles and other phenomena that occurred in the recent crisis. In recent years an alternative view has been developing where markets are considered as complex evolving systems, emphasizing a bottom-up approach to model the financial system and the macro-economy as a network of interacting agents. According to this view, aggregate market phenomena, such as financial crises, are thought of as emerging properties of complex systems resulting from the interaction of many heterogeneous households, firms, banks, investors, etc.

Research on agent-based models of systemic instabilities and financial crises requires a joint effort of experts in complexity research from different fields, such as economics, finance, psychology, physics, mathematics, computer science, biology, and the use state-of-the-art ICT tools. Main topics of the workshop include:

·         evolutionary market dynamics in interacting networks of heterogeneous boundedly rational agents;

·         behavioral models of economic decision making: theory and laboratory experiments with human subjects;

·         agent-based models of linkages and transmission mechanisms between financial markets and the macro-economy;

·         development of ICT based policy support tools from complexity modeling.