Agent-Based Modeling Exploring Sovereign CDS Spreads in Europe

In this thesis, we systematize the studies of an agent-based modeling in finance. We introduce the application of the agent-based modeling to the detection of the linkage between two complex systems on the macro- and micro-levels – the European market of sovereign Credit Default Swaps (macro-level) and the Italian equity market (micro-level). To execute this, we develop a framework for the capacity to slave the simulated dynamics of the first market to the changes in the dynamic direction of the second market in a long term by the transformation of the $-game slaving algorithm in the part of the choice of the optimal strategy. Moreover, by means of the Tremor Price Dynamics, we explain the practicality of the global market dynamics for the predictions of the local market fluctuations. Then, we use the effect of decoupling to obtain the knowledge about the time of (negative) bubble occurrence in the European market of sovereign CDS, and on its base we divide analytically the time series of two markets into separate bubble-periods. As a result, we obtain the information confirming the lead-lag linkage between the European market of sovereign Credit Default Swaps and the Italian equity market which could not been caught by means of the standard approach. Finally, we demonstrate a simple example of how to make the predictions of the changes in the dynamic direction of the lagged local market on the base of the bubble-periods appearing on the leading global market.