Journal of Network Theory in Finance

Welcome to the first issue of Volume 5 of The Journal of Network Theory in Finance.

Nicolas K. Scholtes’s paper “Default cascades and systemic risk on different interbank network topologies” starts us off by investigating the relationship between the topology of interbank networks and their ability to propagate localized, idiosyncratic shocks across the banking sector via banks’ interbank claims on one another. The paper proposes a probabilistic network simulation methodology capable of generating realistic interbank networks that are not random but instead exhibit properties such as a power law degree distribution, disassortative mixing and a tiered/core–periphery structure. In particular, Latin hypercube sampling – a Monte Carlo simulation methodology – is used to obtain a stratified sample of networks on which the same cascading defaults model is run. This paper is not only the first to consider cascading defaults within the context of realistic simulated interbank networks but also the first to posit a link between a variety of topological measures of the network and the extent of the default cascade, which the authors identify empirically. The results mainly highlight the important role played by the network structure in explaining the extent of the default cascades. The impact of the balance sheet and network-related variables varies significantly depending on the default cascade measure used as well as the presence of liquidity effects and whether the shock is random or targeted. The paper ends by enlightening us on future possible extensions and further studies.

“Credit rating analysis based on the network of trading information” by Ximei Wang, Boualem Djehiche and Xiaoming Hu is the second paper in this issue. It highlights two main categories of credit rating methods – traditional credit modeling and statistics-based modeling – and reports that none of these methods explicitly includes the network of trading information as a factor for credit rating, although all economic agents (ie, individual customers, companies, institutions and countries/ areas) belong to some networks with common activities. This is the key issue around which the paper is built. Popular tools, such as assortativity analysis, community detection and centrality measurement, are introduced to analyze the topology structures and properties of the network of trading information (NoTI). This is combined with a study of the correlation between the characteristics of the network and the credit ratings. An empirical investigation of sovereign rating analysis based on the world trade network (WTN) is also presented as a case study. The results show that the economies in the WTN are connected to most of the economies in the world, but only a few of the links are intense. A core and a periphery are identified in the network; each core in the communities is linked actively with their neighbors in the
same community. A correlation-based study between credit ratings and NoTI showed a clear improvement in credit rating prediction accuracy compared with actual metrics. Overall, this demonstrates that network analysis plays an important role in the credit rating problem.

The issue’s third paper, “The liquidity of credit default index swap networks” by Richard Haynes and Lihong McPhail, looks at the implications of recent regulatory reforms such as the mandatory clearing of standardized swap contracts and mandatory trading on centralized execution platforms on credit index swap trading. Market liquidity is studied using quantity-based and price-based liquidity measures such as daily volumes, price impacts and price dispersion. The factors behind credit index trading costs executed on the largest credit default swap trading platform over a sample period of two-and-a-half years are also examined. It is found that trading conditions have generally been stable or have improved. Moreover, the authors claim, trade sizes may have declined slightly, perhaps due to automation or to customers optimizing trade size to achieve the best execution. It is interesting to read that, according to the analysis, customers trading with more dealers (high network degree) and those connected with more active dealers (high network centrality) incur lower trading costs. This provides some evidence that in cases where customers are able to reach more intermediaries they may be able to mitigate the cost of their trades.

Enjoy reading!
 

Tiziana Di Matteo
King’s College London