## EXECUTIVE SUMMARY

## Source details

**Canonical URL:** [EXECUTIVE SUMMARY](https://www.imf.org/-/media/files/publications/cr/2022/english/1irlea2022013.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2022/english/1irlea2022013.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2022/english/1irlea2022013.pdf.json)

---

### Scope and structure
- Document components and major sections include (page starts as presented):
  - INTRODUCTION — 18
  - TOP-DOWN SOLVENCY STRESS TEST OF BANKS — 31 (A–F subsections)
  - BANK LIQUIDITY RISK ANALYSIS — 60 (A–E)
  - BANK INTERCONNECTEDNESS ANALYSIS — 76 (A–D)
  - BANK CLIMATE STRESS TEST ANALYSIS — 86 (A–C)
  - INSURANCE SECTOR SOLVENCY STRESS TEST — 99 (A–E)
  - INSURANCE SECTOR LIQUIDITY RISK ANALYSIS — 110 (A–B)
  - INSURANCE SECTOR CLIMATE RISK ANALYSIS — 113 (A–B)
  - INVESTMENT FUND LIQUIDITY STRESS TEST ANALYSIS — 119 (A–C)
  - INVESTMENT FUND STRESS TEST DESCRIPTIVE AND EMPIRICAL ANALYSIS — 123, 128
- Appendices include I–XII and References (starting pages 133–161).
- Glossary includes acronyms such as CET1, LCR, NSFR, NBFI, PD, PiT, TTC, HQLA, RCR, SCR, etc.

### Background and macro-financial context
- Structural divergence: an innovative, fast-growing international finance sector vs a consolidating retail banking sector.
- Ongoing and emerging risks: persistent inflationary pressures, supply bottlenecks, and the war in Ukraine.
- Pandemic and Brexit impacts uneven; unwinding of public support is a material near-term risk.

### Banking sector — resilience and remaining vulnerabilities
- Strengths:
  - Banks maintained strong capital and liquidity buffers through the pandemic.
- Vulnerabilities (selected):
  - Profitability challenges and long-term mortgage arrears in retail banks.
  - High exposure to the CRE segment.
  - Significant off-balance sheet exposures of large international banks.
- Structural facts and magnitudes:
  - Banking sector assets contracted from 780 percent in 2009 to about 200 percent of GDP in 2021.
  - Funds sector total assets grew from 4.7 times GDP in 2009 to more than ten times in 2021.
  - Retail banks: total assets of €317 billion at Q2 2021; account for more than 90 percent of lending to households; hold 40 percent of total bank assets.
  - Large international banks: total assets €276 billion.
  - Other international banks: total assets €62.5 billion; nine foreign-owned LSIs.
- Pandemic-era indicators:
  - Aggregate CET1 ratio: 22 percent at beginning of 2020.
  - CET1 around 22 percent through to Q3-2021.
  - Liquidity Coverage Ratio: 178 percent as of Q3-2021.
  - Payment relief measures largely phased out later 2021.

### Bank solvency stress test results (scenario-based)
- Sample coverage:
  - 12 banking institutions (about 80 percent of total banking system assets).
  - Scenario-based tests for five retail and three large international banks; sensitivity tests for four other international banks.
- Baseline scenario (capital accumulation):
  - Retail banks: fully loaded CET1 ratios trend from 16.4 percent to 17.8 percent.
  - Large international banks: fully loaded CET1 ratios trend from 19.9 percent to 27.9 percent.
  - No banks fell below hurdle rates under the baseline.
- Adverse scenario (capital depletion):
  - Aggregate fully loaded CET1 ratio change by year 5:
    - Retail banks: declines by about 6.7 percentage points.
    - Large international banks: declines by about 0.4 percentage points.
  - Trough capital depletion:
    - Retail banks: 7.2 percentage points.
    - Large international banks: 2.3 percentage points.
  - Contributors to system-level CET1 decline (percentage points):
    - Credit risk provisioning: about 7.4 percentage points.
    - Interest rate risks: 1 percentage point.
    - Risk weighted assets (RWA): 0.6 percentage point.
    - Market risks: 0.4 percentage point.
  - Dividend distribution contributes about 0.7 percentage point to aggregate capital depletion.
- Hurdle rate thresholds:
  - CET1: 4.5 percent plus O-SII buffer.
  - Tier 1: 6.0 percent plus O-SII buffers.
  - CAR: 8 percent plus O-SII buffer.
  - Capital conservation buffer allowed to be used under the adverse scenario.

### Sensitivity analysis: unwinding of pandemic support
- Assumption: 50 percent of loans with active or expired moratoria deteriorate from stage one and two into stage three assets over five years.
- Result:
  - Additional credit impairment brings an extra 110 bps CET1 depletion to the adverse scenario.
  - One retail bank’s CET1 and Tier 1 ratio fall below hurdle rates under the adverse scenario.
  - Maximum capital shortfall against CET1 hurdle rate: about 0.2 percent of GDP.

### Credit quality, NPLs, and provisioning
- Stage 2 assets doubled since the pandemic.
- As of mid-2021:
  - 66 percent of total NPLs renegotiated and restructured.
  - Retail banks’ NPL ratio: 4.4 percent as of Q2 2021.
  - System NPL ratio: around 2.6 percent at mid-2021.
  - Over 50 percent of total NPLs are in arrears for more than 2 years.
- NPL coverage:
  - Irish banks: around 34 percent.
  - Euro Area average: 47 percent.

### Liquidity resilience and gaps
- LCR and cashflow tests:
  - Aggregate LCR increased from 153 percent at end-2019 to 179 percent at mid-2021.
  - Standard LCR by group (mid-2021):
    - Retail banks: 198 percent.
    - Large international banks: 161 percent.
    - Other international banks: 211 percent.
  - All sampled banks met the 100 percent LCR minimum as of mid-2021.
- Cashflow-based stress test (5-days, 4-week, 3-months, 12-months):
  - Maturity mismatch: 66 percent of total outflows projected within less than 30 days; over 50 percent of inflows materialize beyond first three months.
  - Counterbalancing capacity (CBC) composition: central bank reserves 73 percent; marketable securities 19 percent.
  - CBC depletion (5-day horizon):
    - Mild scenario: decline of 7 percent (system-wide).
    - Severe scenario: decline of 22 percent (system-wide).
    - Large international banks: CBC decline ranges from 8 to 25 percent of total assets.
  - By 12 months CBC depletion reaches up to 12 percent (mild) and 29 percent (severe) of total assets.
  - Under most severe 12-month conditions, two retail, two large international and three other international banks could experience shortfalls up to 4 percent of total sector assets.
  - System-level liquidity shortfalls manageable; large international banks register shortfalls up to 1.8 percent of total sector assets over 12 months in the most severe scenario.
- Funding structure specifics:
  - Loan-to-deposit ratio: 86 percent overall (Large international banks: 89 percent; Domestic retail banks: 75 percent).
  - Non-resident deposits: 88 percent of total deposits for international banks.
  - Off-balance sheet contingent liabilities for international banks: €108 billion or 40 percent of total contractual outflows within 12 months (mid-2021).
  - Retail banks off-balance sheet exposures: around €30 billion (10 percent of total assets).
  - About 40 percent of wholesale funding for large international banks are intra-group funding.
  - Issuance of Insurance Letters of Credit (“ILOCs”) approximately 80 percent collateralized on average.

### Foreign currency liquidity vulnerabilities
- LCR and cashflow exercises in U.S. dollar and U.K. Sterling reveal vulnerabilities:
  - U.S. dollar: international banks had lower starting USD LCR; two banks below 100 percent in mid-2021 in USD reporting subset.
  - Sterling: retail banks have lower starting point and higher depletion rates; one bank in each group below 100 percent in Sterling subset.
- System-level net funding gap under severe scenario (12 months):
  - U.S. dollar: up to 2 percent of total sector assets.
  - Sterling: peaks at 0.1 percent of total system assets.

### Interconnectedness and contagion
- Domestic interbank linkages muted; cross-border and bank–NBFI linkages meaningful.
- Large international banks’ exposures to NBFIs account for 45 percent of their total exposures; around 92 percent of these are with foreign NBFIs.
- Adding top NBFIs increases inward spillover risks:
  - Including domestic NBFIs raises large international banks’ vulnerability index from almost 0 to 15 percent of total capital on average.
  - Including domestic and foreign NBFIs increases indices further; for one large international bank the vulnerability index reaches almost 100 percent of capital.
- Domestic bilateral interbank exposures are small: 0.05 percent of sample bank assets and 0.07 percent of GDP.

### Banking sector climate risk analysis
- Transition risk (carbon-tax shock):
  - Shock modeled: increase carbon tax from €33.5 to €100 per ton of CO2.
  - Covered firms: 1,400 firms (≈68 percent of total corporate sector debt).
  - Firm behavioral assumption: pass-through to consumers allowed; price elasticity of -0.3 applied.
  - PD and bank capital impacts:
    - Direct bank CET1 depletion up to 3.5 percentage points (or 15 percent of existing CET1 capital) under the most severe instantaneous shock with no firm behavioral response.
    - Sectoral PD increases largest in agriculture, electricity and gas, and transportation over five years.
  - Market valuation losses on bank-held corporate debt securities from transition risk peak at 0.04 percent of total CET1 capital; bank holdings of corporate debt securities issued by carbon-intensive industries ≈0.5 percent of total holdings.
  - Policy implication: phased implementation and rapid green transition reduce PD increases; need to monitor bank carbon exposures.
- Physical risk (severe flooding simulation):
  - Upper-bound shock: 5 percent loss of capital stock over 5 years translates into GDP shock 0.4–4 percent.
  - Retail banks: simulated severe flooding leads to CET1 depletion around 240 bps at trough before recovery.
  - Banks’ exposure: roughly 15 percent of NFC credit to high-carbon sectors; more than 20 percent of loans to sectors exposed to high physical hazards (largely floods).

### Insurance sector — solvency, liquidity and climate analyses
- Sample and coverage:
  - Top-down tests cover 25 (re)insurers: 10 life, 8 non-life, 7 reinsurers; aggregated assets €369 billion; primary insurers €270 billion.
  - Coverage: >70 percent market share in each sub-sector.
- Solvency stress test (Solvency II based, instantaneous shocks):
  - Median valuation asset declines:
    - Median life insurer: asset values decline by 10 percent.
    - Median non-life insurer: asset values decline by 2 percent.
  - Median asset-liability ratio changes:
    - Median life insurer: declines to 103.2 percent (down 0.5 percentage points).
    - Median non-life insurer: declines to 113.3 percent (down 3.2 percentage points).
    - Median reinsurer: declines to 123.9 percent (down 2.6 percentage points).
  - Solvency outcomes:
    - Vast majority remain well-capitalized.
    - One insurer would not meet the SCR following the stress without reactive actions; capital shortfall < €10 million.
    - Median SCR coverage decreases by 13 percentage points for median non-life insurer.
    - Median post-stress SCR ratio in life sector is 6 percentage points higher than pre-stress on median, though majority life insurers see SCR ratio declines.
- Liquidity and variation margin analysis:
  - Bottom-up analysis of ten insurers: no sector-wide liquidity vulnerability; aggregate share of liquid assets relatively high.
  - Top-down margin-call analysis for interest rate swaps (sample of five insurers): all five could meet variation margin calls even for +100 bps interest rate change using cash only.
  - Data limitations: incomplete reporting for other derivative types; recommendation to cross-check QRTs with EMIR.
- Insurance climate risk:
  - Physical catastrophe simulations (e.g., Ireland/U.K. windstorm):
    - Gross claims around €550 million; net claims after reinsurance around €140 million.
    - Average SCR ratios decline by less than 3 percentage points.
  - Transition Minsky-style re-pricing (instantaneous NGFS 1.5°C priced to 2050):
    - Median valuation impacts: equity -3.8 percent; corporate bonds -1.6 percent; investment funds -1.4 percent.
    - Median life insurer asset value reduction 2.7 percent; non-life and reinsurers <1 percent.
    - Absolute potential asset decline ≈ €7 billion (2.3 percent of total investment assets).
  - Observations: transition risk larger for life insurers due to asset allocation and unit-linked business; results carry considerable modeling uncertainty.

### Investment funds — liquidity resilience and vulnerabilities
- Objective: assess fixed-income funds’ ability to withstand severe redemption shocks without LMTs or fire-sale price impacts.
- Data and sample:
  - Morningstar coverage: 82 percent of CBI-provided funds; after filters 3,289 funds (≈three quarters of Irish AUM); stress-test sample focused on 274 funds (fixed-income/property focus).
  - Fund categories: HY, IG, EM, Sovereign, Mixed, Property, Other.
- Methodology:
  - Three-stage model: calibrate redemption shock (VaR and ES at percentiles), compute HQLA using Basel III-like weights, compute Redemption Coverage Ratio (RCR = HQLA / redemption shock).
  - Homogeneity vs heterogeneity approaches for shock calibration.
- Key findings:
  - Most fixed-income funds would weather severe shocks across scenarios.
  - Pockets of vulnerability:
    - High-yield (HY) bond funds: majority vulnerable.
      - Under ES(3%) benchmark: Most HY funds (81 percent heterogeneity; 100 percent homogeneity) have RCR <1.
      - Under stricter RCR<0.5: close to 60 percent (heterogeneity) and 80 percent (homogeneity) of HY funds have RCR <0.5.
      - Some HY funds show liquidity shortfalls around 30 percent of TNA.
    - Emerging-market focused fixed-income funds: vulnerable in some scenarios; small fraction by count but can hold ≈20 percent of TNA in that category.
  - Sovereign and IG funds: most resilient; EM and Mixed moderately resilient.
- Policy implications:
  - CBI should complete internal stress-testing framework for funds.
  - CBI and CSO should analyze common asset holdings of funds, banks, and relevant non-banks.
  - Review use of liquidity management tools by funds, focusing on those with demonstrated liquidity challenges.

### Key actionable recommendations (excerpted)
- Banking supervision and liquidity:
  - Continue closely monitoring portfolios that benefited from pandemic payment breaks and other forbearance measures as pandemic support phases out. (Addressee: CBI; Timing: C; Priority: H)
  - Regularly perform liquidity stress tests by individual bank and significant foreign currencies. (Addressee: CBI; Timing: C; Priority: H)
  - Regularly monitor large off-balance sheet exposures (credit and liquidity facilities) and provide early warnings to banks. (Addressee: CBI; Timing: C; Priority: H)
- Interconnectedness and NBFI data:
  - Expand entity-level data collection on interbank and bank–NBFI exposures to complete network analysis. (Addressee: CBI; Timing: MT; Priority: H)
  - Analyze riskiness of bank–NBFI linkages and their direct/indirect impacts, focusing on international banks. (Addressee: CBI; Timing: ST; Priority: H)
- Climate risk monitoring and modelling:
  - Initiate data collection on banking exposures to transition and physical climate risks (bank-specific carbon exposures, geospatial flood data). (Addressee: CBI; Timing: MT; Priority: M)
  - Develop models to assess transition and physical climate risks’ impact on banking sector. (Addressee: CBI; Timing: MT; Priority: M)
- Insurance supervision:
  - Continue strengthening supervision of intra-group transactions and concentrations, focusing on post-Brexit group structures, recovery planning, and liquidity risk management. (Addressee: CBI; Timing: ST; Priority: H)
  - Conduct regular top-down solvency stress tests of insurers to validate insurers’ scenarios and recovery plans. (Addressee: CBI; Timing: ST; Priority: M)
  - Monitor protection gaps at regional and local level, especially flood risks. (Addressees: DoF, CBI; Timing: MT; Priority: M)
- Investment funds:
  - Complete internal stress-testing framework for IFs and MMFs. (Addressee: CBI; Timing: MT; Priority: H)
  - Determine common asset holdings of funds, banks, and relevant non-banks. (Addressees: CBI, CSO; Timing: MT; Priority: M)
  - Review use of liquidity management tools by IFs, focusing on funds with recent liquidity challenges. (Addressee: CBI; Timing: ST; Priority: H)

