Co Mapping Contagion in the Euro Area Banking
Co. Mapping Contagion in the Euro Area Banking Sector Giovanni Covi*, Mehmet Ziya Gorpe**, Christoffer Kok*** Abstract This paper presents a novel approach to investigate and model the network of euro area banks’ large exposures within the global banking system. Drawing on a unique dataset, the paper documents the degree of interconnectedness and systemic risk of the euro area banking system based on bilateral linkages. We then develop a Contagion Mapping (Co. Map) methodology to study contagion potential of an exogenous default shock via counterparty credit and funding risks. We construct contagion and vulnerability indices measuring respectively the systemic importance of banks and their degree of fragility. Decomposing the results into the respective contributions by credit and funding shock provides insights to the nature of contagion which can be used to calibrate bank-specific capital and liquidity requirements and large exposures limits. Disclaimer: This Paper should not be reported as representing the views of the European Central Bank (ECB) and of the International Monetary Fund (IMF). The views expressed are those of the authors and do not necessarily reflect those of the ECB and the IMF. * European Central Bank, Macroprudential Policy and Financial Stability Directorate, Stress Test Modelling Division, Frankfurt am Main. Email: giovanni. covi@ecb. europa. eu ** International Monetary Fund, Monetary and Capital Markets Department, Washington DC. Email: mgorpe@imf. org *** European Central Bank, Macroprudential Policy and Financial Stability Directorate, Stress Test Modelling Division, Frankfurt am Main. Email: christoffer. kok@ecb. int
Rubric OUTLINE Table of Contents 1. Outline 2. Data Infrastructure 3. Co. Map Methodology 4. Key Findings 5. Future Developments 6. Conclusion Mapping Contagion in the EA Banking Sector 2 www. ecb. europa. eu ©
Rubric MOTIVATION Motivation “The financial crisis really was a stress test for the men and the women in the middle of it. We lived by moments of terror. We endured seemingly endless stretches when global finance was on the edge of collapse, when we had to make monumental decisions in a fog of uncertainty, when our options all looked dismal but we still had to choose” (Geithner, 2014: 19). Enter presentation title by changing the footer. 3 www. ecb. europa. eu ©
Rubric OUTLINE Literature Review 1. Market-based Data 2. Bank-Balance Sheet Data q Seminal Work: Allen and Gale (2000), Eisenberg and Noe (2001); q Simulated Networks: Nier et al. (2007), Lu and Zhou (2010), Halaj and Kok (2013), Alter et al. (2014); q Risk Topology: Battiston et al. (2012), Craig and von Peter (2015), Glasserman and Young (2016); q Multilayer: Kok and Montagna (2013), Bargigli et al. (2015); q Liquidity Shock: Brunnermeier and Petersen (2009), Gai et al. (2011); q Fire Sale: Espinoza-Sole (2010); Caballero and Simsek (2013); Caccioli et al. (2014); Cont and Schanning (2017). q Coverage: Country-specific and market-specific (AT, DE, IT Interbank Market). Mapping Contagion in the EA Banking Sector 4 www. ecb. europa. eu ©
Rubric OUTLINE Objectives 1. Macroprudential Surveillance Tool a. b. Identification of Tipping Points (Contagion and Vulnerability) Modelling Credit and Funding Risks. 2. Fine-tuning Prudential Measures Based on Counterfactual Analyses: a. b. Capital Requirements & Capital Buffers (OSII & SRB). Liquidity Coverage Ratio & Large Exposures Limits. 3. Quarterly Updatable Data Infrastructure a. b. Large Exposures Network Bank-Specific Parameters. Mapping Contagion in the EA Banking Sector 5 www. ecb. europa. eu ©
Rubric DATA INFRASTRUCTURE Large Exposures Dataset Granular Bank and Exposure Level Information – Counterparty Information: • Individual vs Connected Clients • LEI code, Country, Sector, NACE – Exposure Breakdown into: • Asset Classes • Maturity Buckets – Large Exposures Coverage: • Interbank Network: 2. 3 EUR Tr. • FC Network: 1. 4 EUR Tr. • GG Network: 3. 7 EUR Tr. • NFC Network: 4. 5 EUR Tr. • Total: 12 EUR Tr. (50% Total Assets) • Updatable on a Regular Basis. Mapping Contagion in the EA Banking Sector 6 www. ecb. europa. eu ©
Rubric DATA INFRASTRUCTURE EA Interbank-Network of Large Exposures • Network Structure: 200 Counterparties for 1. 1 Trillion of Net Bilateral Exposures. Borrower Perspective - Panel (a) Lender Perspective - Panel (b) INT EA EA Source: COREP Supervisory Data: Tables C. 27 -C. 28 Note: the size of the nodes captures the absolute degree of interconnectedness. The colours of nodes are clustered by country of origin, the thickness of the flows summarizes the value of the exposures in EUR billion, the colour of the flows refers to the borrower/lender colour capturing respectively the borrower/lender perspective. • Mapping Contagion in the EA Banking Sector 7 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Contagion Risk: Solvency and Liquidity Default • Distress Threshold Capital Loss Given Default Discount Rate • Mapping Contagion in the EA Banking Sector Funding Shortfall Liquidity Shortfall 8 Asset Pool Liquidity Surplus www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Model Output 1. Contagion Index: system-wide losses induced by bank i in percent of total capital in the system (excluding bank i); 2. Vulnerability Index: average loss experienced by bank i across all simulations in percent of its own capital. 3. Sacrifice Ratio: ratio of system-wide losses due to the failure of a bank over the cost of a rescue package extended to the bank. 4. Amplification: ratio of total losses in all subsequent rounds over the losses that occur only in the first round. 