The recent financial crisis of 2007/2008 raised awareness on the complexity and the fragility of financial systems. Identifying the interbank exposures in financial networks has become of great interest to regulators and policy makers across the world. However, the confidentiality of financial data prevents the full understanding of these complex structures. Thus, people turned to different mathematical approaches for modelling financial networks. The questions naturally emerging next is how these models differ and how to choose which model is the best to use. Following up on these questions, this thesis deals with the comparison of two network reconstruction models:
the fitness model and the exponential random graph model. We compare the structural differences between the models based on the reproduction of the Italian and the German interbank network. Furthermore, we compare the two models regarding the stability of sampled networks by applying the DebtRank algorithm.
«
The recent financial crisis of 2007/2008 raised awareness on the complexity and the fragility of financial systems. Identifying the interbank exposures in financial networks has become of great interest to regulators and policy makers across the world. However, the confidentiality of financial data prevents the full understanding of these complex structures. Thus, people turned to different mathematical approaches for modelling financial networks. The questions naturally emerging next is how the...
»