Banking systems are at the center of the financial infrastructure of any country. It has become apparent after the subprime crisis that such systems cannot be studied by looking at their components individually (that is, in isolation). Thus, an integrated approach is needed.In this paper we introduce a numerically friendly, yet general, algorithm that allows us to represent a banking system in a realistic way. We start with a detailed description of the banks' balance sheets and we incorporate two different features to account for connectivity effects: interbank loans and correlated-exposures to a common universe of credit-risk loans. The driving force behind the model is the progressive deterioration of the loan portfolios to which the banks are exposed. This method is useful to identify the weaknesses of a given banking system, and it is also helpful to assess the merits of different potential responses from the regulator's viewpoint. An example based on the banking system of a Latin American country is used to demonstrate the merits of this approach.Some of our findings are in agreement with those of previous authors, namely, that the probability of developing cascades is proportional to the degree of connectivity of the banking networks, and that the maximum resilience for a given set on banks is achieved for some "intermediate" (optimal?) level of connectivity. However, we also find -- in contrast with a few earlier researchers -- that investigating the features of a banking system by collapsing banks individually (and keeping other features unaltered) can give a misleading view of the system resilience.

In this paper, we investigate a credit rating problem based on the network of trading information (NoTI). First, several popular tools, such as assortativity analysis, community detection and centrality measurement, are introduced for analyzing the topology structures and properties of the NoTI. Then, the correlation between the characteristics of the network and the credit ratings is investigated to illustrate the feasibility of credit risk analysis based on the NoTI. Sovereign rating based on the world trade network is analyzed as a case study. The correlation between the centrality metrics and the sovereign ratings conducted by Standard & Poor’s clearly shows that highly ranked economies with vigorous economic trading links usually have higher credit ratings. Finally, a simulation is conducted to illustrate the degree of improvement in credit rating prediction accuracy if the NoTI is considered as an additional attribute.

Funding: National Funds from FCT (grant nr. UID/ECO/00124/2013) and POR Lisboa (grant nr. LISBOA-01-0145-FEDER-007722 )

We study the topological and structural properties of the bank–sector credit network of Spain over the period 1997–2007. In particular, we start by analyzing assortativity, different types of motifs and the nestedness phenomenon, both in the bipartite structure of the binary version of the network and in its weighted version. In order to assess the statistical significance of the network properties, we employ the family of the so-called bipartite configuration models, imposing initial constraints such as the degree sequence and/or strength sequence of the observed bipartite network. We find that, in the binary version, the so-called bipartite binary configuration model, maintaining the observed degrees, can replicate the main features of many properties of the observed network. In addition, the one-mode projection matrixes, which indicate lending portfolio overlaps between banks and borrowing portfolio overlaps between sectors, can mostly be predicted by information embedded in the degree sequence of the original bipartite structure. In the weighted version, we observe that the so-called bipartite enhanced configuration model, where both the degree and the strength sequences are preserved, outperforms (on average) the so-called bipartite weighted configuration model, which maintains only the strength sequence in replicating the topological features of the network. Moreover, comparing the observed network to all the referenced null models, we still find a number of features of higher-order topological properties that cannot be explained by information embedded in the observed degree and/or strength sequence. In particular, we discover an “excessive” weighted clustering of the banks and sectors with the highest degrees.

This paper proposes a framework to identify the structure of a financial network and its evolution over time, and presents an application to an interbank market with complete actual data. The framework is based on a methodology popular in the social network literature – stochastic blockmodeling – which, we argue, is more general, transparent and rewarding in terms of results than other proposed methodologies. In particular, it allows us to identify the presence of multiple cores and peripheries as well as the different forms of interaction between them. We find that such a varied core–periphery structure exists in almost all periods for the different instruments analyzed. In the case of term deposits, which account for two-thirds of interbank exposures, we find that, far from being static, the structure underwent a transition in the period 2009–15, with the core increasing its size. We also show that facts revealed by our approach cannot be observed in the metrics commonly used to describe networks. Finally, we describe how the elements identified by our method can be used to single out sources and channels of transmission of systemic risk in a network of banks.

Today’s asset management academia and practice are dominated by mean–variance thinking. Consequently, this usually leads to the quantification of the dependence structure of asset returns by the covariance or the Pearson correlation coefficient matrix. The respective dependence measures are linear by construction and hence unable to detect nonlinear dependencies. This paper tackles the described concern with regard to financial networks and their implementation in active investment strategies. We discuss the mutual information measure, which is an information-theoretic concept and is able to detect linear and nonlinear dependencies. The empirical part of this paper extensively compares mutual-information-based networks with correlation-based networks on a stand-alone basis and in the framework of active investment strategies.

Modern society heavily relies on strongly connected, socio-technical systems. As a result, distinct risks threatening the operation of individual systems can no longer be treated in isolation. Consequently, risk experts are actively seeking for ways to relax the risk independence assumption that undermines typical risk management models. Prominent work has advocated the use of risk networks as a way forward. Yet, the inevitable biases introduced during the generation of these survey-based risk networks limit our ability to examine their topology, and in turn challenge the utility of the very notion of a risk network. To alleviate these concerns, we proposed an alternative methodology for generating weighted risk networks. We subsequently applied this methodology to an empirical dataset of financial data. This paper reports our findings on the study of the topology of the resulting risk network. We observed a modular topology, and reasoned on its use as a robust risk classification framework. Using these modules, we highlight a tendency of specialization during the risk identification process, with some firms being solely focused on a subset of the available risk classes. Finally, we considered the independent and systemic impact of some risks and attributed possible mismatches to their emerging nature.