One of the main lessons of the current crisis is that some institutions bear an outstanding risk for the financial system due to their size or special role. During a crisis, these institutions usually get state aid to prevent the whole system from collapsing, which would mean higher costs for the state and the real economy as well. One of the best examples is the AIG. Due to its activity on the CDS market, the Federal Reserve helped the insurer company to avoid defaulting since nobody knew the possible effects of the collapse of the institution.
These lessons motivated central banks and other regulators to put more emphasis on the examination and the regulation of systemically important financial institutions (SIFI). To do this, sophisticated identification of SIFIs is getting more and more important in financial literature. Expanding the former simple techniques, central banks and supervisory authorities tend to use more complicated methodologies based on network theory approaches using transaction data of financial markets. This information is important for investors as well because it helps to diversify their exposure towards the financial sector.
This chapter aims to introduce two techniques based on network theory, which can be used in the identification of SIFIs beyond the commonly used centrality measures.
The global financial crisis highlighted that the size of some financial institutions was too big compared to the real economy, or they had too many connections with important counterparties. Because of this, any problems that affect these institutions can have fatal results on the whole financial system and the real economy. For this reason, governments spared no effort in saving these institutions. There are several global examples where governments or central banks give guarantees, inject capital, lend funding, or support the acquisition of their most important financial institutions (for example, Northern Rock, AIG, or Bear Stearns).
Without these steps, the chance for a collapse seemed to be too high, which would have been accompanied with extreme high costs because of bailouts. All in all, identification of systemically important financial institutions again became a hot topic. One of the main lessons of the crisis was that the biggest and most interconnected institutions have to be handled differently even during normal times. According to the new Basel framework, systemically important institutions have to be more strictly regulated than their less important partners. Due to their central role and their interconnectedness, the failure of these institutions can send shock waves through the financial system, which, in turn, can harm the real economy. The rational choices of individual institutions, which target the maximum possible profit, may be suboptimal on a system-wide level because they do not take into account their possible negative effects during stress periods.
Before the crisis, the systemic role of individual financial institutions was mainly assessed during the decision about the lender-of-last-resort support. Central banks took into account a bank's systemic role in their decision on lending to this bank in case of serious problems. A survey about analysis techniques used in different countries found that in many cases, authorities applied a similar methodology in the assessment of systemic importance. A wide variety of methods exist in practice, from traditional techniques (for example, indicator-based approaches that focus on market shares) and complex quantitative models to qualitative criteria, which include market intelligence (FSB (2009)). Several different types of ratios might be included in indicator-based methods (BIS (2011)). Usually, financial markets, financial infrastructure, and financial intermediation are in the focus of the examination, but the actual set of indicators can vary from country to country, depending on the special characteristics of the investigated banking system.
Indicator-based methods mainly focus on each bank's market share in different parts of banking (from assets to liabilities and from notional values of OTC derivatives to payments cleared and settled, it may cover several fields, BIS (2011)). These indicator-based methodologies sometimes don't contain information about the interconnectedness of the institution on financial markets. Daróczi et al. (2013) provided some suggestions on how to include this information in the identification of systemically important banks. Simple measures of networks applied for each bank can expand the traditional indicator-based methods. In the financial literature, many different measures are used to evaluate the stability of the network or assess the role of individual institutions. Iazetta and Manna (2009) used the so-called geodesic frequency (also known as "betweenness") and degree to assess the resilience of the network.
They found that the use of these ratios helps in the identification of the big players in the system. Berlinger et al. (2011) also used network measures for the examination of individual institutions' systemic role.
In this chapter, we won't include these methods since Daróczi et al. (2013) showed the theory and its application in R. Our focus will be on two different methodologies of network theory, which are relevant in the identification of systemic importance and can be easily applied. First, we will show the core-periphery decomposition of financial markets. Second, we will show a simulation method that helps us to see the contagious effects in case any individual institution defaults.