*Source: EXECUTIVE SUMMARY (1irlea2022013).*

### EXECUTIVE SUMMARY _________________________________________________________________________ 11

### EXECUTIVE SUMMARY

### Scope and structure
- Document components and major sections with starting page references as presented:
  - INTRODUCTION — 18
  - TOP-DOWN SOLVENCY STRESS TEST OF BANKS — 31
    - A. Solvency Stress Tests of the Banking Sector — 31
    - B. Macroeconomic Scenarios — 31
    - C. Banking Sector Vulnerabilities — 34
    - D. Methodology for Scenario-Based Solvency Stress Test — 41
    - E. Sensitivity Analysis of Retail and Large International Banks — 55
    - F. Sensitivity Analysis for Other International Banks — 57
  - BANK LIQUIDITY RISK ANALYSIS — 60
    - A. Introduction — 60
    - B. Liquid Assets and Funding Structure — 61
    - C. LCR-based Liquidity Stress Test — 64
    - D. Cashflow-based Liquidity Stress Test — 69
    - E. NSFR-based Liquidity Stress Test — 75
  - BANK INTERCONNECTEDNESS ANALYSIS — 76
    - A. Interbank and Intersectoral Network — 76
    - B. Domestic Interbank Contagion — 78
    - C. Bank – NBFI Interconnectedness — 80
    - D. Cross-border Interbank Contagion — 82
  - BANK CLIMATE STRESS TEST ANALYSIS — 86
    - A. Ireland Exposure to Climate Risks — 86
    - B. Transition Risk Sensitivity-based Stress Test — 87
    - C. Physical Risk Scenario-based Stress Test — 96
  - INSURANCE SECTOR SOLVENCY STRESS TEST — 99
    - A. Scope and Sample of the Solvency Stress Test — 100
    - B. Scenarios for the Solvency Stress Test — 102
    - C. Capital Standard and Modeling Assumptions — 104
    - D. Results of the Solvency Stress Test — 105
    - E. Sensitivity Analyses — 108
  - INSURANCE SECTOR LIQUIDITY RISK ANALYSIS — 110
    - A. Results of the Central Bank’s Liquidity Risk Analysis — 110
    - B. Analysis of Risks from Variation Margin Calls — 111
  - INSURANCE SECTOR CLIMATE RISK ANALYSIS — 113
    - A. Physical Risk — 113
    - B. Transition Risk — 117
  - INVESTMENT FUND LIQUITY STRESS TEST ANALYSIS — 119
    - A. Introduction — 119
    - B. Potential Spillovers from the Funds Sector — 121
    - C. Liquidity-Stress Testing Model — 121
  - INVESTMENT FUND STRESS TEST DESCRIPTIVE ANALYSIS — 123
    - A. Data and Sample Selection — 123
    - B. Portfolio Composition — 125
    - C. Redemption Shock — 127
  - INVESTMENT FUND STRESS TEST EMPIRICAL RESULTS — 128
    - A. Liquidity Resilience of Funds — 128
    - B. Liquidity Shortfall — 130
    - C. Conclusions of the Investment Fund Liquidity Stress Test Analysis — 131

### Key inventories (figures and tables included in the content)
- Selected Figures (with figure numbers and titles as presented)
  - 1. Financial Sector Overview
  - 2. Bank Liquidity and Capital Position
  - 3. COVID-19 Pandemic-Related Payment Relief
  - 4. Bank Profitability
  - 5. Credit Quality of Lending Portfolios
  - 6. Credit Quality of the CRE and SME Segments
  - 7. Bank Funding Conditions
  - 8. Bank Exposure to Climate Risks
  - 9. Stress Testing Framework
  - 10. ST Macroeconomic Scenarios
  - 11. Sample Bank Balance Sheet Decomposition
  - 12. Bank Credit Exposure and Quality by Geography and Segment
  - 13. Loans by Stages
  - 14. Loan Forbearance
  - 15. Liability Composition
  - 16. Sample Banks Securities Holdings
  - 17. Structure of Solvency Stress Testing Framework
  - 18. PD Projections
  - 19. Use of PiT and TTC PDs for RWAs and Provisioning
  - 20. Interest Rate Estimation by Portfolio
  - 21. Take-up of Bank-facing COVID-19 Support Measures
  - 22. Concentration Analysis
  - 24. Methodology of Sensitivity Analysis for Other International Banks
  - 25. Result of Sensitivity-Based Stress Test for Other International Banks
  - 27. Bank Contractual Cashflows and Funding Profile
  - 28. Liquidity Indicators
  - 29. LCR Component
  - 30. Results of the LCR-Based Stress Test
  - 31. Counterbalancing Capacity
  - 32. Asset Encumbrance
  - 33. Results of the Cashflow-Based Stress Test
  - 34. NSFR Development
  - 35. Methodology – NSFR Stress Test
  - 36. Results of the NSFR Stress Test
  - 37. Interbank and Bank-NBFI Network
  - 38. Total Assets of Financial Vehicle
  - 39. Interconnectedness Analysis
  - 40. Cross-Border Interconnectedness
  - 41. Result for Cross-Border Interbank Contagion Analysis
  - 42. Methodology – Transition Risks
  - 44. Bank Capital Impact under Carbon Tax Shock
  - 45. Bank Exposure to Transition Risk
  - 46. Shock Calibration – Physical Risks
  - 47. Physical Risk - Macroeconomic Scenario
  - 48. Result of the Physical Risk Analysis
  - 49. Components of Insurance Risk Analysis
  - 50. Own Funds and Long-Term Guarantee Measures
  - 51. Insurance Solvency ST—Valuation Impact
  - 52. Insurance Solvency ST—Solvency Impact
  - 53. Insurance Solvency ST—Sensitivity Analyses
  - 54. Insurance Liquidity Risks
  - 55. Insurers’ Variation Margin Calls
  - 56. Insurers’ Physical Climate Risks—Natural Catastrophes
  - 57. Insurers’ Physical Climate Risks—Parametric Approach
  - 58. Insurers’ Sectoral Exposures
  - 59. Insurers’ Transition Risks
  - 60. Portfolio Composition of Irish-Domiciled Fixed-Income Funds
  - 61. HQLA Buffers - Irish-domiciled Fixed-Income Funds
  - 62. RCR and the Resilience of Funds
  - 63. Liquidity Shortfall and RCR

- Selected Tables (with table numbers and titles as presented)
  - 1. Recommendations
  - 2. The Banking System
  - 3. PD Portfolio Mapping
  - 4. Interest Rate Portfolio Mapping
  - 5. Result of Scenario-Based Stress Test
  - 6. LCR Stress Test Parameters
  - 7. Cash Flow Based Stress Test Parameters
  - 8. Estimation of Expected Default Frequency for Irish Firms
  - 9. Projection of Corporate Balance Sheet and Vulnerability Indicators
  - 10. Insurance Stress Test Specification
  - 11. Investment Fund Data Availability in Morningstar
  - 12. Sample of Investment Funds in the Stress Test
  - 13. Liquidity Weights
  - 14. Average Redemption Shocks for the ES and VaR Models

### Appendices and technical annexes
- Appendices listed with starting pages:
  - I. Risk Assessment Matrix — 133
  - II. Housing Price Developments — 135
  - III. The Estimation of the PD Satellite Models — 137
  - IV. The Estimation of the Interest Rate Satellite Models — 140
  - V. Shock Calibration of the Physical Risk Analysis of the Banking — 142
  - VI. Banking Sector Stress Testing Matrix (STeM) — 144
  - VII. Insurance Sector Stress Testing Matrix (STeM) — 151
  - VIII. Investment Fund Liquidity Stress Testing Matrix (STeM) — 154
  - IX. Insurance Sector Interest Rate Scenarios — 155
  - X. Investment Fund Category Correspondence Table — 156
  - XI. RCR for a Range of Shock Scenarios — 158
  - XII. RCR and Liquidity Shortfall under the Heterogeneity Approach — 160
  - References — 161

### Glossary (selected acronyms as provided)
- AC — Amortized Cost
- AE — Asset Encumbrance
- ASF — Available Stable Funding
- AUM — Assets Under Management
- BIS — Bank for International Settlements
- BMA — Bayesian Model Averaging
- BPS — Basis Points
- BSCR — Basic Solvency Capital Requirement
- CAR — Capital adequacy ratio
- CBI — Central Bank of Ireland
- CCB — Capital Conservation Buffer
- CCyB — Countercyclical Buffer
- CET1 — Core Equity Tier 1
- CFLST — Cash Flow-based Liquidity Stress Test
- COREP — Common Reporting Framework
- CRE — Commercial Real Estate
- CRR — Capital Requirements Regulation (EU)
- CSO — Central Statistics Office
- EA — Euro Area
- EAD — Exposure at default
- EBA — European Banking Authority
- ECB — European Central Bank
- EDF — Expected Default Frequency
- EIOPA — European Insurance and Occupational Pensions Authority
- EMIR — European Markets and Infrastructure Regulation
- EOF — Eligible Own Funds
- EU — European Union
- FINREP — Financial Reporting Framework
- FSAP — Financial Sector Assessment Program
- FSB — Financial Stability Board
- FV — Fair Value
- FVC — Financial Vehicle Corporations
- FVOCI — Fair Value through Other Comprehensive Income
- FVTPL — Fair Value through Profit and Loss
- FX — Foreign Exchange
- GAAP — Generally Accepted Accounting Principles
- GDP — Gross domestic product
- GFC — Global Financial Crisis
- GFM — Global Macro-financial Model
- HQLA — High-quality liquid assets
- IF — Investment Funds
- IIFA — International Investment Fund Association
- IMF — International Monetary Fund
- IRB — Internal ratings-based (approach)
- IRRBB — Interest Rate Risk in the Banking Book
- LCR — Liquidity Coverage Ratio
- LGD — Loss Given default
- LSI — Less Significant Institution
- LTG — Long-Term Guarantee
- MBF — Market-based Finance
- MFI — Monetary Financial Institutions
- MCR — Minimum Capital Requirement
- MMF — Money Market Funds
- NBFI — Non-Bank Financial Institution
- NFC — Non-Financial Corporate
- NGFS — Network for Greening the Financial System
- NII — Net Interest Income
- NIM — Net Interest Margin
- NPL — Nonperforming Loan
- NSFR — Net-Stable Funding Ratio
- OFI — Other Financial Institution
- ORSA — Own Risk and Solvency Assessment
- OSII — Other Systemically Important Institution
- PEPP — Pandemic Emergency Purchase Program
- PD — Probability of Default
- PiT — Point-in-Time
- QRT — Quantitative Reporting Template
- RAM — Risk Assessment Matrix
- REIT — Real Estate Investment Trust
- RFR — Risk-Free Rate
- ROW — Rest of the World
- RWA — Risk-Weighted Assets
- RWD — Risk Weight Density
- SCR — Solvency Capital Requirement
- SI — Significant Institution
- SMEs — Small- and Medium-Sized Enterprises
- SPE — Special Purpose Entity
- SPV — Special Purpose Vehicle
- ST — Stress Test
- STE — Short-Term Exercise
- STeM — Stress Testing Matrix
- TN — Technical Note
- TTC — Through-The-Cycle
- UCITS — Undertaking for the Collective Investment in Transferable Securities
- U.K. — United Kingdom
- U.S. — United States
- VA — Volatility Adjustment
- WEO — World Economic Outlook

*Source: EXECUTIVE SUMMARY (document listing and section headings as provided).*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### Background and context
- Irish financial landscape: increasing divergence between an innovative and fast-growing international finance sector and a consolidating retail banking sector facing post-GFC operating restrictions and rising competition from non-bank players.
- Shocks and risks: global pandemic and Brexit left uneven marks; risks from unwinding of public support; ongoing and emerging risks include persistent inflationary pressures (supply bottlenecks) and the war in Ukraine that may impede recovery and magnify vulnerabilities to downside shocks.

### Banking sector — resilience and remaining vulnerabilities
- Banks maintained strong capital and liquidity buffers through the pandemic, but pockets of vulnerability remain:
  - Profitability challenges and long-term mortgage arrears in retail banks.
  - High exposure to the CRE segment.
  - Significant off-balance sheet exposures of large international banks may create domestic and cross-border challenges.
- Stress-testing framework:
  - Scenario-based stress tests for retail and large international banks; streamlined sensitivity tests for other international banks (all LSIs).
  - Liquidity stress tests and contagion analysis covering domestic and cross-border interbank exposures and cross-sectoral interlinkages between banks and non-bank financial institutions.
  - Bank risk analysis included climate change-related risks (transition and physical).

### Bank solvency stress test results (scenario-based)
- Baseline scenario (capital accumulation):
  - Retail banks: fully loaded CET1 ratios trend from 16. 4 percent to 17.8 percent.
  - Large international banks: fully loaded CET1 ratios trend from 19.9 percent to 27.9 percent.
  - No banks fell below the hurdle rates under the baseline.
- Adverse scenario (capital depletion):
  - Aggregate fully loaded CET1 ratio declines by about 6.7 percentage points for retail banks and 0.4 percentage points for large international banks by the 5th year.
  - Trough capital depletion: 7.2 percentage points for retail banks and 2.3 percentage points for large international banks.
  - Credit risk provisioning is the largest contributor to capital decline: cumulative effect of 7.4 percentage points over five years.
- Hurdle rate thresholds considered:
  - CET1: 4.5 percent plus O-SII buffer.
  - Tier 1: 6.0 percent plus O-SII buffers.
  - CAR: 8 percent plus O-SII buffer.
  - Capital conservation buffer allowed to be used under the adverse scenario.

### Sensitivity analysis: unwinding of pandemic support
- Assumption: 50 percent of loans with either active or expired moratoria deteriorate from stage one and two into stage three assets.
  - Result: one retail bank’s CET1 and Tier 1 ratio fall below the hurdle rates under the adverse scenario.
  - Capital shortfall against CET1 hurdle rate amounts to 0.2 percent of GDP.

### Liquidity resilience and gaps
- LCR-based stress test:
  - Meaningful decline in LCR ratios across stress scenarios.
  - Retail banks: larger impact under retail scenario.
  - International banks: higher stress under the wholesale scenario.
  - All banks can withstand the most severe 30-day shock due to high initial liquidity buffers.
- Cashflow-based stress test (beyond 30 days):
  - Banks broadly resilient in short-term due to counterbalancing capacities.
  - Liquidity position weakens beyond three months due to maturity mismatch (frontloaded outflows, backloaded inflows).
  - Large international banks more prone to short-term liquidity shortfalls due to high share of wholesale funding and larger off-balance sheet exposures.
    - About 40 percent of the wholesale funding for large international banks are intra-group funding.
    - Off-balance sheet contingent liabilities amount to €108 billion or 40 percent of total contractual outflows within 12 months for large international banks.
    - One component, issuance of Insurance Letters of Credit (“ILOCs”), is approximately 80 percent collateralized on average.
- Currency-specific vulnerabilities:
  - LCR and cashflow exercises for major foreign currencies reveal vulnerabilities to U.S. dollar and U.K. Sterling denominated outflows, particularly for international banks in U.S. dollars.
  - Contributing factors: weaker initial positions in those currencies, non-trivial off-balance sheet exposures (mainly credit lines and FX-related derivative transactions), and reliance of international subsidiaries on foreign currency backstops from parents.
  - EBA reporting differences: inflows/outflows in the LCR of significant foreign currencies include principal exchange on FX derivative contracts, whereas aggregate LCR reports FX derivative flows on a net basis.

### Interconnectedness and contagion
- Interbank and bank-NBFI linkages:
  - Interbank linkages appear muted domestically and cross-border.
  - High degree of connectivity between large international banks and NBFIs: NBFIs account for 45 percent of the total exposures of the large international banks.
  - Adding top NBFIs by exposure size into the network increases inward spillover risks of large international banks, indicating larger vulnerability to shocks from NBFIs.
- Policy implication: need for expanded entity-level data collection and analysis of bank–NBFI linkages.

### Banking sector climate risk analysis
- Transition risk (carbon-tax shock single-factor analysis):
  - Meaningful impact on Irish corporates; energy-intensive sectors experience largest PD increases.
  - Direct impact on banks: notable CET1 capital depletion up to 3.5 percentage points (or 15 percent of existing CET1 capital) under the most severe scenario.
  - Non-linear effects over the risk horizon; policy implication: speedy transition to greener technology to ensure business viability for carbon-intensive sectors.
- Physical risk (severe flooding simulation):
  - Simulated severe flooding leads to non-trivial bank CET1 depletion around 2.4 percentage points at trough, before recovery close to pre-shock levels.
  - Suggests need for enhanced monitoring of exposures to climate-sensitive sectors and precautionary actions (e.g., high quality collateral, insurance against physical damage).

### Insurance sector — stress test analysis and resilience
- Scope: top-down approach covering 25 (re)insurers (>70 percent of market in each sub-sector: life, non-life, reinsurance).
- Solvency results:
  - Vast majority of insurers remain well-capitalized.
  - One insurer would not meet the solvency capital requirement (SCR) following the stress without reactive management actions; capital shortfall would be less than €10 million.
  - Non-life insurers most affected; median SCR coverage decreases by 13 percentage points for the median non-life insurer.
  - Median post-stress SCR ratio in the life sector is 6 percentage points higher than pre-stress, though the SCR ratio declines for the majority of life insurers.
- Key drivers of asset value declines: higher corporate and sovereign spreads; equity and property shocks smaller. Higher interest rates and Euro depreciation partly offset other adverse shocks for many insurers with a slight excess of foreign-denominated assets over liabilities.
- Liquidity analysis:
  - Bottom-up analysis of ten insurers indicates no sector-wide liquidity vulnerability; aggregate share of liquid assets is relatively high and inflows may increase in the adverse scenario for some firms due to derivative-related collateral flows.
  - Top-down analysis of margin-call risks for interest rate swaps indicates low vulnerabilities; reporting data for other derivative types is incomplete.
  - Importance of supervisor understanding liquidity flows, liquidity risk management practices, and insurance group structures.
- Insurance exposure to climate risks:
  - Exposure mainly through non-life underwriting.
  - Large natural catastrophes, when isolated, would likely not have a pronounced impact on solvency levels.
  - Domestic important perils: windstorms, floods, freezes; many (re)insurers underwrite globally (e.g., U.S. hurricanes, wildfires).
  - Major windstorm in Ireland and the United Kingdom would affect profits and reduce SCR ratios by less than 3 percentage points on average.
  - Modeling challenges persist; heavy reliance on intra-group reinsurance.

### Insurance sector transition risk
- Assessed via top-down application of NGFS orderly 1.5 degree scenario priced instantaneously by investors to 2050.
- Transition risk manageable overall but larger for life insurers due to comparatively riskier asset allocation (importance of unit-linked business and type of funds chosen by customers).
- In absolute amounts, asset values could decline by around €7 billion (or 2.3 percent of total investment assets).
- Results carry considerable modelling uncertainties.

