Systemic Risk in Financial Networks

Many models of financial systems do not take the network structure - connecting financial agents - into account, even though it has been conjectured sinee long and demonstrated recently, that this network structure is of prime importance. Network structure influences global stability of financial systems, the default risk of individual agents or institutions, the efficiency of the system and the possibility to share risk within the system. With the possibility to model complex systems an an agent based footing, together with the availability of strong computing power it becomes possible to address and model realistic financial networks and their function an a quantitative basis.In the proposed project we combine modern (dynamic) network theory with standard utility based economics to computationally model a global network of interconnected banks. These banks are in contaet with each other by trading financial instruments, the utility functions of the banks are based an profit maximization and risk reduction. Starting from preliminary work (Thumer et al., Quantitative Finance 3, 306-319, 2003) we aim to develop a model which can be used to answer hitherto un-answerable real-world questions in the context of systemic risk in our globalized financial world.