A complex system is not simply the sum of its elements. It is possible that all entities are safe in themselves, but the system as a whole is still vulnerable. Systemic risk is the risk of the entire system collapsing due to one or several shocks. If we wish to identify the systemically important financial institutions (SIFIs) as it was proposed by BCBS (2011), we have to consider five factors contributing to systemic risk: size, interconnectedness, lack of substitutes, cross-jurisdictional activity, and complexity of the activities. When measuring interconnectedness, we can rely on network data and can apply several methods, for example, centrality measures, stress test, and core-periphery models.
Now, we illustrate the first method based on an index of some centrality measures, as described in Komárková et al.(2012) and von Peter (2007). Banks with the highest index-value can be considered as the most central ones, thus with the most...