### Investment funds — liquidity resilience and vulnerabilities
- Focus: liquidity resilience of fixed-income investment funds facing severe but plausible redemption shocks without use of liquidity management tools or sale of less liquid assets.
- Most fixed-income funds would weather severe shocks across a wide range of scenarios.
- Pockets of vulnerability: high-yield bond funds and emerging-market focused fixed-income funds are more susceptible to liquidity mismatch and may be less resilient under severe market stress.
- Progress on recommendations from 2016 FSAP:
  - Central bank has made significant strides in building a stress-testing framework for funds, but the model remains a work-in-progress.
  - Given the size, accelerating growth, and systemic importance of the investment fund sector, the central bank should reinforce efforts to complete the model and simulations and to conduct regular stress tests of Irish-domiciled funds.

### Key actionable recommendations (excerpted from Table 1 — Banking risk analysis)
- Continue to closely monitor portfolios that benefited from pandemic payment breaks and broader policy support, including other forms of forbearance measures to ensure banking sector resilience with the phase-out of pandemic supportive policies. (Addressee: CBI; Timing*: C; Priority**: H) (¶3, ¶48)
- Continue to regularly perform liquidity stress tests by individual bank and significant foreign currencies to identify and provide early warning signals to banks with liquidity gaps over short and long-term horizons. (Addressee: CBI; Timing*: C; Priority**: H) (¶66, ¶76)
- Continue to regularly monitor large off-balance sheet exposures (via credit and liquidity facilities). Provide early warning signals to banks with associated liquidity risks. (Addressee: CBI; Timing*: C; Priority**: H) (¶58, ¶75)
- Expand data collection on interbank and bank-NBFI exposures at entity level (in addition to existing large exposure dataset) to complete interbank and bank-NBFI network analysis. (Addressee: CBI; Timing*: MT; Priority**: H) (¶85, ¶93)
- Analyze the riskiness of banks-NBFI linkages and their direct and indirect impact on the banking sector, with a particular focus on international banks. (Addressee: CBI; Timing*: ST; Priority**: H) (¶81, ¶94)
- Initiate data collection on banking exposures to both transition and physical climate risks, such as bank specific carbon exposures as well as leveraging geospatial data to assess exposures to corporate and mortgage borrowers facing high flooding risks, to allow in-depth analysis on climate related risks to financial stability. (Addressee: CBI; Timing*: MT; Priority**: M) (¶110, ¶111, ¶112, ¶120)
- Develop models to properly assess the impact of transition and physical climate risks on the banking sector. (Addressee: CBI; Timing*: MT; Priority**: M) (¶110, ¶111, ¶112, ¶120)

*Italicized source attribution line.*

*Source: EXECUTIVE SUMMARY (1irlea2022013).*

### 8. Continue strengthening the supervision of intra-group

### 8. Continue strengthening the supervision of intra-group transactions and concentrations, with a focus on post-Brexit group structures, recovery planning, and liquidity risk management. (¶142, ¶152)

### Recommendations and supervisory actions
- Continue strengthening supervision of intra-group transactions and concentrations, with specific focus on:
  - post-Brexit group structures,
  - recovery planning, and
  - liquidity risk management. (¶142, ¶152)
  - Responsible authority: CBI. Timing: ST. Priority: H. (Insurance – Risk Analysis)
- Conduct regular top-down solvency stress tests of insurers, also to validate scenarios and assumptions used by insurers in their recovery plans. (¶138)
  - Responsible authority: CBI. Timing: ST. Priority: M.
- Monitor protection gaps at regional and local level, including price dynamics and the cost of insurance, and consider options to narrow protection gaps wherever material, particularly with respect to flood risks. (¶155)
  - Responsible authorities: DoF, CBI. Timing: MT. Priority: M.

### Investment fund sector risk recommendations
- Complete the internal stress-testing framework for investment funds and money market funds. (¶191)
  - Responsible authority: CBI. Timing: MT. Priority: H.
- Determine the common asset holdings of investment funds, banks, and relevant non-banks. (¶192)
  - Responsible authorities: CBI, CSO. Timing: MT. Priority: M.
- As part of ongoing policy development on IF liquidity risk management and taking into account developments at the EU and international level, review the use of liquidity management tools by IFs, with a particular focus on funds that have demonstrated liquidity challenges in recent periods of stress. (¶193)
  - Responsible authority: CBI. Timing: ST. Priority: H.

### Banking sector structure and recent evolution (selected facts)
- Banking sector assets contracted from 780 percent in 2009 to about 200 percent of GDP in 2021.
- Funds sector total assets grew from 4.7 times GDP in 2009 to more than ten times in 2021.
- Retail banks: total assets of €317 billion at Q2 2021; account for more than 90 percent of lending to households; hold 40 percent of total bank assets; loans account for 60 percent and debt securities account for 13 percent of their total assets. Four retail banks are SIs; one is an LSI.
- Large international banks: total assets €276 billion; account for 70 percent of the banking system’s total non-resident deposits and about 70 percent of lending to non-residents; three entities (two US-based subsidiaries, one U.K.-based subsidiary); all three considered SIs by the ECB.
- Other international banks: total assets €62.5 billion; nine foreign-owned LSIs.

### Resilience during the pandemic (selected indicators)
- Aggregate CET1 ratio: 22 percent at beginning of 2020 (EA average at 15 percent).
- CET1 ratio sustained at around 22 percent through the pandemic to Q3-2021.
- Liquidity Coverage Ratio: 178 percent as of Q3-2021 (comfortably above the 100-threshold).
- Payment relief measures have been largely phased out later 2021; full impact on asset quality and financial stability has yet to be seen.

### Key vulnerabilities and structural risks
- Subdued profitability and high-cost base:
  - Net profitability drivers include low credit demand, constrained net interest margin, rising share of fixed rate new mortgages, and high operational costs.
- Problem assets and legacy portfolios:
  - NPL ratio declined from a peak at 25 percent in 2014 to around 2.6 percent at mid-2021.
  - Retail banks’ NPL ratio: 4.4 percent as of Q2 2021.
  - Over 50 percent of total NPLs are in arrears for more than 2 years (mostly household mortgage segment).
  - Irish mortgage portfolios have persistently higher risk weight densities than other Euro Area countries.
- Credit exposure to CRE and SMEs:
  - Banks’ exposure to real estate activities: around €18 billion total and €11 billion domestic CRE market.
  - CRE exposures equal to 20 percent of total corporate loans and 30 percent of domestic corporate loans.
  - CRE NPL ratio: around 12.2 percent.
  - Mortgage segment constitutes about 40 percent of total bank lending; residential housing price growth reached 15 percent in March 2022.
  - SME lending constitutes about 35 percent of total corporate loans; SME NPL ratio: 8.6 percent as of March 2021; large corporates NPL ratio: 6.5 percent.
- Coverage ratio:
  - NPL coverage ratio of Irish banks: around 34 percent; Euro Area average: 47 percent.
- Funding and liquidity risks:
  - International banks rely heavily on wholesale and non-resident funding; loan-to-deposit ratios trending downwards but liquidity buffers are high.
  - Off-balance sheet exposures: €108 billion (40 percent of total assets) for large international banks as of mid-2021; retail banks around €30 billion (10 percent of total assets).
  - High shares of wholesale funding and non-resident deposits may pose funding stability risks in event of adverse confidence shocks.

### Sample bank stress-test coverage (selected figures from Table 2)
- Total assets of sample banks: 621,745 (Mil. EUR) — 100.0 percent of sample; total asset of the banking system: 782,083 (Mil. EUR).
- Individual sample asset figures (Mil. EUR) and asset share of banking system:
  - Bank 1 (Retail, SI): 122,867 — 15.7 percent
  - Bank 2 (Retail, SI): 129,095 — 16.5 percent
  - Bank 3 (Retail, LSI): 21,503 — 2.7 percent
  - Bank 4 (Retail, SI): 30,634 — 3.9 percent
  - Bank 5 (Retail, SI): 12,272 — 1.6 percent
  - Bank 6 (Large International, SI): 69,489 — 8.9 percent
  - Bank 7 (Large International, SI): 141,945 — 18.1 percent
  - Bank 8 (Large International, SI): 64,889 — 8.3 percent
  - Bank 9 (Other International, LSI): 9,236 — 1.2 percent
  - Bank 10 (Other International, LSI): 5,093 — 0.7 percent
  - Bank 11 (Other International, LSI): 3,219 — 0.4 percent
  - Bank 12 (Other International, LSI): 11,503 — 1.5 percent

*Source: IRELAND, INTERNATIONAL MONETARY FUND (content unit: 1irlea2022013).*

### 5. Bank exposure to nonbank financial institutions (NBFIs), mostly comprised of financial

### 5. Bank exposure to nonbank financial institutions (NBFIs), mostly comprised of financial

### NBFI exposures and channels of bank-NBFI linkage
- Bank exposure to nonbank financial institutions (NBFIs), largely financial vehicle corporations and special purpose vehicles (FVCs and SPVs), is concentrated in international banks with a cross-border focus.
- The majority of the FVCs are domiciled in Ireland and Italy, with over 400 billion assets in each jurisdiction.
- Banks use securitization by FVCs to offload loans and associated credit risks to investors in FVC securities, but recently have begun to:
  - buy back a portion of the FVC securities, or
  - provide direct sponsorship to the FVCs,
  thereby retaining a portion of the risk associated with the securitized loans.
- Retained asset-backed securities could be used as eligible instruments to access various types of central bank liquidity facilities.
- SPVs are used as investment vehicles (private equity, syndicated loans, distressed debt).
- International banks have substantial exposures to FVCs and SPVs via derivative transactions and settlements, and various short- and long-term asset receivables that carry counterparty risks.
- Other linkages between banks and NBFIs:
  - Banks’ investment in property funds: Around €6.2bn (53 per cent of total) financing of property funds comes from banks, of which €3.2bn comes from Irish retail banks.
  - Banks’ wholesale deposits placed by NBFIs for transactional purposes are an important source of funding for the Irish banks (around 9 percent of total banking sector liabilities).

### Bank–sovereign nexus
- Domestic and foreign sovereign debt securities holdings account for about 2.4 and 4.2 percent of total banking sector assets as of mid-2021.
- These holdings are lower than the average holdings by other Euro Area countries.
- Retail banks hold the majority of domestic sovereign bonds.
- The relatively longer duration sovereign bonds, currently at around 9 years, may expose banks to adverse market valuation amid sudden decompression of sovereign risk premium and tightened global financial conditions.
- State ownership of three large retail banks, which hold more than one-third of system assets, may form a two-way channel enabling spillover of sovereign and banking stress.
  - Current state holding: AIB 70.97%; BOI 4.93%; PTSB 75% (Mar 2022).

### Climate-related exposures
- The banking system’s exposure to climate risks is material:
  - Roughly 15 percent of bank NFC credit is to sectors with a high carbon footprint.
  - More than 20 percent of loans are to sectors exposed to high physical hazards (largely floods).
  - These shares are well above the Euro area average, suggesting heightened vulnerability of banks to carbon tax shocks and sea level rises.
- The government adopted an ambitious climate action plan in 2021.

### Exposures to Russia and indirect channels
- Direct and indirect banking sector exposures to Russia are limited.
- Direct channels:
  - Irish banks’ lending to Russian companies predominantly comes from internationally focused banks and as of Q4 2021 stood at €1.1 billion, representing about 0.2 percent of total assets.
  - No Irish banks had loans secured against collateral located in Ukraine or Russia.
- Indirect channels:
  - Real sector: Ireland has limited trade exposure to Russia (around 0.4 percent of both goods exports and imports share) and Ukraine (0.1 percent).
  - Energy sector:
    - Ireland imports only 6.3 percent of its energy imports (coal, coke, and briquettes, petroleum fuels and gas) from Russia.
    - Breakdown of energy imports: 49 percent from the United Kingdom, 15 percent from the United States, 6 percent from Netherlands, 2.9 percent from Sweden.
    - Only 5 percent comes directly from Russia (context: the source of U.K. energy is diversified).
  - Household sector: The share of energy in final household expenditure is around 7 percent in Ireland, lower than the EU average.
- Conclusion: Both direct and indirect exposure of Irish banks to the Russia-Ukraine conflict is considered negligible at present. Further pressure from high global energy prices into domestic inflation and output has been incorporated into both the baseline and adverse scenario.

### Stress testing framework and macro-financial scenarios
- Objective: assess the capacity of the banking system to withstand severe but plausible macro-financial shocks and explore channels of propagation.
- The stress tests covered solvency, liquidity, contagion, and climate risks:
  - Solvency: top-down scenario-based stress tests on retail and large international banks; streamlined sensitivity tests for other international banks.
  - Liquidity: regulatory approaches including LCR and NSFR, and a cash-flow based approach using maturity ladders for large withdrawals across multiple time horizons.
  - Contagion: domestic and cross-border interbank exposures and cross-sectoral interlinkages between banks and NBFIs (FVCs, insurers, pension funds, investment funds).
  - Climate: transition risk via single factor sensitivity analysis to study impact on firms’ probabilities of default (PDs) following an increase in carbon taxation (sectorally differentiated); physical risk via a scenario simulating the macroeconomic impact of a severe flooding event translated into bank losses via satellite models.
- The stress test covered 12 banking institutions, constituting around 80 percent of total banking system assets.
  - For five retail and three large international banks, scenario-based stress tests were conducted; sensitivity analysis for the four other international banks.
  - Data used: confidential supervisory data as of mid-2021 at the highest consolidation level within Ireland.
- Macrofinancial risks identified for adverse scenario (joint realization):
  - Russia’s invasion of Ukraine leading to escalation of sanctions and higher commodity prices and tighter financial conditions.
  - Outbreaks of lethal and highly contagious COVID-19 variants with extended supply chain disruptions.
  - De-anchoring of inflation expectations in the United States and/or advanced European economies prompting faster-than-anticipated tightening of monetary policy.
  - Geopolitical tensions and deglobalization causing trade disruption and lower investor confidence.
  - Continued trade frictions and uncertainty related to post-Brexit arrangements increasing costs for Irish businesses.
- Adverse scenario features:
  - Significant contraction in growth, tightening global financial conditions, rising sovereign risk premia, and inflationary pressures.
  - The exercise uses data as of mid-2021 and features one baseline and one adverse scenario spanning a five-year horizon.
  - The adverse scenario results in shocks to GDP and GNI* growth equivalent to 2.6 and 3.1 standard deviations, respectively, from their baselines.
    - When measured against historical mean growth, the shock to GDP and GNI* is equivalent to 2.1 and 2.4 standard deviations, respectively.
  - Residential housing price: cumulative decline of about 30 percent over a three-year horizon in the adverse scenario.
  - CRE price: calibrated to decline by about 40 percent over the same horizon.

### Banking sector vulnerabilities and asset composition
- Credit risk is the largest risk factor for the banking system:
  - RWAs of credit risk account for 85 percent of total RWAs in the sample banks.
- Use of IRB approach:
  - Retail banks use the IRB approach for most of their credit RWAs, at 60 percent.
  - For international banks the IRB share stands at 4 percent.
- Risk weight density (RWD):
  - Historically high in Ireland, at an average of 51 percent since 2014 relative to the regional average at 37 percent.
  - Drivers: asset composition, level of IRB rollout, legacy mortgage portfolios from the GFC.
  - The Covid-19 shock and accompanying public supports contributed to a notable decline in risk weighted densities, driven by strong growth of high-quality liquid assets funded by surging customer deposits.
- Asset composition (sample banks, 2021Q2):
  - Loans represent 51 percent of assets, followed by central bank reserves and non-interest earning assets.
  - By sector, loans are mostly concentrated in households and corporates, followed by financial institutions and central banks and governments.
  - Differences by bank type:
    - Retail banks: disproportionately large share of household (mostly mortgages) and corporate loans due to domestic focus.
    - International banks: higher share of non-interest bearing assets (reverse repo and derivative assets) and bulk of credit exposure to banks and NBFIs (cross-border exposures), resulting in a more diversified and liquid asset profile.
- Asset quality and segmentation:
  - Cross-border focus of international banks yields higher share of credit exposure to U.K., Euro Area, and U.S.
  - Retail banks are highly exposed to the Irish market with some U.K. exposures.
  - Credit quality varies across markets and segments:
    - Exposures in the Irish and U.K. markets underperform those in the rest of the world.
    - Corporate portfolios (particularly real estate and SME segments) bear higher credit risks than other loan products.
  - NPL ratios for NBFIs are lower than other segments but remain uncertain due to high-risk profile of FVCs and dense interconnections with large international banks.
  - A recent revision in definition of default and reassessment of unlikely-to-pay loans contributed to increases in NPLs across portfolios.
  - The recent declining trend of NPL ratios may reflect supportive policies mitigating full pass-through of the Covid-19 shock to loan loss provisions.

*Source: IMF staff, Ireland Financial Sector Assessment (chapter excerpts).*

### 17. The global pandemic, coincided with Brexit, has resulted in a significant increase in

### 17. The global pandemic, coincided with Brexit, has resulted in a significant increase in 

### Credit risk and non-performing loans (NPLs)
- Stage 2 assets on aggregate have doubled since the pandemic.
- Part of the increase in 2020 reverted to stage 1 in early 2021, although the overall level remains elevated.
- An assessment of the impact on bank solvency in the event of a large transition from stage 2 to stage 3 is warranted because of:
  - delayed recognition of NPLs;
  - the tapering of policy support; and
  - limited policy space to counteract additional downside shocks in light of heightened global uncertainties.
- Banks are proactively restructuring loans, which may avert short-term defaults but may also push full recognition of NPLs into the medium- and long-term.

