International transmission of shocks and fragility of a bank network

Abstract

The weighted and directed network of countries based on the number of overseas banks is analyzed in terms of its fragility to the banking crisis of one country. We use two different models to describe transmission of shocks, one local and the other global. Depending on the original source of the crisis, the overall size of crisis impacts is found to differ country by country. For the two-step local spreading model, it is revealed that the scale of the first impact is determined by the out-strength, the total number of overseas branches of the country at the origin of the crisis, while the second impact becomes more serious if the in-strength at the origin is increased. For the global spreading model, some countries named "triggers" are found to play important roles in shock transmission, and the importance of the feed-forward-loop mechanism is pointed out. We also discuss practical policy implications of the present work. I. BACKGROUNDThe increasing globalization provides important advantages in terms of risk sharing and risk diversification in banking and financial markets, but it also facilitates the risk spreading among different nations [1]. The recent financial crisis or shock, originated from United States (US) and spread to other countries, has witnessed that many overseas banks cut back their loans to the local markets and withdrew their representations. The withdrawals cause the liquidity shortage of the host countries which then lead to subsequent withdrawals of the host countries' overseas branches in the crisis attacked countries. This process can continue and the shocks in this situation are hence defined as sequential shocks in this study.There are abundant literature on transmission mechanism of shocks in global banking: [2] examines how Japanese asset bubbles have been transmitted to US via lending responses of Japanese overseas banks in US in the 80's; [3] explores the mechanism of 1998 Russian debt default as a negative liquidity shock to international banking and its impacts on the banking of Peru; [4] studies the effect of financial distress in foreign parent banks on the local SME (small and medium enterprises) financing in 14 central and eastern European countries during the early stage of the 2007 financial crisis.Within the framework of complex network theory, this paper introduces models for shock transmission and examines the different impacts of the sequential shocks on the global banking network. We find that the damage of the shocks is closely related to the directionality of the edges and the network topology plays an important role. * Corresponding author: beomjun@skku.edu II. LITERATURE REVIEW ON COMPLEX NETWORK METHODOLOGYThe research using complex system approach in banking has also experienced three stages of the development. In the earlier stage, [6] claims that the possibility of contagion effect depends on the structure of interbank linkages. They believe that a "complete structure of linkages" will share the risk more easily than "an incomplete structure", hence the risk sh...

Assessment of EU banking network regionalization during post-crisis period

“…All negative flows ('net repayments') are replaced with zeros and ignored in the analysis, which is in line with the research of Minoiu and Reyes (2013). Visual images of cross-border banking claim flows networks are drawn using network analysis software Gephi, which is used in a number of previous research (Feng et al, 2014;Feng & Hu, 2013). In addition, countries of the EU 12 and EU 28 banking networks are classified into communities, according to community detection algorithm in Gephi -modularity class, created by Blondel et al (2008).…”

Evaluation of the post-crisis EU banking network connectedness in the global context

“…Each cell w represents the value of the flow from country i to country j at time t. Then, these matrices are transformed into their binary matrices A = a , where each cell a takes value 1 if w > 0 and 0 otherwise. Visual images of banking networks are drawn using network analysis software Gephi (in line with Feng et al (2014) and Feng and Hu (2013)). Countries are classified into communities using community detection algorithm in Gephi -modularity class, created by Blondel et al (2008).…”

Synchronization in simplicial complexes of memristive Rulkov neurons

“…The study of complex systems has yielded essential insights into the structures and dynamics of numerous interconnected systems, ranging from social networks (Shahal et al, 2020 ) and biological systems (Ma and Tang, 2017 ) to network marketing and trading networks (Kim et al, 2006 ; Feng et al, 2014 ; Cho et al, 2023 ), in the discipline of network science. Networks have traditionally been depicted as graphs, with nodes and edges capturing entities and their pairwise connections (Burgio et al, 2020 ; Ghosh et al, 2022 ).…”