IMF Publishes Paper on Mapping Contagion in Euro Area Banking Sector

Overall, this paper aims to overcome certain data and modeling gaps in the interconnectedness literature by studying the degree of contagion and vulnerability of euro area significant institutions within the global banking system. The paper presents the data infrastructure and details the topology of euro area interbank network of large exposures. The comprehensive data infrastructure allows to build a detailed modeling framework capturing the specificities of prudential regulations such as minimum capital requirements, macro-prudential capital buffers, the liquidity coverage ratio, and large exposure limits and their interplay with credit, funding, and fire-sale risks. The paper discusses the results based on the contagion and vulnerability score, and provides a multitude of sensitivity tests to determine non-linearities and to check robustness of the results. The paper shows how a range of macro-prudential tools can be fine-tuned to reduce contagion and derives policy implications. 

This euro-centric systemic risk assessment highlights that the degree of bank-specific contagion and vulnerability depend on network specific tipping points directly affecting the magnitude of amplification effects. This leads to the conclusion that the identification of such tipping points and their determinants is the essence of an effective micro- and macro-prudential supervision. In this paper, it is argued that, in isolation, variations in bank-specific characteristics seem to play a lesser role than the network structure in changing the degree of amplification effects (non-linearities). Large and uneven shifts across banks in contagion and vulnerability indicators observed when changes in bank-specific characteristics are combined with changes in network structure point to the importance of non-linearities arising from their interactions and their heterogenous impact on banks. In a variety of tests, heterogeneity in the magnitude of bilateral exposures and of bank-specific parameters is detected as a key driver of the total number of defaults in the system.

The results are network and model dependent based on an incomplete set of bilateral exposures. Therefore, both dimensions need to be extended to include additional channels of contagion and in turn improve the loss estimates of an extreme event. Work should be done to incorporate euro area less significant institutions to complete the euro area banking system, to incorporate financial corporations to model the complex interactions within the financial system, and to incorporate exposures to the real economy to capture feedback loops. It is also important to investigate the role of additional prudential requirements that are currently missing in the framework, such as binding leverage and net stable funding ratio, and are being implemented according to the internationally agreed Basel standards. Finally, given the significance of network structure in determining contagion risks, it would be important to study how the network structure changes, over time and in response to systemic shocks, as well as how different network structures impact model results.

Related Link: Working Paper

Keywords: International, Europe, EU, Banking, Large Exposures, Systemic Risk, Macro-Prudential Policy, Stress Testing, Network Analysis, Data, Contagion Mapping, Research, IMF