### Forbearance and restructuring
- As of mid-2021:
  - 66 percent of the total non-performing loans has been re-negotiated and restructured with borrowers.
  - 65 percent for household NPLs.
  - 67 percent for corporate NPLs.
- Moratoria have expired for most Euro Area countries including Ireland, but various forbearance measures remain at banks’ discretion.
- Properly used forbearance can:
  - provide buffers to viable firms and households facing transitory liquidity difficulties;
  - relieve stress on banks by lowering probability of default and loan loss provisions.
- Risks if forbearance is misused:
  - delay in recognition of inevitable credit losses;
  - limitation of banks’ lending and income generation capacity;
  - potential escalation to systemic risks.
- Banks should make reasonable efforts to identify and distinguish viable borrowers from non-viable ones.

### Interest rate risk and repricing gaps
- As of September 2021:
  - 65 percent of newly issued loans were denominated in variable rates.
  - household mortgage new issuance with variable rates was 20 percent.
- Liability-side characteristics:
  - large share of overnight and term deposit liabilities subject to short-term repricing.
  - slower repricing on the asset side (especially new mortgage loans with high share of fixed rate issuance) and competition from nonbank lenders may enlarge the repricing gap.
- A larger repricing gap may more than offset the benefit of adjustable-rate lending schemes and increase interest rate risk.
- Note: higher share of floating rate loans for segments other than household mortgages may translate into heightened credit risks in a stressful event.

### Market risk and sovereign exposures
- Holdings of sovereign securities:
  - 43 billion euros in sovereign securities.
  - 7 percent of total banking assets.
  - 70 percent of total debt securities.
- Foreign sovereign securities:
  - on average 4.3 percent of total banking assets.
  - represented 43 percent of total debt securities held by the banks.
  - were held mostly by international banks.
- Irish sovereign securities:
  - ranked second and were held exclusively by retail banks.
- Accounting treatment and duration:
  - A significant portion of sovereign securities placed in fair value category (mostly through other comprehensive income).
  - Unhedged portion of the debt securities represents about 12.3 percent of the total marketable securities.
  - Weighted average duration for sovereign securities holdings of the retail banks in sample is 5 years.
  - Comparative durations: holdings of bonds issued by banks 3.4 years, corporates 4.7 years, nonbank financial institutions 5.8 years.
- Risks:
  - high concentration of sovereign debt combined with relatively longer duration may expose banks to adverse market conditions under imperfect hedging and may adversely affect banks’ solvency and liquidity.

### Scope and sample of the solvency stress test
- Sample and treatment:
  - Eight sampled banks were assessed.
  - For the five retail and three large international banks, a scenario-based stress test was conducted.
  - Sensitivity analysis was conducted for the four other international banks.
- Capital metrics used:
  - total capital adequacy ratio (CAR);
  - Tier 1 capital (T1);
  - Common Equity Tier 1 (CET1) capital;
  - leverage ratios.
- Capital requirements included:
  - O-SII buffer, where applicable;
  - baseline scenario also subject to bank specific Pillar II requirements.
- Banks are allowed to deplete their capital conservation buffers (CCBs) under stress.

### Solvency framework, risks covered, and modelling approach
- Risks covered:
  - credit risk (all exposures),
  - market risks,
  - sovereign risk,
  - interest rate risk in the banking book.
- Exclusions:
  - derivatives book was not considered due to lack of sufficiently granular information.
- Geographic segmentation applied to exposure classes: Ireland, United Kingdom, United States, rest of the Euro Area, and rest of the world.
- Macro scenarios translated into evolution of PDs, LGDs and interest rates using satellite models; these affect:
  - growth of balance sheet items,
  - pre-provision net income,
  - other base components.
- Shocked risk parameters drive RWAs, provisions (via IFRS9 transition matrices), asset repricing, and market valuation losses.
- Final output:
  - combined P&L items and full balance sheet evolution to obtain CET1, Tier 1, total capital, and leverage ratios over the stress-testing horizon.

### Balance sheet growth assumptions
- Quasi-static approach:
  - asset allocations and composition of funding assumed to remain the same;
  - balance sheet growth follows the weighted average GDP growth of countries where banks have significant exposure.
- Floor on rate of change of balance sheets set at zero percent to prevent deleveraging.
  - This constraint is binding in the adverse scenario.
- Other adjustments:
  - revaluation of assets and liabilities for foreign exchange movements;
  - conversion of a portion of off-balance sheet items (credit lines and guarantees) to on-balance-sheet exposures.
- The stress test assumes no write-off of existing NPLs over the risk horizon.

### IFRS9, mapping to regulatory classification, and PD/LGD methodology
- IFRS9 required forward-looking expected loss calculations and asset stage classifications using transition matrices and life-time expected loss.
- Accounting vs regulatory classification:
  - accounting classification based on balance sheet exposures, distinguishes loans and securities, is sector-based and less granular;
  - regulatory classification based on on- and off-balance sheet amounts and different borrower types.
- The FSAP team mapped accounting and regulatory portfolios using approximations assuming similar risk characteristics.
- PD estimation:
  - five portfolio types modeled separately: corporate CRE, corporate non-CRE, household retail, household mortgage, and financial institutions.
  - Bayesian Model Averaging (BMA) methodology used to estimate point-in-time PDs with macro variables as independent variables.
  - For household mortgage portfolios, U.K. mortgage exposures in the adverse PD shocks were reduced to 50 percent of the original shock to reflect lower historical default rates for U.K. mortgages.
- Sovereign PDs:
  - extracted from sovereign yields using a Merton-based reduced-form structural model.
  - LGD assumption used in sovereign PD extraction: 45 percent.
- Corporate exposures outside Ireland:
  - PDs proxied by Moody’s expected default frequency (EDF) for United Kingdom, United States, other Euro Area, and rest of the world.
- PD forecasting details:
  - a logit transformation was applied before BMA/OLS estimates to address the truncated nature of default rate distribution and ensure forecasts remain within 0-1 bounds.

### PD and LGD projections and stress impacts
- Baseline vs adverse:
  - In the baseline scenario, PDs in most segments projected to remain flat.
  - In the adverse scenario, PDs sharply increase, with larger impacts on:
    - household retail,
    - corporate (in particular CRE),
    - financial institutions (in particular NBFIs),
    - compared with household mortgages which are less impacted.
  - The magnitude of projected PD shocks under the adverse scenario is milder relative to historical stress episodes because of structural changes since the GFC.
- Relevant macro drivers for PDs:
  - real domestic output,
  - unemployment rate,
  - short-term and long-term interest rates,
  - asset prices.
- LGD projection:
  - point-in-time LGD paths under baseline and adverse scenarios projected using historical time series on coverage ratio for the total loan portfolio provided by the CBI.
  - stressed coverage ratio under adverse scenario applied to household and corporate portfolios separately, considering differences in collateralization.
  - forward paths attached to bank-specific Point-in-time LGDs with a floor in the adverse scenario at the downturn LGDs.
  - for banks under standardized approach without LGD statistics, LGDs for IRB banks provided by national authorities used as proxy.

### Interaction of credit risk with capital ratios
- Credit risk affects capital ratios through:
  - loss provisions (numerator) driven by IFRS9 provisioning, loan exposures, stage transitions (guided by stressed PiT PDs) and PiT LGDs under stress;
  - risk weights (denominator) via RWAs which differ by regulatory approaches (STA vs IRB).
- The stress-testing framework uses PiT PDs for provisioning and a mix of PiT and TTC PDs for RWAs per the regulatory approach.

*Source: IMF staff estimates and analysis as presented in the provided chapter.*

### 34. To compute capital requirements, standardized (STA) and internal ratings-based (IRB)

### 34. To compute capital requirements, standardized (STA) and internal ratings-based (IRB)

### Capital requirements: STA and IRB portfolio treatment
- For standardized (STA) portfolios, RWAs can change due to:
  - balance sheet growth (푔푔푡푡𝑐𝑐),
  - new flows of NPLs (∆푁푁푁푁푁푁𝑡𝑡),
  - new provisions for credit losses (푁푁푃푃푃푃푃푃𝑡𝑡),
  - exchange rate movements (∆퐹퐹퐹퐹𝑡𝑡𝑐𝑐),
  - triggered portion of off-balance sheet items (푈푈푈푈푁푁𝑖𝑖,𝑡𝑡−1𝑐𝑐,𝑗𝑗).
- Triggered portion assumptions: baseline = zero percent; adverse = 20 percent of off-balance sheet items over a 5-year horizon.
- For IRB portfolios:
  - Bank-specific regulatory credit risk parameters (TTC PDs, downturn LGDs, stressed EAD) were used to derive RWAs under baseline and adverse scenarios.
  - Weighted average through-the-cycle (TTC) PDs adjusted according to:
    - ∆TTCU_NP_t = (∆NiT_NP_t) * 0.5
    - This reflects adjustments to long-run average (TTC) PDs based on recent point-in-time (PiT) PDs and migration within non-defaulted rating grades.
  - Change to EAD governed by an equation that includes:
    - prior period EAD net of provisions,
    - credit growth g_t_c (demand effects included; supply effects disallowed),
    - fraction of foreign currency loans f_i_c,
    - FX depreciation c_t_FX∆,
    - shocks to triggered credit lines and guarantees,
    - undrawn guarantees UCL.
  - Stressed credit conversion factors on undrawn credit lines and guarantees were informed by historical off-balance sheet migration during stress, drawing on banks’ Pillar III disclosures.
- RWAs computation specifics:
  - Dependent on stressed credit risk parameters, correlation assumptions, and effective maturity; sector-specific.
  - In line with Basel III: applied scaling factor of 1.06 to credit RWAs and used a 1.25 multiplier to the correlation parameter of all exposures to financial institutions.
  - Difference in granularity of RWA calculation recognized by applying original scaling factor (ratio of model-calculated RWAs to reported RWAs at time T0).

### Interest rate risk in the banking book (IRRBB)
- Assessment decomposed into:
  - base effect: change in interest income/expense due to changes in outstanding amounts of interest-earning assets/liabilities absent rate shocks; computed as effective interest rate × outstanding amount per balance sheet item over stress horizons.
  - gains or losses under interest rate shocks: add-on using gap-analysis assessing cash-flow effects from general increase in interest rates across repricing buckets (less-than-1-year to 5-year), applied as positions reach repricing.
- Funding risks included as part of interest rate risk assessment: rising funding rates (deposit rates, debt security interest) affect repricing of sensitive liabilities.
- Net interest income projection = base component + gains/losses due to interest rate shocks.
- Interest payments assumed to accrue only on performing exposures under baseline and adverse scenarios; interest revenue on performing exposures calculated on gross carrying amount. No interest projected on non-performing exposures (conservative approach).

### Ireland-specific IRRBB inputs and modeling
- Inputs: historical aggregated interest rates time series and sensitive assets/liabilities from IRRBB template.
- Cost of funding and lending rates treated as functions of macroeconomic scenario variables.
- Rates for new business (front-book) mapped into:
  - Assets: loans and debt securities (decomposed by counterparties; corporate, household retail, household mortgage, sovereign, corporate debt securities, financial).
  - Liabilities: overnight deposits, term deposits, debt securities.
- Same interest rate on bank bonds applied to asset and liability sides of banking book.
- One sample bank reporting only total sensitive assets/liabilities: shocks computed on aggregated sub-components when needed.
- Stressed interest rate paths produced gains/losses per bucket.

### Interest rate results and pass-through
- Satellite model results: short-term and long-term rates significant drivers.
  - Asset side: lending rates highly correlated with short-term rate (dominant floating-rate loans).
  - Liability side: overnight deposit cost largely driven by short-term rate; term deposits and bond rates driven by both short-term and long-term rates.
  - Strong pass-through from Irish sovereign bond yield and short-term rate to funding rates, particularly term deposits and corporate debt securities.
- Projected net interest margin:
  - Under adverse scenario, net interest margin declines by about 0.3 percentage point on average for sample banks.
  - Historical net interest margin of total banking sector ≈ one percent since 2008.

### Securities valuation and other market risks
- Securities-level data used to measure gains/losses from changes in risk-free rates and credit spreads.
  - Bank-specific duration and yield derived from unique securities holdings.
  - Gains/losses calculated using modified duration approach.
  - Covers debt securities in FVTPL and FVOCI. Rebalancing not allowed through horizon.
  - For Amortized Cost (AC) securities: loss provisions calculated using risk parameters of AC lending portfolios as proxy.
  - To account for imperfect hedging under stress: assumed 20 percent of AC portfolios exposed to full valuation shock under adverse scenario.
- Equities held with trading intent:
  - Fair value impact follows projected equity prices under scenarios, subject to minimum threshold using approach similar to EBA 2018 methodology.
  - Market impact constraint: ∆Eq ≤ 1.5 ∗ (−0.20% ∗ (EEELLLL + EEEESShhLLoott))
    - VaR scaling factor set to upper bound of 1.5.
    - Trading position includes fair value of long positions offset by short positions in equity instruments in trading book.
  - Note: gross position of equities in trading book = 0.4 percent of total assets or 5 percent of total CET1 capital; net exposure below 1 percent of both total assets and total CET1 capital (considered immaterial).

### Net income projection, other income/expenses, and dividend policy
- Net income (profit and loss) projected using all risk factors; mainly driven by gains/losses from credit risk, market risk, and interest rate risk.
- Other income statement items projected to grow with balance sheet size (net fee and commission income, operational and administrative expenses).
  - Under adverse scenario, growth in non-interest income and expenses subject to a zero percent floor.
  - Extraordinary income/loss assumed not to recur.
- Corporate income tax:
  - Factored into profit and loss calculations.
  - Set at banks’ effective tax rate with a cap at 30 percent.
- Dividend policy assumption:
  - Dividends paid at 25 percent of current period net income after taxes by profitable banks (only if net income positive) and in compliance with supervisory capital requirements.
  - Banks not allowed to issue new shares or repurchases during stress horizon.

### Scenario-based solvency stress test results and contributions
- Overall finding: stress test confirmed Irish banking sector resilience to severe macro-financial shocks while revealing pockets of vulnerability as economy exits pandemic support.
- Baseline scenario results:
  - Retail banks: fully loaded CET1 ratio trending from 16.4 percent to 17.8 percent.
  - Large international banks: fully loaded CET1 ratio trending from 19.9 percent to 27.9 percent.
  - No banks falling below hurdle rates in baseline.
  - Slower capital accumulation for retail banks reflects low pre-provision income and limited income generation capacity.
- Adverse scenario results:
  - No banks fall below hurdle rates, supported by high initial capital positions and high pre-provision income of large international banks.
  - Aggregate CET1 ratio changes by the 5th year:
    - Retail banks: declines by about 6.7 percentage points.
    - Large international banks: declines by about 0.4 percentage point.
  - Trough capital depletion can reach:
    - Retail banks: 7.2 percentage points.
    - Large international banks: 2.3 percentage points.
  - Distribution of ending CET1 ratios:
    - Retail banks: generally lower, ranging from 6 to 16 percent.
    - Large international banks: range from 12 to 27 percent.
  - Within retail group, two banks experience larger capital depletion than peers.
- Contributors to system-level capital ratio decline in adverse scenario (percentage points):
  - Credit risk provisioning: about 7.4 percentage points.
  - Interest rate risks: 1 percentage point.
  - Risk weighted assets (RWA): 0.6 percentage point.
  - Market risks: 0.4 percentage point.
- Credit loss composition by borrower type:
  - Retail banks: corporate and NBFI record 77 percent of total credit losses for retail banks.
  - Large international banks: corporate and NBFI record 87 percent of total credit losses for large international banks.
- Additional contributors:
  - Dividend distribution contributes about 0.7 percentage point to aggregated capital depletion in the adverse scenario, primarily driven by profit generated by large international banks.

*Source: IMF staff.*

### 45. Although loan payment moratoria have now largely expired, the initial take-up rate

### Although loan payment moratoria have now largely expired, the initial take-up rate

### Loan moratoria and credit quality
- Irish retail banks provided over €24 billion of payment breaks to households and corporates; 99 percent have now expired.
- Majority of loans with expired moratoria have returned to full payments, but a notable share has been recorded in stage 2 and stage 3:
  - As of mid-2021, 40 percent of mortgages in stage 2 were previously under moratoria; corporate loans at 23 percent.
  - Among the total, about 40 percent of mortgage balance with payment breaks had some forbearance history.
- Loans subject to public guarantees show much lower previous-moratoria shares, with the majority remaining in stage 3.
- Policy implication: The end of policy support warrants close monitoring of loans as borrowers exit support programs.

### Sectoral take-up of payment breaks
- Contact-intensive and severely hit sectors had the highest usage of payment breaks (data as of May 2021, consolidated domestic banking data):
  - Accommodation and food: 54 percent take-up rate.
  - Entertainment and recreation: 24 percent.
  - Real estate activities: 15 percent.
- Implication: Corporate portfolios with high payment-break usage have higher risk profiles; monitoring solvency and liquidity implications in these segments is important for financial stability.