5. Channels: • Type: Credit Risk and Funding Risk • Regional: Global, EA, XEA, Country-Specific. Mapping Contagion in the EA Banking Sector 9 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Results (1) – Main Outputs Mapping Contagion in the EA Banking Sector 10 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Results (2) – Systemic Risk Map Note: Contagion and vulnerability indexes are normalized by dividing each index for the entity’s maximum value. The vulnerability index is here reported in absolute terms, i. e. considering the EUR value of capital depletion, and not as a % of the capital base. Mapping Contagion in the EA Banking Sector 11 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Results (3 a) – Bank Heterogeneity Note: Special Case 1 refers to Average Parameters, whereas Special Case 2 uses average parameters and set θ = Infinity and γ = 0. Mapping Contagion in the EA Banking Sector 12 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Results (3 b) – Bank Heterogeneity Breakdown Contagion Index: Credit and Funding Shocks Note: Euro Area Contagion Index refers to induced capital losses to Euro Area banking System (only) as % of EA banking system’s total capital. The contagion index is further decomposed into the respective contributions by credit and funding shocks. Mapping Contagion in the EA Banking Sector 13 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Result (3): Stress Test Scenario – Recession Weakened Capital Base – 2016 Adverse Scenario CET 1 Capital Shortfall Benchmark vs Stress Test Scenario Mapping Contagion in the EA Banking Sector 14 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Results (4 a): Prudential tools Fine-tuning prudential Measures Objective: Minimize Second Round Effects (20 default events) • LCR Adjusted = 25 bn Increase (8 banks treated) • Pool Adjusted = 25 bn Increase (8 banks treated) • CET 1 Adjusted = 25 bn Increase (8 banks treated) • Mix = 25 bn Increase (4 banks treated) • CET 1 Adjusted = 10% RWA (9 banks treated) Mapping Contagion in the EA Banking Sector 15 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Results (4 b): Macroprudential tools Calibration OSII Interconnectedness Score • Network measures correlate highly with the more simple sizebased interconnectedness indicators, constructed following the EBA guidelines on the calibration of O-SII buffers. • There is nevertheless value for policymakers to take into account network-based measures. Source: Covi, Kok, and Meller (2018) “Using large exposure data to gauge the systemic importance of SSM significant institutions”, Macroprudential Bulletin, Issue 5 Notes: The institutions represented are euro area significant institutions in the large exposures sample. The cut-off date for data refers to the third quarter of 2017. Mapping Contagion in the EA Banking Sector 16 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Future Developments From Contagion Modelling to Stress Test Modelling 1) Adaptive: Behavioural Rules of Contagion • Distress vs Defaults • Capital, Liquidity, and Leverage Requirements 2) Granular Multi-Sectoral Network • Financial Sector • Non-Financial Corporations • Governments 3) Macro Scenario as Aggregation of Micro Shocks • Modelling Feedback Loops Between Sectors Mapping Contagion in the EA Banking Sector 17 www. ecb. europa. eu ©
Rubric CONTAGION MAPPING METHODOLOGY (Co. Map) Conclusion Key Insights: 1. Large exposures are a powerful tool to assess systemic risk stemming from interconnectedness and contagion. 2. Cross-border exposures important channel of contagion to the euro area banking system. 3. Bank-specific parameters strongly improve results accuracy. § Funding risk calibration is key to disentangle contribution to total risk. 4. Stress test conditions seem to amplify contagion non-linearly. § Highlights the importance of capturing other risk channels and the full picture of bilateral exposures (towards financial and non- financial corporations). 5. Calibration of prudential requirements may benefit extensively from network-based measures of contagion and vulnerability. 6. Co. Map Methodology prudential and monitoring tool to assess threats to euro area financial stability. Mapping Contagion in the EA Banking Sector 18 www. ecb. europa. eu ©
Rubric DATA INFRASTRUCTURE Appendix – Data Sources (1) COREP and FINREP Supervisory Data • C. 27 & C. 28 – Interbank Network of Large Exposures • C. 30 – Large Exposure Maturity Breakdown • C. 67. 00. a – Ten largest funding Sources • C. 66. 00. a – Cumulated Counterbalancing Capacity – Cumulated Net Funding Gap • C. 72. 00. a – HQLA Mapping Contagion in the EA Banking Sector 19 www. ecb. europa. eu ©
Rubric DATA INFRASTRUCTURE Appendix – Data Sources (2) COREP and FINREP Supervisory Data • C. 02. 00 – Total Risk Exposure (RWA) • C. 03. 00 – Capital Adequacy (CET 1, TIER 1, Own Funds) – Pillar 2 Requirements • C. 06. 01 – Group Solvency (Capital Buffers) • F. 01 & F. 32. 01 – Total Assets and Financial Assets – Non-encumbered Assets • BANKSCOPE – International Bank Mapping Contagion in the EA Banking Sector 20 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK Calibration • Mapping Contagion in the EA Banking Sector 21 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK Bank-Specific Model Parameters (1) Mapping Contagion in the EA Banking Sector 22 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK Bank-Specific Model Parameters (2) Mapping Contagion in the EA Banking Sector 23 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK Bank-Specific Model Parameters (3) Default Thresholds Mapping Contagion in the EA Banking Sector 24 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK EA Network of Large Exposures Mapping Contagion in the EA Banking Sector 25 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK Results (1 a) – Bank Heterogeneity Euro Area Contagion Index - Model Comparisons Mapping Contagion in the EA Banking Sector 26 www. ecb. europa. eu ©
Rubric CONTAGION MODELLING FRAMEWORK Results (1 a) – Bank Heterogeneity Euro Area Contagion Index - Model Comparisons Mapping Contagion in the EA Banking Sector 27 www. ecb. europa. eu ©
- Slides: 27