Journal of Network Theory in Finance

Welcome to the second issue of Volume 4 of The Journal of Network Theory in Finance.

This issue starts with “Relation between regional uncertainty spillovers in the global banking system” by Sachapon Tungsong, Fabio Caccioli and Tomaso Aste, a paper that focuses on the quantification of systemic risk from market data. It is inspired by a 2009 work by Francis Diebold and Kamil Yilmaz, which proposed a method based on forecast error variance decomposition to estimate, using market data, networks of interdependencies between firms, and used the connectedness of the estimated networks to quantify spillovers of uncertainty between variables. The present paper improves on that work, representing a step forward in our understanding of the connectedness of banking systems. In particular, this paper generalizes Diebold and Yilmaz’s methodology to an exponentially weighted daily returns and ridge regularization on vector autoregression and forecast error variance decomposition. Moreover, it estimates the time evolution of connectedness in the following three regional banking systems: North America, Southeast Asia and the European Union. This allows the authors to perform a comparative analysis of the three regions and a quantification of the existence of causal relations between different regions. Aside from the new results and robust analysis this paper reports, I very much enjoyed reading its very well-crafted and exhaustive literature review. I am sure our readers will enjoy it too.

Our second paper, “The quest for living beta: investigating the implications of shareholder networks” by Matthew Oldham, looks at financial markets as complex systems and applies network theory to study their behavior. A vast literature has developed around this idea, and, in my opinion, the econophysics community has added significantly to its development. This paper contributes nicely to these studies by analyzing the dynamical evolution of bipartite networks as well as the subsequent stock-by-stock and investor-by-investor networks formed for each quarter between 2007 and 2010, comprised of the stocks in the Standard & Poor’s 500 (S&P 500) and the US institutional investment managers that held them. It presents an interdisciplinary and alternative type of study to the traditional standard mainstream economic and finance type of analysis, capturing a novel result: namely, a parallel movement in the density of the investment network and the volatility and value of the S&P 500 index. Nowadays, it is always important to tackle problems by looking at data and using approaches that are interdisciplinary by nature as well as new and able to be applied to different systems.

“A stock-flow consistent macroeconomic model with heterogeneous agents: the master equation approach” by Matheus R. Grasselli and Patrick X. Li, our third and final paper, is another example that proposes the alternative approach of agent-based modeling. It addresses a key theme in macroeconomics: the distinction between the actions of individual agents and aggregate behavior. This is an alternative to both the aggregate-level Keynesian model and the representative-agent-based dynamic stochastic general equilibrium (DSGE) model, and it considers agents not constrained by utility-maximizing behavior and aggregation not achieved through equilibrium. The authors’ main proposal is a mean-field-type approximation to a stock-flow consistent agent-based model with two types of firms and two types of households. This allows them to investigate the behavior of aggregate variables with respect to parameters that are difficult to estimate outside the model, such as the fraction of external financing that firms raise by issuing new debt as opposed to equity. Agent-based modeling relies on numerical simulations, making it both computationally time- consuming and difficult to interpret; however, this paper shows how approximating by means of a mean-field approach might provide results that are not achievable otherwise.

Let me end this editorial letter by thanking this issue’s authors for their valuable contributions, and by reminding you that Risk Journals is proud to sponsor NetSci 2018, the flagship conference of the Network Science Society, which aims to bring together leading researchers and practitioners working in the emerging area of network science. The conference fosters interdisciplinary communication and collaboration in network science research across computer and information sciences, physics, mathematics, statistics, the life sciences, neuroscience, environmental sciences, social sciences, finance and business, arts and design (see www.netsci2018.com).

Tiziana Di Matteo
King’s College London