### Sensitivity analysis of moratoria unwind (add-on to adverse scenario)
- Stress-test assumptions for additional flows from expired moratoria over a 5-year horizon:
  - 50 percent of credit portfolios with expired moratoria under stage 2 assumed to flow to stage 3 evenly across a 5-year horizon.
  - 50 percent of credit portfolios with expired moratoria under stage 1 assumed to flow to stage 2 evenly across a 5-year horizon.
  - Rationale: partial 50 percent flow avoids double counting with macro-financial shocks in the adverse scenario and allows part of loans to improve or remain in current category.
- Results of this add-on:
  - Additional credit impairment brings an extra 110 bps CET1 depletion to the adverse scenario.
  - One retail bank falls below the CET1 and Tier 1 hurdle rates.
  - Maximum capital shortfall against CET1 hurdle rate amounts to about 0.2 percent of GDP.
- Policy implication: Continued monitoring of portfolios that benefited from payment breaks and broader policy support is necessary as policies phase out.

### Concentration risk analysis (simultaneous default of top 5 exposures)
- Method: hypothetical simultaneous default of five largest borrowers per bank; two recovery assumptions tested (zero-recovery and 50 percent provisioning).
- Zero-recovery results (aggregate CET1 starting at 19.5 percent):
  - Simultaneous default of five largest NFC borrowers causes aggregate CET1 to decline by 6.6 percentage points from 19.5 percent to 12.9 percent.
  - Large international banks experience higher depletion (8.9 ppts) than retail banks (1.7 ppts).
  - No banks fall below the 4.5 percent CET1 threshold.
  - Default of five largest NBFIs causes CET1 decline by 5.7 percentage points (large international banks at 8.3 ppts; retail banks at 5.7 ppts).
- 50 percent provisioning results (aggregate CET1 starting at 19.5 percent):
  - Simultaneous default of five largest NFC borrowers causes aggregate CET1 to decline by 3.3 percentage points from 19.5 percent to 16.2 percent.
  - Large international banks depletion 4.4 ppts; retail banks 0.8 ppts.
  - No banks fall below the 4.5 percent CET1 threshold.
  - Default of five largest NBFIs causes CET1 decline by 2.8 percentage points (large international banks at 4.2 ppts; retail banks at 1.2 ppts).

### Other international banks: profitability and sensitivity analysis
- Business model and income composition (as of end-2021, percentages of total RWAs):
  - Net fees and commission: 3.7 percent of total RWAs.
  - Net interest income: 1.1 percent of total RWAs.
  - Net trading income: 0. 8 percent of total RWAs.
  - Loan loss provisions: 0.2 percent of total RWAs.
  - Residual component (around 90 percent is administrative expenses): -5.4 percent of total RWAs.
  - Other comprehensive income: (- 0.3 percent).
  - Result: compressed revenue and high cost-base led to an overall loss and capital depletion of about 0.1 percentage point of RWA in 2021.
- Sensitivity analysis methodology:
  - Net profit decomposed into five subcomponents transformed into P&L ratios; tests impose lower bound (10th percentile) for NIM, net trading income, and net fees and commission ratios, and upper bound (90th percentile) of net loan loss ratios using historical 2008–2020 series.
  - Off-balance sheet credit exposure converted with ad-hoc factor 50 percent onto balance sheet.
  - Interest rate risk also assessed structurally via maturity ladder and IRRBB templates; rate shocks selected at 10th percentile of historical changes in lending minus funding spreads.
- Sensitivity test results (aggregate, relative to mid-2021 baseline):
  - Loss contributions to capital ratios (percentage points of RWA):
    - Net fees and commissions: 3.1 percentage point decline.
    - Credit risks: 2.5 percentage points.
    - Interest rate risks: 0.9 percentage points.
    - Trading risks: 0.6 percentage points.
    - Off-balance sheet exposure: 0.5 percentage points.
  - Under structural approach, interest rate risks contribute 1.4 percentage points (broadly consistent with ratio-based approach).
  - Heterogeneity: bank-specific capital impact (sum of all risk factors) ranges from 2.3 to 27.3 percentage points relative to baseline.
  - No bank in this group falls below the hurdle rate of 4.5 percent CET1 ratio given high initial CET1 buffers.

### Liquidity risk analysis — key findings
- Tests conducted: LCR, cash-flow based (12-month), and NSFR (100 percent threshold) for five retail banks, three large international banks, and four other international banks.
- Data sources: COREP reports as of mid-2021; resident and non-resident deposits from CBI supervisory templates; comparisons 2019–2021 to assess liquidity buffer buildup.
- Aggregate liquidity positions:
  - Loan-to-deposit ratio: 86 percent overall.
    - Large international banks: 89 percent.
    - Domestic retail banks: 75 percent.
  - Non-resident deposits: 88 percent of total deposits for international banks.
  - Contingent liabilities for international banks: €108 billion or 40 percent of total contractual outflows within 12 months.
  - Banks’ total assets increased by about 17 percent since 2019.
- LCR and liquid assets:
  - Aggregate LCR increased from 153 percent at end-2019 to 179 percent at mid-2021.
  - Liquid assets to total assets rose from 25 to 28 percent over the same period; euro area average at 20 percent as of mid-2021.
- Maturity mismatches and funding structure:
  - 77 percent of deposits were placed within the overnight bucket.
  - Over 50 percent of cash inflows would materialize beyond the first three months, creating a frontloaded outflow / backloaded inflow mismatch.
  - Implication: Vulnerability to sustained liquidity stress and reliance on non-resident funding for international banks; continued monitoring of maturity structures and off-balance sheet exposures recommended.

*Sources: CBI, ECB, IMF staff estimates.*

### 60. The funding profile of banks reflects their diverse business models   and market

### 60. The funding profile of banks reflects their diverse business models   and market

### Funding profiles by bank group
- Retail deposits account for 61 percent of retail banks’ total funding.
- International banks:
  - Deposits from credit facilities: 35 percent.
  - Operational deposits: 22 percent.
  - Other unsecured funding: 14 percent.
  - Retail deposits: 1 percent.
- Changes since 2019 in contractual outflows:
  - Retail banks: increase of €155 billion.
  - Large international banks: increase of €177 billion.
  - Other international banks: change of €2 billion (broadly constant).
- Other international banks rely more on parent funding; COVID liquidity support measures did not expand their funding sources.

### Composition of HQLA and LCR levels
- HQLA level 1 comprised 99 percent of total HQLA.
- Central bank reserves and holdings of sovereign securities account for around 60 percent and 38 percent of total HQLA, respectively.
- Retail and large international banks accumulated HQLAs sharply during the pandemic; other international banks broadly maintained their level of liquid assets.
- Asset encumbrance ratio: around 15 percent (as of mid-2021).

### LCR-based stress test: scenarios and parameters
- Sample: five retail banks, three large international banks, four other international banks.
- Four scenarios:
  - Scenario S1 (Regulatory LCR): standard CRR parameters.
  - Scenario S2 (Retail run-off): higher run-off rates for retail-related claims; liquid assets usable with CRR haircuts.
  - Scenario S3 (Wholesale run-off): higher run-off rates for wholesale-related claims; liquid assets usable with CRR haircuts.
  - Scenario S4 (Combined run-off + price shock): stressed run-off rates for retail and wholesale (whichever is higher) and haircuts when liquidating assets informed by ECB valuation haircuts.
- Non-resident deposits subject to higher shocks: unstable wholesale deposits from financial institutions and nonfinancial corporations at around 100 and 50 percent of outflow rates in the most severe scenario, respectively.
- Selected run-off and haircut parameters (as presented):
  - stable retail deposits: S1 5% | S2 10% | S3 5% | S4 10%
  - other retail deposits: S1 10% | S2 20% | S3 10% | S4 20%
  - operational deposits: S1 5-25% | S2 5-25% | S3 15-35% | S4 15-35%
  - non-operational deposits: S1 20-40% | S2 20-40% | S3 30-50% | S4 30-50%
  - committed facilities to retail customers: S1 5% | S2 10-15% | S3 5-10% | S4 10-15%
  - committed facilities to corporate customers: S1 10-30% | S2 10-40% | S3 20-50% | S4 20-50%
  - level 1 assets haircuts in S4: nono-5/0% (table notation preserved).

### LCR-based stress test: results
- As of mid-2021 all sampled banks met regulatory minimum threshold of 100 percent.
- Standard LCR by group (mid-2021):
  - Large international banks: 161 percent.
  - Retail banks: 198 percent.
  - Other international banks: 211 percent.
- Interpretation:
  - Large international banks’ lower standard LCR may reflect higher unsecured wholesale funding and contingent liabilities.
  - Other international banks’ high standard LCR reflects low contractual outflows relative to HQLAs.
- Stress impacts:
  - Retail banks: larger impact under the retail scenario.
  - International banks: higher stress under the wholesale scenario.
  - All banks were able to withstand the most severe 30-day shock due to high initial liquidity buffers.

### Foreign currency LCR stress testing
- Separate LCR-based stress tests for U.S. dollar and Sterling using same assumptions.
- Coverage:
  - Seven banks report LCR in U.S. dollar.
  - Six banks report LCR in Sterling.
  - Three banks (mostly retail) not included due to insignificant positions.
- Findings:
  - U.S. dollar: international banks had lower starting LCR in USD, with two banks below 100 percent as of mid-2021; international banks experienced higher stress than retail banks across scenarios, driven by weaker initial positions and off-balance sheet exposures.
  - Sterling: retail banks had a lower starting point and higher depletion rate across scenarios, with one bank in each group already below 100 percent; retail banks show higher vulnerability to Sterling funding shocks consistent with concentrated U.K. market exposure.
- Policy implication: continued monitoring of foreign currency liquidity conditions is important to identify potential liquidity gaps.

### Cashflow-based liquidity stress test: methodology and data
- Purpose: assess adequacy of liquid assets to offset cash inflow and outflow shocks over time using net liquidity position (cumulated net funding gap minus cumulated counterbalancing capacity).
- Data: maturity ladder (COREP C66.00) as of mid-2021 with 21 maturity buckets (overnight to >5 years), secured/unsecured funding, initial stock and changes in counterbalancing capacity.
- Maturity mismatch:
  - About 66 percent of total outflows projected within less than 30 days.
  - Open maturity bucket holds about 53 percent of total outflows.
  - Including committed facilities, share of outflows increases to 73 percent for buckets below 30 days and to 63 percent for the open maturity bucket.
  - For inflows, 50 percent of total inflows concentrated in loans and advances with maturity longer than 30 days.
- Additional observations:
  - Around 75 percent of total corporate and retail loans have residual maturities more than 12 months.
  - Among total retail and wholesale deposits of the sample banks, 44 percent are non-resident deposits; these are mostly concentrated in funding from credit institutions (94 percent) and nonbank financial institutions (51 percent) and dominated by international banks.
- Counterbalancing capacity composition (mid-pandemic):
  - Central bank reserves and marketable securities accounted for 73 percent and 19 percent of total counterbalancing capacity, respectively.
- Treatment of inflows: in principle 100 percent of contractual inflows except inflows from loans to retail and corporate customers set to 0 percent (to reflect loan moratoria and assumption that banks are not allowed to deleverage).

### Cashflow-based stress test: scenarios, parameters and haircuts
- Horizons: 5-days, 4-week, 3-months, 12-months.
- Three stress severity levels: mild market stress, medium market stress, severe market stress.
- Two approaches to counterbalancing capacity (CBC):
  - Full CBC: fully endogenous central bank liquidity supply if banks have unencumbered eligible collateral.
  - Full CBC with market haircuts: full CBC but with market-specific haircuts and bank-specific market price effects on CBC elements.
- Selected run-off and haircut parameter ranges (as presented):
  - Unsecured bonds: 40-100% (outflows).
  - Regulated covered bonds: 25-70% (outflows).
  - Securitisations and others: 100% (outflows).
  - Repos across all asset classes: 100% (outflows).
  - Stable retail deposits: 2-10% (outflows).
  - Other retail deposits: 5-30% (outflows).
  - Operational deposits: 5-30% (outflows).
  - Non-operational corporate deposits & other: 20-100% (outflows).
  - Derivatives: 100% (outflows).
  - Committed facilities: 10-100% (outflows).
  - Loan inflows from retail and corporates: 0% (inflows).
  - Loan inflows from banks and NBFIs: 30-100% (inflows).
  - Haircuts on market prices:
    - Level 1 assets: 95%
    - Level 1 covered bonds: 90%
    - Level 2A assets: 85%
    - Level 2B assets: 50-75%
    - Other tradable assets: 50%
    - Non tradable assets: 50%

### Cashflow-based stress test: results and vulnerabilities
- General result: banks can withstand mild and medium liquidity outflows over the short-term; liquidity positions weaken beyond three months.
- International banks are more susceptible to shocks even in the short term due to larger share of wholesale funding and off-balance sheet exposures.
- CBC depletion (as share of total assets) over a 5-day horizon:
  - Mild scenario: decline of 7 percent (system-wide).
  - Severe scenario: decline of 22 percent (system-wide).
  - Large international banks: decline ranges from 8 to 25 percent of total assets.
- Specific outcomes:
  - Under the 5-day severe scenario one large international bank would experience a funding gap (driven by high off-balance sheet exposure) with a small liquidity shortfall relative to total banking system assets (less than 1 percent).
  - By end of 12 months, CBC depletion reaches up to 12 percent (mild) and 29 percent (severe) of total assets.
  - Under the most severe condition by 12 months, two retail, two large international and three other international banks could experience shortfalls up to 4 percent of total sector assets.
  - System-level liquidity shortfalls remain manageable: only large international banks register a liquidity shortfall under the most severe scenario at 0.5 percent and 1.8 percent of total sector assets over the 3-month and 12-month horizons, respectively.

*Source: IMF staff analysis as presented in the chapter.*

### 76. The cash flow-based test echoes findings from the LCR-based stress test on

### 1irlea2022013 - 76. The cash flow-based test echoes findings from the LCR-based stress test on

### Cash flow-based stress test: foreign-currency liquidity vulnerabilities
- International banks that rely on wholesale and non-resident parent funding are more vulnerable to dollar denominated funding stress, with net funding gap at the system level reaching to 2 percent of total sector assets under the most severe scenario over a 12-month horizon.
- The funding gap associated with Sterling denominated outflows peaks at 0.1 percent of total system assets.
- The cashflow-based test (12-month horizon) and the LCR analysis (30 days horizon) arrive at similar conclusions despite methodological differences, including:
  - lower liquidity capacity of large international banks due to large off-balance sheet exposures; and
  - weak liquidity position of the banks in significant foreign currencies such as the U.S. dollar over the one-month horizon.
- Off-balance sheet items include issuance of Insurance Letters of Credit (“ILOCs”), which are approximately 80 percent collateralized on average and considered by the CBI as less unstable relative to other committed facilities.

### NSFR-based liquidity stress test: methodology and results
- Aggregate NSFR for sample banks as of mid-2021 stood at 150 percent, above the minimum requirement of 100 percent; no single bank was below the threshold.
- Drivers of elevated NSFR: high share of retail deposits for retail banks (treated as stable funding) and reliance on long-term corporate and interbank deposits for international banks; high share of HQLA assets reduces required ASF.
- Stress-test approach: volume-based migration from long-term to short-term funding applied to ASF instruments (retail deposits—stable and unstable; liabilities from non-financial customers; liabilities from financial customers and central banks; interdependent liabilities such as credit and liquidity facilities; and other liabilities such as trade payables). Equity instruments were not stressed.
- Assumed migration rates in stress scenario:
  - 50 percent of funding within six- to twelve-month bucket flows to less than six-month bucket;
  - 35 percent of funding with more than one year maturity migrates to the six- to twelve-month bucket;
  - 15 percent flow rate from over one-year bucket to the less than six-month bucket.
- Applicable required stable funding and ASF factors were unchanged under the stress scenario.
- Results:
  - Most banks maintain a stable funding profile under the adverse migration shock.
  - Only one retail bank falls marginally below the 100 percent NSFR threshold in the adverse scenario.
  - Large international banks experience a larger decline in NSFR than retail banks; other international banks retain the highest levels of stable funding both before and after the shock.

### Interconnectedness analysis: topology and exposures
- Network topology: meaningful interconnections between large international banks, cross-border banks, and NBFIs; domestic interbank linkages are limited.
- Large international banks’ exposures to NBFIs account for 45 percent of their total exposures; around 92 percent of these exposures are with foreign NBFIs (dominated by United Kingdom, France, and the United States).
- Sector composition of those NBFI exposures: over 80 percent classified as “financial service activities, except insurance and pension funds” or “activities auxiliary to financial services”; remainder belongs to insurance and pension funds.
- Domestic interbank exposure (loans and debt securities):
  - Total bilateral exposures represent 0.05 percent of the sample bank assets and 0.07 percent of GDP.
  - The majority of domestic interbank assets and liabilities are vis-à-vis retail banks: 99 percent and 82 percent of total interbank assets and liabilities, respectively.
- Funding composition and interbank roles:
  - Retail banks’ main source of funding: household and corporate deposits (72 percent of total liabilities).
  - International banks obtain most funding from domestic and cross-border banks, NBFIs and corporates (57 percent of total liabilities).
  - As of mid-2021, banks which hold about 60 percent of total assets in the banking system are classified as net borrowers; the remaining banks are net lenders.
- Domestic interbank contagion indices:
  - Retail banks generally have higher contagion and vulnerability indices than international banks, but index levels measured as losses in percent of capital are small, around 5 percent.
  - Conclusion: limited interbank exposures within Ireland and sufficient capital to withstand domestic interbank shocks via direct balance-sheet exposures.

