Financial spillovers from cross-border bank exposures: A network analysis

Identifying and estimating spillovers has become a key agenda for financial stability oversight. This note advances the analysis of financial spillovers and contagion of systemic risks in the cross-border banking system. Extending the network analysis by Espinosa-Vega and Sole (2010), this note provides a quantitative assessment of systemic risk and spillover channels, illustrating how and to what extent a systemic bank failure in a country could have spilled over to other countries by analyzing the interbank claims, cross exposures on sovereign securities, and off-balance sheet exposures. The empirical results show that bank failures in peripheral European countries could have wider ramifications beyond the local banking systems, particularly if financial centers such as United Kingdom and United States are affected.

Background:

This paper identifies key systemically important spillover channels for a range of banking systems among advanced countries. The methodology is based on Espinosa-Vega and Sole (2010), which simulates the failure of banks and tracks the spillovers to other countries. Wider spillovers would suggest higher importance of potential linkages. This approach does not only consider spillovers through direct linkages through exposures, but also through third parties by considering the "domino effect".

The approach also tracks the spillover effects on affected counterparties from both asset and liability sides. It considers two separate shocks: (i) the impact of a banking system defaulting on its liabilities to foreign banks (credit shock), and (ii) the impact of a banking system deleveraging by withdrawing funding from foreign banks, forcing the latter to deleverage as well by selling assets at a discount (funding shock). While this may be considered as tail risks, the analysis help illustrate relative importance of systemic linkages  through the global banking network.

The scope of the analysis covers bilateral exposures among 30 countries. These countries are Australia, Austria, Belgium, Brazil, Canada, China, Chinese Taipei, Denmark, Finland, France, Germany, Greece, Hong Kong SAR, India, Ireland, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, South Korea, Sweden, Switzerland, Turkey, United Kingdom, and the United States. All bilateral exposure data come from the BIS consolidated banking statistics as of end-2011, including the sectoral breakdowns of bilateral exposures across sectors. Data on banking sector capital are obtained mainly from Financial Soundness Indicators database (www.fsi.org).

The paper considers three simulations. First considers only on interbank exposures (Simulation 1). The second simulation captures potential exposures at default, i.e. outstanding derivative contracts or contingent liabilities (guarantees, credit commitments) vis-à-vis the defaulting banking system (Simulation 2). This allows an analysis of spillover risks from countries that have a large presence in the derivatives or off- balance sheet markets such as the U.S. The last scenario captures the potential knock-on effects of banking sector distress on the non-bank and sovereign sectors of each country (Simulation 3). This allows us to capture spillover effects of banking problems within countries, as recently witnessed in several Euro zone countries.