Journal of Network Theory in Finance

The Journal of Network Theory in Finance is being launched at a time of ever-increasing connectivity and complexity in financial markets and in the financial system. This complexity and connectivity is currently being tackled by regulators, who collect ever-more-detailed information on links between financial institutions for macroprudential supervision, by asset managers, who need to understand the complex time-varying dynamics of interdependencies in financial markets, and by risk managers, who want to understand how emerging systemic risks can be identified as they cascade through the financial markets.

In recent years, network theory has proven useful in applications ranging from cancer research to the social graph. Applications of network theory are quickly becoming ever more present in finance, with network analysis providing answers to questions where traditional analysis methods are weak, and leading to improved models across wide types of financial risks. In fact, networks underlie virtually every type of risk, including liquidity, operational, insurance and credit risk.

The Journal of Network Theory in Finance aims to bring together research currently carried out in disparate areas at universities and by policymakers and industry practitioners. This research has often been published in a wide variety of journals across physics, finance, economics and other disciplines, or it remains unpublished due to the avant garde nature of the field. The publisher and the editorial board therefore see great value in launching this interdisciplinary journal for publishing academically rigorous and practitioner-focused research on the application of network theory in finance and related fields.

On September 23, 2014 a conference entitled "Network Theory and Financial Risk" was held at the Centre for Risk Studies at Cambridge University to inaugurate the journal, and submissions to that conference form the backbone of the journal's first issue. The papers reflect the scope of the journal well, ranging from applications in asset management to the measuring of counterparty exposures between financial institutions.

The first paper in the issue, "Eccentricity in asset management" by Hakan Kaya, considers connectivity among financial assets and investigates whether a node's position in the network can predict the magnitude of the asset's returns, and whether the network structure can explain systemic events. The author finds that assets that are located near the center of the network tend to have higher returns and shows that an investment strategy based on this information has historically provided value. Importantly, the paper shows how methodologies from network theory can help reformulate long-standing questions in finance and provide new insights.

Understanding interbank exposures has been a focus of financial stability analysis in recent years. The issue's second paper, "Emergence of the EU Corporate Lending Network" by Grzegorz Hałaj, Urszula Kocha´nska and Christoffer Kok, extends the analysis to bank-firm relationships by developing a network formation algorithm that estimates these relationships based on largely public data. The authors find that contagion can be amplified when this transmission channel is taken into account. The work will likely find direct applications in stress testing both by regulators and at banks.

Our third paper, "Risk diversification: a study of persistence with a filtered correlation-network approach" by Nicoló Musmeci, Tomaso Aste and T. Di Matteo, addresses the problem of correlation structures observed in the past not always persisting into the future. Finding persistent structures is important for risk diversification and the authors develop a new clustering algorithm to identify such structures. Clusters of assets identified from correlation structures are more economically meaningful and can be more stable than, for example, industry-based classifications, and they are therefore important for better investment and risk management decisions.

Much of the initial analysis on financial interlinkages has focused on particular financial instruments for which data has been available. The issue's fourth paper, "A multiplex network analysis of the Mexican banking system: link persistence, overlap andwaiting times" by José-Luis Molina-Borboa, Serafin Martínez-Jaramillo, Fabrizio López-Gallo and Marco van der Leij, overcomes this simplification and specifically studies the interaction between several layers of interconnectivity across markets (repos, uncollateralized loans, cross holdings, etc). The paper provides very rich insight into the complex multiplex nature of the Mexican financial system and will help other researchers understand and model how these networks interact in other countries where data at such a detailed level is not available, or where the data is confidential.

Last but not least, on behalf of the editorial board I welcome readers to the inaugural issue of The Journal of Network Theory in Finance. I hope the journal will foster both weak and strong links among the community of researchers interested in financial networks, and that the techniques and findings presented therein find valuable application among practitioners.

Kimmo Soramäki
Financial Network Analytics Ltd.