### Bank–NBFI interconnectedness and spillovers
- OFIs (other financial institutions: FVCs, SPVs, OFI residual) are large: total assets of OFIs are 400 percent of GDP as of end-2020.
- Ireland represents the largest FVC market in the Euro Area by regional comparison.
- Bank–OFI linkages and magnitudes:
  - Irish bank assets and liabilities to OFIs account for 4 and 7 percent of total assets and liabilities, respectively.
  - Large international banks’ credit exposure to NBFIs: 29 percent of their total credit exposure; retail banks: 1 percent.
  - OFIs provide funding to NFCs and households at close to 10 and 32 percent of total loans to NFCs, respectively; this is larger than the Irish banking sector’s role at close to 6 percent of total NFC loans.
- Bank–NBFI network analysis approaches:
  - Approach 1: Irish domiciled NBFIs only—OFIs account for 7 of 10 firms and 89 percent of total exposure.
  - Approach 2: Domestic and foreign NBFIs—OFIs account for 100 percent of total exposure; large exposures originate mainly from United States (33 percent), United Kingdom (19 percent), and France (11 percent).
- Vulnerability (inward spillover) results via credit channel:
  - Including domestic NBFIs raises large international banks’ vulnerability index from almost 0 to 15 percent of total capital on average; retail banks increase from 4.4 to 4.8 percent of total capital.
  - Including domestic and foreign NBFIs further increases indices; for one large international bank the vulnerability index reaches almost 100 percent of capital.
- Data limitations: limited information on NBFIs’ asset exposures and funding channels implies the analysis mainly captures inward credit spillovers and may under-estimate funding-channel impacts.

*Source: IMF staff.*

### 95. The cross-border contagion analysis aimed at assessing Ireland banking systems’

### 95. The cross-border contagion analysis aimed at assessing Ireland banking systems’ 

### Cross-border interbank linkages and trends
- Based on the Bank for International Settlements (BIS) consolidated banking statistics as of 2021Q3:
  - Irish domestic banks’ foreign asset claims declined from $234 to $15 billion since the GFC.
  - Irish domestic banks’ foreign liabilities declined from $358 to $31 billion since the GFC.
- BIS locational data (unconsolidated, includes foreign subsidiaries) shows a similar downward trend but at a higher level of exposure relative to consolidated data, driven by higher cross-border linkages of international banks partly via intragroup exposures.
- Top asset exposures of Irish domestic banks (asset side): United Kingdom (25 percent), France (23 percent).
- Top liabilities exposures of Irish domestic banks (liabilities side): France (35 percent), Switzerland (11 percent), United States (10 percent), United Kingdom (7 percent).
- BIS consolidated statistics for total nonfinancial private sector cross-border claims of Irish banks:
  - Nonfinancial private sector accounts for 60 percent of total cross-border claims.
  - Country shares of these claims: United Kingdom (69 percent), United States (9 percent), France (7 percent), Spain (4 percent), Netherlands (2 percent), Germany (2 percent).
- On the liability side, nonbank financial institutions dominate total claims of foreign banks on the Irish domestic economy at 52 percent, reflecting high foreign investment in investment funds and other financial institutions such as FVCs and SPVs.
- Country ranking for foreign investor exposure on Ireland (total exposure and share of total foreign claims):
  - United States: 116 billion, 32 percent
  - United Kingdom: 21 percent
  - France: 16 percent
  - Germany: 9 percent
  - Spain: 7 percent
  - Italy: 5 percent
  - Belgium: 4 percent

### Cross-border contagion simulation design
- Coverage: 29 banking sectors, most operating within Europe and with direct linkages to Ireland.
- Data: BIS consolidated banking statistics as of Q3 2021.
- Shock assumptions in combined credit and funding shock simulations:
  - 30 percent loss on asset claims in case of default of a counterparty.
  - 65 percent roll-over of funding (i.e., 35 percent decline in interbank funding).
  - 30 percent haircut on assets that may be forced to be liquidated due to loss of funding.
- Scenarios allow decomposition of bank losses into credit and funding channel contributions to impacts on bank capital.

### Contagion and vulnerability results for Irish banks
- Ireland’s scores:
  - Vulnerability index: 0.2 (percent of total capital).
  - Contagion index: 0.1 (percent of total capital).
  - Irish banking sector ranks lower in contagion and vulnerability indices than many counterparties, but vulnerability exceeds contagion for Ireland.
- Top inward spillover risks to Irish banks originate from:
  - France: capital loss contribution of 2.2 percent of capital (combined channels).
  - United Kingdom: capital loss contribution of 1.1 percent of capital.
  - United States: capital loss contribution of 0.5 percent of capital.
- Channel-specific observations:
  - For France, loss contribution via the funding channel is higher than via the credit channel.
  - For the United Kingdom, loss contribution via the credit channel is higher than via the funding channel, signaling non-negligible credit exposure of Irish domestic banks to U.K. banks.
- Both inward and outward spillover risks are characterized as muted for Irish banks in aggregate given reduced interbank integration.

### Bank exposure to climate risks: overview and direct exposures
- Policy context:
  - Ireland’s 2021 Climate Act commits to net-zero greenhouse gas (GHG) emissions by no later than 2050 and to more than halving emissions by 2030.
  - Government has introduced an incremental charge on carbon tax, from €33.5 to €100 per ton of CO2 emissions by 2030, while recycling revenue to mitigate higher energy bills, finance climate-related investment and ensure a just transition.
  - Current carbon tax (introduced in 2010) is at €41/ton (note: slightly different start date and timing of increases for different fuels).
- Physical risks:
  - Severe coastal flooding risk from sea level rise: an average sea level rise of between 0.5 to 1 meter, in combination with storm surge events, could inundate as much as 1,000km2 of coastal lands around Ireland if no protective measures are undertaken.
  - Growing weather variability: more frequent extreme rainfall increasing river flooding and exacerbating water overflow; re-emergence of droughts adversely affecting crop cycles.
- Banking system exposures:
  - Roughly 15 percent of bank NFC loans are to sectors with a high carbon footprint.
  - More than 20 percent of loans are to sectors exposed to high physical hazards, dominated by floods — described as well above Euro Area average.

### Transition risk sensitivity-based stress test: coverage and purpose
- Purpose: quantify financial-stability implications of transition to a low-carbon economy via impact on corporate PDs and bank capital from carbon tax increases.
- Shock modeled: increase in carbon tax from €33.5 to €100 per ton of CO2 emission.
- Covered population: 1,400 firms, constituting around 68 percent of total corporate sector debt.
- Data sources:
  - Firm-level balance sheet data: Capital IQ, covering 1995 to 2020.
  - Firm-specific PDs proxied by Moody’s expected default frequency: Moody’s Analytics, covering 1999 to 2020.
  - Industry-level carbon intensity: Eurostat.
  - Banking sector corporate loan exposure and NPLs by industry: regulatory reporting templates.
- Behavioral response inclusion:
  - Firms allowed to pass the entire carbon tax shock to end consumers.
  - Demand response modeled using a homogenous short-run price elasticity of -0.3 across all carbon producing industries (from Carbon Pricing Assessment Tool development by the IMF Fiscal Affairs Department).
  - Analysis does not consider other behavioral/transitional responses (e.g., energy efficiency improvements), which are assumed to occur only over the long term.
- Projection horizons and scenario types (four main scenario variants, each over 1-year to 5-year horizons):
  - Instantaneous carbon tax shock from €33.5 to €100, no firm behavioral response.
  - Incremental carbon tax shock from €33.5 to €100 (linear over 5 years), no firm behavioral response.
  - Instantaneous carbon tax shock from €33.5 to €100, assuming firms pass entire cost to consumers.
  - Incremental carbon tax shock from €33.5 to €100 (linear over 5 years), assuming firms pass entire cost to consumers.
- Static and dynamic firm balance sheet projections:
  - Static: one-year instantaneous PD impact.
  - Dynamic: five-year horizon with either instantaneous or incremental implementation; firm balance sheets evolve dynamically based on predefined accounting principles to capture non-linear solvency and liquidity effects.

### Transition stress test methodology steps and parameters
- Step 1 — Bridge estimation:
  - Derive a bridge equation estimating firm PDs in logit form with three firm-level balance sheet indicators: interest coverage ratio (ICR), current ratio, and leverage ratio.
  - Estimation using a fixed effects panel regression with unique model averaging and sign constraints on a subset of publicly listed firms.
  - Regression results (Table 8) — Estimated effects on Expected Default Frequency in logit form (RHS variables measured in fraction):
    - Interest coverage ratio: -0.01*
    - Leverage ratio: 1.69*
    - Current ratio: -0.01
    - (Intercept): -7.39*
    - Individual firm fixed effects: Yes
    - R-square: 0.38
    - (* denotes p value less than 0.1)
- Step 2 — PD projection for extended firm sample:
  - Apply estimated coefficients to extended firm sample covering publicly and non-publicly listed firms.
  - Project vulnerability indicators at firm level under each scenario using firm-specific balance sheet information and carbon emissions, then compute PDs before and after shock.
  - Dynamic projection uses accounting identities to evolve balance sheet and profit components; forward-looking projections assume constant economic conditions and no shifts in business models, energy usage, or technological advancement to lower emissions.
- Step 3 — Link to banking sector:
  - Aggregate firm-level PD projections at industry level using total debt as weights.
  - Attach changes in aggregated PDs to bank portfolio starting PDs by industry to compute additional loan loss provisions and impact on bank capital.
  - Bank PDs by industry proxied using new NPL flows in 2020 (change in NPL stock between 2019 and 2020 assuming an average write-off rate of 30 percent).
  - Loss given default (LGD) assumed homogenous across industries at 30 percent.
  - The relative composition of corporate portfolios between green and brown segments determines the size of the capital impact, reflecting non-negligible carbon intensive portfolios held by banks.

*Source: Excerpt from IMF staff analysis in the provided PDF content.*

### 109. The regression output suggests that interest coverage ratio and firm leverage

### 1irlea2022013 - 109. The regression output suggests that interest coverage ratio and firm leverage

### Regression drivers of firm PDs
- Regression output indicates interest coverage ratio (ICR) and firm leverage dominate the evolution of firm PDs (Table 8).
- High P values and sizable coefficient estimates of ICR and leverage ratios for firms in the regression sample.
- Intuition: lower interest coverage ratio and higher leverage ratios are expected to drive up firms’ PDs.
- Note: the high dominance of leverage ratios can be explained by construction, as leverage ratios are key input into the derivation of the market based EDF indicator.
- Footnote: "PD here refers to probability of default in logit form."

### PD projections under carbon tax shocks — sectoral and temporal patterns
- Projections suggest transition risks could be material for Irish firms over the medium term (Figure 43).
- Energy intensive sectors with largest PD increases over a five-year horizon across all scenarios: agriculture, electricity and gas, and transportation.
- Shock-response patterns:
  - Instantaneous shock with no firm behavioral response: impact can be significant for these sectors.
  - Full passthrough of carbon price shock to ending consumers can partially alleviate financial burden on firms and improve credit quality.
  - Firms subject to instantaneous carbon charges reach significantly higher PDs by the fifth year than those subject to incremental charges — implying a benefit from phase-in policies with gradual increases in carbon tax.
  - Higher increase in PDs as stress persists into the longer horizon confirms important non-linear effects of carbon tax on corporate financial health under assumption of constant carbon emissions.
- Policy implication highlighted: importance of speedy transitioning to greener technology to improve energy efficiency and ensure business viability for carbon-intensive segments.
- Note on modeling assumption: "All figures represent annual probability of default (n0n-logit and non-cumulative)."

### Banking sector spillovers from corporate PD increases
- Immediate impact on banking sectors appears contained, but cumulative capital impairment over the medium term can be substantial under severe scenarios:
  - Cumulative capital impairment can reach to almost 3.5 percentage points of CET1 capital (or 15 percent of existing CET1 stock) under the most severe scenario with instantaneous carbon tax shock and no firm behavioral response.
- This may curtail bank resilience to sustained transition shocks in absence of preemptive monitoring of carbon-intensive credit portfolios.
- Observation: underscores importance of advancing monitoring toolkit and developing relevant modeling approaches to properly assess banking-sector impacts of material climate risks.
- Caveat: calculation of capital impairment under carbon tax shock does not consider pre-provision income made in the baseline over the risk horizon; thus overall capital depletion should not be seen as akin to a traditional solvency stress test.

### Market valuation risk of bank-held debt securities
- Method: market losses computed using modified duration approach where shocks to corporate bond spreads are proxied by simulated corporate PDs under the Merton approach.
- Assumptions and parameters:
  - LGD is set to be homogenous at 30 percent.
  - D represents duration of debt securities; B initial bond yield; S corporate bond spread computed using Merton formula; r risk-free rate.
  - For sectoral granularity, duration and outstanding amount of bank debt securities provided by CBI under NACE Rev.2 classification.
  - Note in balance-sheet modeling: "50 percent of additional borrowing assumed in the form of short-term debt."
- Findings:
  - Market valuation risk associated with transition risk is found to be insignificant.
  - Cumulative losses over the medium term under various carbon tax shock scenarios peak at 0.04 percent of total CET1 capital.
  - Banks on aggregate are less exposed to securities issued by carbon intensive industries, currently at 0.5 percent of total holdings of corporate debt securities (Figure 45).
  - Carbon-intensive industries experience sharper decompression of corporate bond spread under various carbon tax shocks, at 5 percentage points on average, in contrast to 0.1 percentage points for “greener” industries in sample.

### Caveats and limitations of the transition-risk analysis
- Several caveats that may lead to over- or mis-estimation:
  - Absence of scenario-consistent dynamic projections that incorporate redistribution of tax receipts to offset initial impact and firms’ technological transformation may overestimate negative impact of carbon tax.
  - Dataset covers almost exclusively large publicly listed firms; estimated impact may not be fully representative of entire corporate sector which includes many small and medium size enterprises (mostly non-listed).
  - Analysis does not differentiate banks into retail banks (more sensitive to domestic carbon tax shocks) and international banks (foreign-oriented and less susceptible to domestic policies), which may mask heterogeneity within banking sector.
  - Analysis does not consider public subsidies towards certain segments post carbon tax levy (e.g., agriculture) to mitigate immediate financial difficulties.
  - Different metrics of carbon intensities (greenhouse gas emissions vs carbon emissions) can materially affect PD impacts, especially for agriculture which emits mostly methane and nitrous oxide (CO2 equivalent) rather than CO2.
  - Indirect tax effects via scope 2 and 3 emissions, as well as carbon border adjustment, are not considered but can be important for an open economy such as Ireland.

### Physical risk scenario-based stress test (severe flooding)
- Motivation: heightened flooding risks due to growing weather variability and sea level rise; frontloads long-term physical hazard impacts into a 5-year horizon.
- Shock calibration methodology:
  - Multi-pronged approach combining backward-looking (historical extreme flooding across Europe) and forward-looking (recent publications on potential coastal flooding damages) to set damage rates as percentage of total capital stock.
  - Damage rates over a 5-year horizon expected to range from 0.5 to 5 percent of total capital stock.
- Translation to GDP:
  - Historical elasticity between total capital stock and GDP estimated using OLS; coefficient on total capital stock (elasticity between capital and GDP) is estimated to be 0.83.
  - Therefore, shock to GDP in response to total capital loss is valued between 0.4 to 4 percent, relative to the baseline scenario, over the 5-year horizon.
- Scenario narratives and macro channels:
  - Shock to capital stock → drop in labor productivity → subsequent recovery via investment.
  - Lower capital → smaller future wealth → drop in private consumption and demand for labor → lower employment.
  - Drop in housing prices due to curtailed income and savings and falling housing market expectations led by flooding damage.
  - Rise in sovereign risk premium due to higher fiscal cost associated with recovery.
- Modeling: scenario calibration relies on the GFM modeling used in bank solvency stress test; scenario treated as standalone (not an add-on to adverse scenario) given timing uncertainty.

### Physical risk results and banking implications
- Focused on upper bound scenario (5 percent loss of capital stock) and on retail banks (international banks minimally affected).
- Impact:
  - Severe flooding event triggers non-trivial depletion of retail banks’ CET1 capital of around 240 bps before returning close to pre-shock level.
- Policy implication: prompt enhanced monitoring of banking sector exposure to climate-sensitive segments and precautionary actions (e.g., ensure loans secured with high quality collateral or insured against physical damages) to mitigate climate-related risks to financial stability.

### Insurance sector solvency stress testing — scope and summary metrics
- FSAP performed solvency and liquidity stress tests and climate risk analysis covering up to 25 (re)insurers.
- Top-down solvency ST used same macrofinancial scenario as banking ST; complemented by sensitivity analyses and a liquidity ST based on 2021 EIOPA stress test scenario.
- Climate risk analysis: bottom-up natural disaster risks, parametric increases in severity and frequency of weather-related loss events; transition risk analyzed top-down assuming change in market sentiment towards certain investment exposures.
- Sample and market coverage:
  - Top-down solvency ST performed for 25 large insurers on a solo-entity basis: 10 life insurers, 8 non-life insurers, 7 reinsurers.
  - Stress test reached representative market coverage of at least 70 percent in all three sub-sectors based on gross written premiums.
  - Participants’ aggregated balance sheet assets amount to €369 billion, of which €270 billion can be attributed to primary insurers which undertake predominantly life business.
- Pre-stress solvency and capital composition:
  - All 25 participants record before-stress solvency ratios above regulatory threshold of 100 percent, but individual levels differ widely.
  - SCR composition on aggregate relatively balanced: life underwriting risks dominate; market risks and non-life underwriting risks contribute roughly equal amounts.
  - Eight participating insurers calculate SCR with a full or partial internal model.
  - Insurers hold high-quality capital: 95 percent of eligible own funds are unrestricted Tier 1 capital, while only 1 percent is comprised of Tier 3.
- Long-term guarantee (LTG) measures and transitionals:
  - LTG measures and transitionals have limited effect in the Irish life insurance sector.
  - Nine insurers in the sample have permission to use the Volatility Adjustment (VA), the most relevant LTG measure.
  - Without any LTG measure or transitional being used:
    - Technical provisions of firms using such measures would be 0.2 percent higher.
    - Eligible own funds would be 0.7 percent lower.
    - SCR would be 1.1 percent higher, on average.
    - Median SCR ratio of LTG users in the sample would be around 168 instead of 171 percent.

