Authors: Georgios Sarantitis, Periklis Gogas, Theophilos Papadimitriou

Title: Modelling the transmission of financial crises using simulation methods

Abstract

In the modern, globalized financial setting, economies and markets internationally are becoming increasingly interconnected. Political, financial and trade linkages connect countries worldwide, regardless of their actual geographic distances. These linkages are able to induce economic convergence and development, but may also become pathways of systemic shocks’ dispersion if a crisis manifests in one or more countries.

In this study, we use the simulation method called Susceptible-Infected-Susceptible (SIS) model to study the diffusion of an economic crisis between 43 countries from Europe and the wider Mediterranean basin. The SIS belongs to the family of parametric models in which a researcher builds a complex network of agents and sets the initial parameters of the model and the response of its agents to a given stimulus. The researcher then introduces a shock to one or more of the system’s agents and is able to observe the evolution of the 'infection' spreading to the rest of the network over time.

Our empirical findings indicate that the countries that are most likely to spread a crisis are the larger ones in terms of Gross Domestic Product and total bilateral trade. However, this phenomenon is non-linear since we find relatively small countries like Belgium which can initiate a systemic crisis while there are countries with high GDP levels and trade volumes that only trigger a moderate crisis’s contagion (such as the United Kingdom). Finally, we observe that the relatively smaller countries of the South, such as Greece and Portugal, do not cause a major crisis’s dispersion when it initiates in only one of them, however, they can induce a systemic shock in the 43-country network when a crisis manifests in both of them simultaneously.