*Source: IMF staff calculations and analysis in the provided content.*

### 124. The macrofinancial scenario specified by the FSAP for the banking sector stress test

### 124. The macrofinancial scenario specified by the FSAP for the banking sector stress test

### Scenario design and key adjustments for insurance ST
- The macrofinancial scenario used by the FSAP for the banking sector stress test was adjusted for the insurance stress test to be directly applicable to an insurer’s balance sheet.
- Core scenario features: lower and more volatile growth due to further outbreaks of lethal and highly contagious Covid-19 variants, de-anchoring of inflation expectations in the United States and/or advanced European economies, and extended global supply chain disruptions.
- For the insurance stress test, all shocks were assumed to occur at the beginning of the first year (instantaneous shock); market shocks were front-loaded so the maximum drawdown during the projection horizon is realized immediately after the reference date.
- Reference date: June 30, 2021.
- Instantaneous shock assumption: market shocks (declines in equity and property prices, etc.) realized immediately after the reference date.

### Market risk specification (adverse scenario shocks)
- Equity shocks (adverse scenario):
  - Ireland -58.0%
  - United States, Euro area -18.0%
  - Other advanced economies -20.0%
  - Emerging economies -35.0%
  - Holdings in related undertakings -15.0%
- Property shocks:
  - Residential, domestic -19.9%
  - Commercial, domestic -34.1%
  - Residential, other countries -5.0%
  - Commercial, other countries -18.0%
  - Alternative funds -8.0%
  - Private equity funds -10.0%
  - Infrastructure funds -5.0%
- Risk-free interest rates (adverse scenario):
  - EUR, 1 year +0.06%
  - EUR, 10 years +0.51%
  - USD, 1 year +0.30%
  - USD, 10 years +1.30%
  - GBP, 1 year +0.32%
  - GBP, 10 years +0.26%
- Sovereign bond spreads (adverse scenario):
  - Ireland +1.60%
  - Other low-yield advanced economies +0.25%
  - High-yield advanced economies +1.40%
  - Emerging and developing economies +1.80%
  - Supranationals 0.00%
- Corporate bond spreads (adverse scenario examples):
  - Non-financials, credit quality step 0 +0.60%
  - Non-financials, credit quality step 1 +0.75%
  - Non-financials, credit quality step 2 +1.00%
  - Non-financials, credit quality step 3 +1.30%
  - Non-financials, credit quality step 4 +1.90%
  - Non-financials, credit quality step 5 +2.90%
  - Non-financials, credit quality step 6 +4.20%
  - Non-financials, unrated +2.00%
  - Financials, CQS 0 +0.75%
  - Financials, CQS 1 +0.95%
  - Financials, CQS 2 +1.25%
  - Financials, CQS 3 +1.60%
  - Financials, CQS 4 +2.20%
  - Financials, CQS 5 +3.20%
  - Financials, CQS 6 +4.50%
  - Financials, unrated +2.40%
- Currencies and other shocks:
  - EUR external value -11.5%
  - Structured notes -8.0%
  - Collateralised securities -5.0%
  - Other investments, not classified elsewhere -5.0%

### Additional assumptions and sensitivity tests
- Volatility adjustment: Given the increase of credit spreads, the volatility adjustment increases following the Solvency II calculation method; for insurers using the volatility adjustment, the higher discount rate largely offsets the negative impact of the credit spread shock.
- Downgrades: To capture higher capital requirements from deterioration in corporate bond quality, it was assumed that a third of the bonds is downgraded by three notches.
- Investment funds: Stressed with the respective shock for the underlying asset class; for bond funds, maturities and ratings were assumed to resemble those of directly held bonds.
- Unit-linked liabilities: Decline in liabilities mirrored the market value loss of underlying assets.
- Sensitivity tests (single-factor) used to complement the adverse scenario; results of sensitivity analyses were not added to adverse scenario results.
- Specified single-factor sensitivity scenarios:
  - Interest rates: parallel upward and downward shift of the risk-free term structure (across all currencies) by 100 basis points.
  - Currencies: Increase and decrease of the Euro external value by 10 percent.
  - Equity: Decline of domestic and foreign equity prices by 40 percent.
  - Counterparty risk: Default of the largest banking counterparty with assumptions:
    - Equity exposures fully written off (100 percent haircut).
    - Loss given default (LGD) of 50 percent for unsecured bonds.
    - LGD of 15 percent for secured bonds.
    - LGD of 30 percent for other on-balance sheet exposures.

### Capital standard, data sources and modeling approach
- Solvency II implemented in the European Union in 2016; forms the basis for the insurance stress test. Solvency II principle: assets and liabilities are valued mark-to-market, with LTG and transitional measures allowed for the discount rate.
- Main output: effect on own funds eligible for coverage of the solvency capital requirement (SCR); SCR was partially recalculated after stress.
- QRTs used for TD solvency stress test:
  - Balance sheet (S.02.01)
  - Asset-by-asset investment holdings (S.06.02)
  - Derivative positions (S.08.01)
  - Cash-flow projections (S.13.01, S.18.01)
  - Impact of long-term guarantee measures and transitionals (S.22.01)
  - Own funds (S.23.01)
  - Calculation of the solvency capital requirement (S.25.01, S.25.02, S.25.03)
- Data limitations: some inconsistencies and gaps (notably derivative data); look-through to individual securities held by funds was not applied.
- Re-calculation of SCR after stress:
  - Market risk module: capital charges for equity risk, spread risk and property risk proportionately adjusted with changes in exposures.
  - Equity risk capital charge corrected for the symmetric equity adjustment which changes from +5.7 to -6.0 percentage points after the fall in equity prices in the adverse scenario.
  - Spread risk incorporates the downgrade of one-third of corporate bond holdings.
  - Life underwriting risk capital charge assumed to change proportionately with technical provisions after application of the stressed discount curve.
  - Other components of basic SCR (counterparty default risk, non-life underwriting risk, operational risk) assumed unchanged.
  - Internal model users: SCR calculations, aggregation and diversification effects approximated via a simplified approach building on the standard formula.
  - Loss-absorbing capacity of deferred taxes re-calculated based on modeled valuation losses in the excess of assets over liabilities.
- Reactive management actions and some risk mitigants (financial hedges, stop-loss arrangements, financial reinsurance) not modeled due to static balance sheet assumption and data granularity limits.

### Results of the solvency stress test (key findings)
- Valuation impacts (median and sample results):
  - Asset values decline by 10 percent for the median life insurer, and by 2 percent for the median non-life insurer.
  - Asset-liability ratio median changes:
    - Median life insurer: decline of 0.5 percentage points, down to 103.2 percent.
    - Median non-life insurer: ratio declines by 3.2 percentage points, to 113.3 percent.
    - Median reinsurer: ratio declines by 2.6 percentage points, to 123.9 percent.
- Drivers of valuation losses:
  - Most of the decline in asset values stems from higher corporate and sovereign spreads.
  - Equity and property shocks have a more muted impact.
  - Higher interest rates and Euro depreciation offset other adverse shocks for most insurers with a slight overhang of foreign-denominated assets.
  - For some companies with U.S. dollar financial accounts but material non-U.S. dollar positions, Euro depreciation tends to be negative.
- Solvency outcomes:
  - Non-life insurers affected most: coverage of the SCR decreases by 13 percentage points for the median non-life insurer.
  - Life sector: median post-stress SCR ratio is 6 percentage points higher than pre-stress (but this overstates the sectoral impact as the majority of life firms see SCR ratios decline).
  - A global -40 percent shock to equity prices (sensitivity) could bring a small number of life insurers close to, or below, 100 percent SCR coverage.
- Systemic assessment and recommended supervisory follow-up:
  - The stress test did not reveal systemic vulnerabilities under the tested scenario.
  - Microprudential follow-up recommended:
    - The vast majority of insurers remain well-capitalized.
    - One life insurer drops below the regulatory threshold (100 percent SCR coverage) with a shortfall in eligible own funds to meet the SCR amounting to less than €10m.
    - One reinsurer shows a post-stress SCR ratio close to the threshold; both firms had relatively low pre-stress SCR ratios and are subsidiaries of large internationally active insurance groups.
  - Recommendation: Central Bank to closely monitor risk exposures of these insurers and forthcoming recovery plans (submitted by several larger insurers as of April 2022). Use results of this ST and future TD ST exercises to validate insurers’ own adverse-scenario assumptions and recovery options.

### Sensitivity analyses (supplementary findings)
- Sensitivity analyses broadly confirm resilience but highlight vulnerability to equity price declines for a few life insurers:
  - Equity shock has greatest impact in the life sector; one life insurer could approach an SCR coverage of 100 percent under the equity sensitivity.
  - Non-life insurers and reinsurers are mostly immune to equity sensitivity given smaller exposures (though reinurers show wider dispersion depending on life vs non-life focus).
  - Lower interest rates (not part of the adverse macrofinancial scenario) could have a slightly beneficial effect for many insurers.

### Liquidity risk analysis and findings
- Central Bank liquidity analysis built on EIOPA 2021 bottom-up ST; EIOPA’s shock to cash flows included mass lapse, higher mortality and pandemic claims, and lower premiums; liquid assets shocked with an adverse market risk scenario.
- Central Bank extended EIOPA approach to a sample of ten insurers: four life insurers, three non-life insurers and three reinsurers.
- EIOPA 2021 conclusion: liquidity risks generally less of a concern than solvency risks given insurers’ large holdings of liquid assets, but unexpected cashflows may require liquidation of liquid investment assets in some cases.
- Central Bank findings:
  - Liquidity shock unlikely to have systemic impact in Irish insurance sector; results driven by company specifics.
  - Aggregate share of liquid assets is relatively high and remains stable in stressed situations.
  - Aggregated cash inflows increase in the adverse scenario (compared to baseline) as a result of effects on derivative-related collateral flows at some firms—this is not likely sector-wide given limited use of derivatives across the market.
  - Recommendation: Central Bank should follow up the exploratory 2021 exercise with a broader sample and different combinations of cash outflows and tighter market liquidity conditions to better understand liquidity flows, group structures and liquidity risk management.

*Source: IMF staff (Ireland IMF Financial Sector Assessment Program, Insurance Stress Test chapter).*

### 143. The FSAP carried out a TD analysis of potential liquidity risks from variation margin

### 1irlea2022013 - 143. The FSAP carried out a TD analysis of potential liquidity risks from variation margin

### Top-down analysis of variation margin for interest rate swaps
- Methodology: Used a top-down (TD) approach recently employed by EIOPA and in the 2021 U.K. FSAP.
- Main finding: Variation margin calls for interest rate swaps could be met, even when using very narrow definitions of liquid assets.

### Derivatives holdings and data quality
- Coverage in ST sample:
  - Asset-side derivatives market value: 0.8 percent of assets.
  - Liability-side derivatives: 0.9 percent of total liabilities.
- Composition by product type:
  - Life insurers: forwards 33 percent of notional; put options 17 percent; interest rate swaps 16 percent.
  - Non-life insurers: interest rate swaps 26 percent; other swaps and forwards also relevant.
- Data limitations:
  - Reporting of derivatives in QRTs is incomplete or inconsistent.
  - Detailed position data not available from all companies; quality differs across derivative types.
  - Suggestion: Cross-check QRT data with EMIR reporting to improve availability and enable further analysis.

### Sample and scope of the TD interest rate swaps analysis
- Sample size and coverage:
  - Limited data availability produced a sample of five insurers.
  - Market coverage of the sample: around 10 percent.
- Typical positions:
  - Life insurers typically are fixed-rate receivers in swaps (vulnerable to interest rate increases).
  - Non-life insurers’ swap positions can be fixed-receiving or fixed-paying.

### Stress scenarios modeled
- Interest rate shocks:
  - A 25 basis points (bps) increase occurs overnight (parallel shift across relevant currencies). For this overnight shock, only cash deposits (narrowest definition) are assumed available to meet margin calls.
  - 50 bps and 100 bps increases unfold over five days. In these scenarios, insurers may liquidate high-quality assets: cash deposits plus unencumbered sovereign bonds of credit quality steps 0 and 1, revalued after the interest rate increase and subject to an additional haircut.
- Rationale: Central counterparties often allow high-quality securities to meet margin calls; scenarios test both very narrow and slightly broader liquid asset definitions.

### Stress test outcomes for the five-insurer sample
- All five insurers could meet variation margin calls with cash equivalents only, even for a 100-basis points interest rate change.
- Average variation margin amounts (for insurers with a positive margin call):
  - Higher interest rate scenario (+100 bps): €96 million (55 percent of cash holdings).
  - Lower interest rate scenario (-100 bps): €35 million (27 percent of cash holdings).
- Impact when adding high-quality sovereign bonds (CQS0&1) to cash holdings:
  - Share of variation margin falls to 8 percent (higher rate scenario) and 2 percent (lower rate scenario).
- Encumbrance:
  - Encumbrance levels of high-quality sovereign bonds are less than 5 percent for insurers in the sample, providing buffers in liquidity stress environments.

### Policy implications and recommendations (liquidity)
- Improve data: Cross-check QRT derivatives reporting with EMIR reporting to enhance data completeness and consistency.
- Supervisory monitoring: Given reliance on cash and low encumbrance of high-quality sovereign bonds, monitor liquidity buffers and encumbrance levels.
- Consider central counterparties’ practices in supervisory assessments (use of high-quality securities to meet margin calls).

### Physical climate risk findings (brief)
- Exposure channel: Non-life underwriting domestically; major perils include windstorms, floods, and freezes; some firms underwrite globally (e.g., U.S. hurricanes, California wildfires).
- Catastrophe simulations:
  - Losses inflation-adjusted and scaled up by 150 percent.
  - Example: Ireland/U.K. windstorm — gross claims around €550 million; net claims after reinsurance around €140 million; average SCR ratios decline by less than 3 percentage points.
- Reinsurance:
  - Irish insurers use relatively high levels of reinsurance for extreme events, but rely largely on group-internal reinsurance, introducing counterparty concentration risk.
- Recommendations:
  - Central Bank follow-up supervision on catastrophe risk modeling, especially for flood where modeled occurrence probabilities vary significantly.
  - Monitor local protection gaps where flood insurance is unavailable or unaffordable.

### Transition climate risk findings (brief)
- Investment allocation:
  - About 8 percent of insurance investments allocated to sectors with carbon equivalents >1,000 gram per Euro of output.
  - “Electricity, gas, steam and air conditioning supply” ~4 percent of investments.
- Top-down “Minsky moment” re-pricing scenario (instantaneous pricing of an orderly NGFS 1.5°C path):
  - Median valuation impacts:
    - Equity investments decline by 3.8 percent.
    - Corporate bonds decline by 1.6 percent.
    - Investment funds decline by 1.4 percent.
  - By insurer type:
    - Median life insurer: combined asset value reduction of 2.7 percent.
    - Non-life and reinsurers: declines of less than 1 percent.
  - Absolute potential asset decline: around €7 billion (likely an upper bound given instantaneous shock assumption).
- Interpretation and caveats:
  - Impact larger for life insurers due to riskier asset allocation and fund choices by customers.
  - Considerable modeling uncertainties; risks could materialize earlier or differently than NGFS path implies.
- Policy implication:
  - Monitor asset exposures to high-carbon sectors and consider time-phased adjustments given modeling uncertainties.

*Source: IMF staff calculations and analysis based on Central Bank data and company submissions.*

### 162. The FSAP assessed the liquidity resilience of investment funds in Ireland through a

### 1irlea2022013 - 162. The FSAP assessed the liquidity resilience of investment funds in Ireland through a

### Stress-test objective and scope
- Objective: assess the ability of individual funds to withstand severe but plausible redemption shocks.
- Stress test is micro-prudential in nature and:
  - Does not consider the use of liquidity management tools (LMTs).
  - Does not estimate aggregate price effects of fire sales (left for future analysis).
- Noting systemic relevance: collective inability of funds to meet redemptions could lead to large-scale asset liquidations and potential fire sales with systemic impact.

### Key vulnerabilities and findings
- Majority of Irish-domiciled funds in the sample are resilient to severe but plausible redemption shocks.
- Pockets of vulnerability:
  - A significant share of funds investing primarily in high-yield bonds could experience liquidity shortfalls when faced with severe redemption shocks.
  - Emerging market (EM) focused funds are also susceptible in certain scenarios.
- Heterogeneity: funds with fewer liquid assets (e.g., HY, some EM) are the primary focus due to liquidity mismatch between highly liquid liabilities and less liquid assets.

### Potential spillovers
- Fire sales by funds meeting redemptions can affect other financial system segments via common asset holdings, potentially creating downward price spirals.
- Direct linkages:
  - Funds’ financial asset claims on Irish-domiciled OFIs amounted to around 21 percent of GDP in 2020Q4.
  - Real estate funds are a significant funding source for the domestic CRE property market; liquidity shocks to funds could pressure the CRE sector and cause losses for other financial institutions with larger CRE exposures.
- Policy implication: fund-level resilience assessment should be supplemented with monitoring of common asset holdings between funds and banks/non-banks to assess spillover risk.

### Liquidity-stress testing model (three-stage)
- Stage 1: Calibration of the redemption shock (at individual fund level or at fund category level).
- Stage 2: Calculation of liquidity buffers using a liquidity buckets approach informed by Basel III high-quality liquid assets (HQLA) weights.
  - Caveats of HQLA approach:
    - May overweight fund cash buffers (operational needs, margin calls).
    - May penalize funds with less liquid assets.
    - Ratings-based discounts and homogenizing of equities (discounted at 50 percent) may misstate true liquidity.
- Stage 3: Assessment of resilience using:
  - Redemption Coverage Ratio (RCR) = HQLA / redemption shock.
    - RCR > 1 indicates sufficient liquidity buffers.
    - RCR < 1 indicates potential need to sell less liquid assets, possibly at fire sale prices.
  - Liquidity shortfall = difference between HQLA and simulated redemption shock for funds with RCR < 1.
- Note: A macro-prudential step—analyzing market price impact of fund sales—was not performed due to data/strategy dependencies and is left for future work.

### Calibration of redemption shocks and scenarios
- Redemption shock defined as net outflows in percentage of total net assets and calibrated using a historical distribution approach.
- Two approaches:
  - Homogeneity: pooled category-level distribution; same shock applied to all funds within a category.
  - Heterogeneity: fund-specific shocks based on each fund’s historical net flows.
- Shock metrics:
  - Value-at-Risk (VaR) and Expected Shortfall (ES) models.
  - VaR: shock as a percentile (examples: 1st, 3rd, 5th percentiles).
  - ES: average net outflows below the VaR percentile to capture tail events.
- Combined design: VaR and ES at 1st, 3rd, and 5th percentiles × homogenous vs. heterogenous → total of twelve redemption shock scenarios.
- Tail-statistics construction notes:
  - Outlier monthly net flows above 50 percent of TNA are dropped from tail statistics.

### Data, sample, and coverage
- Data source: Morningstar commercial database with fund list provided by CBI (fund IDs and ISINs).
- Availability and filters:
  - 82 percent of CBI-provided funds were available in Morningstar, covering 86 percent of total AUM.
  - After requiring portfolio reporting date of January 1, 2021 or later: 3,289 investment funds available (around three quarters of AUM of Irish funds).
  - Dataset frequency: monthly, 2007-21.
  - Required variables: net flows, fund size (TNA), portfolio composition (percent cash, bonds, equities, other), credit quality of selected fixed-income instruments.
  - Selection for stress-test sample:
    - Focus on funds with primary focus on fixed-income or property.
    - Seven broader categories mapped from 89 Morningstar categories: HY, IG, EM, Sovereign, Mixed, Property, Other.
    - These seven categories comprised 929 funds active in 2021 (around one-fourth of Morningstar Irish AUM).
    - Final sample after additional filters (≥50 monthly flow observations; complete portfolio composition data): 274 Irish investment funds.

### Portfolio composition and liquidity buffers
- Portfolio breakdown: bonds (corporate, sovereign, securitized), cash, equities, other; bond credit ratings used to classify IG (AAA to BBB) vs. HY (below BBB).
- Liquidity weights: informed by Basel III HQLA liquidity weights.
- Category patterns:
  - HY funds: substantial portion in high-yield corporate bonds, low investment-grade and sovereign bonds, relatively limited cash buffers.
  - Sovereign funds: mostly sovereign bonds, low securitized and cash shares.
  - IG funds: high shares of investment-grade corporate and sovereign bonds.
  - EM, Other, Mixed: more diversified; mixed funds also invest in equities.
- HQLA buffers (by category):
  - HY funds: median HQLA buffer at below ten percent.
  - EM funds: median HQLA at just over 40 percent.
  - Sovereign funds: high liquidity buffers.
  - IG, Mixed, Other: moderate liquidity buffers.

### Model lineage and comparators
- Technical approach resembles micro-prudential aspects of recent IMF FSAP investment fund stress-testing frameworks.
- Closest to the 2020 U.S. FSAP micro-prudential model; other FSAPs cited with IF stress tests include Hong Kong (2021), Brazil (2018), Luxembourg (2017), Ireland (2016), and Sweden (2016).

*INTERNATIONAL MONETARY FUND*

### 182. Over the 2007-2021 sample period, HY, EM, and sovereign bond funds tend to face

### 1irlea2022013 - 182. Over the 2007-2021 sample period, HY, EM, and sovereign bond funds tend to face

### Liquidity resilience and Redemption Coverage Ratio (RCR)
- The Redemption Coverage Ratio (RCR) compares liquidity buffers (HQLA) to the size of the simulated redemption shock; RCR>=1 implies sufficient liquidity, RCR<1 implies the fund would need to sell less liquid assets, potentially at fire-sale prices.
- The benchmark stress scenario uses the ES shock model with the shock calibrated at the third percentile of net outflows (ES Model, 3% shock).
- Findings:
  - "Most HY funds (81 percent and 100 percent, under the heterogeneity and homogeneity approaches, respectively) would experience RCR <1 following a redemption shock."
  - Mixed funds show vulnerability only under the homogeneity assumption with 6 percent of funds experiencing RCR <1.
  - EM funds with RCR<1 constitute a small fraction by count but hold around a fifth of TNA within that category.
  - Sovereign and IG bond funds appear most resilient, followed by the "other" category, reflecting their significantly larger high-quality liquidity buffers.
  - Under the stricter criterion RCR<0.5:
    - Around 3 percent of EM funds would experience RCR <0.5.
    - Close to 60 and 80 percent of HY funds (under the heterogeneity and homogeneity approaches, respectively) would experience RCR <0.5.

### Liquidity shortfall
- Definition: liquidity shortfall (as percent of TNA) = HQLA minus the redemption shock for funds with RCR<1.
- Findings:
  - High-yield (HY) bond funds face the largest liquidity shortfall; a few HY funds experience shortfalls around 30 percent of TNA.
  - Emerging market fixed-income and "other" bond funds face liquidity shortfalls in certain scenarios but at a much lower scale than HY funds.

### Tools, stress-testing assumptions, and methodology notes
- For computational simplicity and conservative estimates, the stress test assumes:
  - Funds sell all HQLA at the hair-cut value as per Table 4.
  - Funds cannot sell other assets without a significant discount.
  - Funds cannot use credit lines or liquidity management tools.
- RCR results are presented under heterogeneity and homogeneity approaches; full set of twelve shock scenarios provided in Appendices XI and XII (not reproduced here).
- The liquidity shortfall and RCR figures are presented for heterogeneity and homogeneity approaches, using ES and VaR models with the redemption shock calibrated at the third percentile of net outflows as a percent of TNA.

### Conclusions from the Investment Fund Liquidity Stress Test Analysis
- The FSAP liquidity stress test of Irish-domiciled investment funds indicates:
  - High-yield bond funds are vulnerable to severe but plausible redemption shocks.
  - Emerging market fixed-income funds are also vulnerable in certain shock scenarios.

### Key policy recommendations and supervisory priorities
- CBI should prioritize the completion of its internal stress-testing model to enable more frequent liquidity and market risk stress tests for IFs and MMFs.
- CBI and CSO should coordinate to analyze common asset holdings of investment funds, banks, and relevant non-banks.
- As part of ongoing IF liquidity risk management policy development and considering EU and international developments, CBI should review the use of liquidity management tools by IFs, with particular focus on funds that have demonstrated liquidity challenges in recent periods of stress (notably HY bond funds and emerging market fixed-income funds).
- CBI should expedite completion of the internal stress-testing framework and enhance internal resources and capacity as appropriate.
- The stress testing exercise should ideally include an estimate of market price impact of fund liquidation and sales in episodes of market stress to analyze potential spillovers from the fund sector to other financial system segments.

### Housing market developments and housing price-at-risk (HaR)
- Recent house price developments:
  - March 2022 growth rate of 15 percent is cited as fueling continuous growth and gaps to the upside.
- Housing Price at Risk (HaR) estimates (staff exercise using three main drivers: domestic financial condition index (FCI), leverage indicator including credit-to-GDP gap, and household affordability indicator represented by house price to disposable income):
  - When shocking these risk factors by one standard deviation simultaneously, over a 3-year horizon:
    - Compounded growth rate at the 5-percentile of the housing price growth distribution is estimated to be -31 percent (using GDP input).
    - When replacing GDP with GNI* in computing house-price-related risk factors, the 5-percentile HaR becomes -43 percent cumulative growth over a three-year horizon.
  - One-year forward Housing Price Growth distribution: 5% HaR = -18%
  - Two-year forward Housing Price Growth distribution: 5% HaR = -12%
  - Three-year forward Housing Price Growth distribution: 5% HaR = -11%
- Implication: continuing upward trend in housing prices, if synchronized with pandemic resurgence or other global downside shocks, may undermine borrowers' affordability—particularly for lower-income tranches—and pose considerable risks to housing market and financial stability more broadly.

*Source: IMF staff calculations and analysis drawn from the provided IMF FSAP chapter text.*

### 194. The satellite models for credit risk were estimated using linear Bayesian Model

### 194. The satellite models for credit risk were estimated using linear Bayesian Model

### Bayesian Model Averaging (BMA) framework and model selection
- Satellite models for credit risk were estimated using linear Bayesian Model Averaging (BMA) framework to remove model uncertainty.
- BMA averages over the best models in the model class according to approximate posterior model probability to avoid over-confident inferences from a single model.
- The framework enables sign restriction in the estimation to ensure reasonable relationships among input variables and robustness of out-of-sample forecasts conditional on a constrained sample size.
- Model selection criteria used in the BMA:
  - R-square.
  - Durbin Watson statistics.
  - Number of significant variables with high posterior inclusion probability.
  - Quality of in-sample forecast (small root-mean-square-error).
  - Size of the impact in the forecasting horizon (historically coherent size of impact).
- BMA allows users to select different model specifications (number of autoregressive lags, number of explanatory variables under permutation, number of lags for each explanatory variable).

### Key drivers of credit risk and quantitative estimates (Table 1: Probability of Default and Coverage Ratio)
- Real output, unemployment rate, short-term and long-term interest rates, and asset prices were relevant for the buildup of credit risk, reflected in higher-than-prior posterior inclusion probabilities and sizable long-run multiplier estimates for sectoral PDs.
- R-square by series (in unit, logit transformed):
  - PD Household retail: 0.99
  - PD Household mortgage: 0.97
  - PD Corporate non-CRE: 0.96
  - PD Corporate-CRE: 0.97
  - PD Financial: 0.75
  - Coverage ratio: 0.98
- Number of lags of independent variables by series:
  - PD Household retail: 2
  - PD Household mortgage: 2
  - PD Corporate non-CRE: 3
  - PD Corporate-CRE: 3
  - PD Financial: 2
  - Coverage ratio: 2
- Selected long run multipliers and posterior inclusion signals (asterisk * denotes higher posterior inclusion probability than the prior):
  - Real GNI* growth, 4-quarter moving sum, percent, y-o-y: PD Corporate non-CRE -0.069*; PD Corporate-CRE -0.042*; PD Financial -0.052
  - Unemployment rate, percent: PD Household retail 0.086*; PD Household mortgage 0.159*; PD Corporate non-CRE 0.069*; PD Corporate-CRE 0.181*; PD Financial 0.119*; Coverage ratio 1.85*
  - Three-month interbank rate, percent: PD Household retail 0.123*; PD Household mortgage 0.024*; PD Corporate non-CRE 0.257*; PD Corporate-CRE 0.645*; Coverage ratio 0.363
  - Ten-year government bond yield, percent: PD Household retail 0.062*; PD Household mortgage 0.015; PD Corporate non-CRE 0.065*; Coverage ratio 2.056*
  - Stock price growth, percent, y-o-y: PD Household mortgage -0.006*; PD Corporate-CRE -0.012* 
  - House price growth, percent, y-o-y: PD Household mortgage -0.011*; PD Corporate non-CRE -0.024*; Coverage ratio -0.089
  - CRE price growth, percent, y-o-y: PD Household retail -0.01*; PD Corporate-CRE -0.017*; Coverage ratio -0.023
- Foreign sector long run multipliers and significance (foreign corporate portfolios, in unit logit transformed):
  - UK Unemployment rate, percent: 0.272*
  - UK Three-month interbank rate, percent: 0.107*
  - UK Ten-year government bond yield, percent: 0.642*
  - UK Exchange rate, Sterling per USD: 9.799*
  - US Real GNI* growth, 4-quarter moving sum, percent, y-o-y: -0.039*
  - US Unemployment rate, percent: 0.013*
  - US Ten-year government bond yield, percent: 0.046*
  - US Stock price growth, percent, y-o-y: -0.011*
  - EA Real GNI* growth, 4-quarter moving sum, percent, y-o-y: 0.264*
  - EA Ten-year government bond yield, percent: 0.499*
  - EA Exchange rate, EUR per USD: 4.698*
  - EA Stock price growth, percent, y-o-y: -0.02*
  - ROW Short term interest rate, percent: 0.355*
  - ROW GDP gap, percent: -0.343*
- R-square for foreign corporate portfolios:
  - UK/US/EA/ROW combined table: 0.90, 0.67, 0.93, 0.66 (as presented under "R-square" for the foreign corporate portfolios).

*The model was developed by Marco Gross and Javier Población (2017).*

### Interest rate satellite models (Appendix IV) and bank interest rate estimation
- Bank funding costs and lending rates estimated based on interest rate for new business (front book) for each bank using quarterly data from 2000 Q1.
- Data sources: ECB statistical warehouse, Refinitiv Datastream, IMF International Financial Statistics, IMF WEO database, Bloomberg, Haver Analytics.
- Satellite models (using BMA) estimate aggregate funding and lending rates on portfolio level, including:
  - interest rates on corporate, household retail and household mortgage loans,
  - coupon rates for sovereign, corporate and bank bonds,
  - overnight and term deposits.
- Period changes of aggregate rates mapped with outstanding sensitive assets and liabilities reported in the IRRBB template to derive impact on net interest margins.
- Main inputs for explaining and projecting bank interest rates: Euro Area interbank rate (EURIBOR) and Ireland long-term bond yield.

Key quantitative results (Table 2: Satellite Model Estimation – Bank Interest Rates; In percent):
- R-square by rate:
  - Corporate loans: 0.93
  - Household retail loans: 0.83
  - Household mortgage loans: 0.80
  - Overnight deposits: 0.92
  - Term deposits: 0.99
  - Debt securities - sovereign: 0.31
  - Debt securities - corporate: 0.81
  - Debt securities - financial: 0.81
- Number of lags of independent variables by rate:
  - Corporate loans: 2
  - Household retail loans: 2
  - Household mortgage loans: 2
  - Overnight deposits: 2
  - Term deposits: 2
  - Debt securities - sovereign: 3
  - Debt securities - corporate: 3
  - Debt securities - financial: 3
- Selected long run multipliers and posterior inclusion signals:
  - Three-month interbank rate, percent:
    - Corporate loans: 0.633*
    - Household retail loans: 0.031
    - Household mortgage loans: 0.433*
    - Overnight deposits: 0.193*
    - Term deposits: 0.731*
    - Debt securities - corporate: 0.393*
    - Debt securities - financial: 0.381*
  - Ten-year government bond yield, percent:
    - Household retail loans: 0.388*
    - Term deposits: 0.005
    - Debt securities - sovereign: 0.032*
    - Debt securities - corporate: 0.136*
    - Debt securities - financial: 0.221*
    - Debt securities - corporate (alternate column): 0.187*
  - Real GNI* growth, 4-quarter moving sum, percent, y-o-y:
    - Household mortgage loans: -0.020
    - Household mortgage loans (alternate): -0.013*
    - Overnight deposits: -0.045*
    - Term deposits: -0.000

- Interpretation:
  - Lending rates highly correlated with short-term interest rate, reflecting dominance of floating rate loans.
  - Cost of overnight deposits largely determined by short-term rates.
  - Term deposits and bond interest rates driven by both short-term and long-term interest rates.
  - Long-run pass-through from Irish sovereign bond yield and short-term interest rate on funding rates is large, particularly to term deposit rate and corporate debt securities.

### Physical risk shock calibration (Appendix V)
- EC JRC PESETA IV project estimates for Europe (three warming levels vs pre-industrial):
  - 3°C: annual welfare loss of 0.6 percent of GDP for Ireland and United Kingdom.
  - 2°C: welfare loss of 0.2 percent of GDP.
  - 1.5°C: welfare loss of 0.1 percent of GDP.
- Dottori and others (2018) river floods estimates:
  - With 1.5 °C, direct flood damage increases by 160–240 percent, with a relative welfare reduction between 0.23 and 0.29 percent.
  - In a 2 °C world direct economic damage doubles and welfare losses grow to 0.4 percent.
- IPCC and other studies indicate upper-bound global exposure:
  - Possible damage from exposure to extreme sea levels could amount to 10 percent of global GDP by the end of the century in the absence of adaptation.
  - Kirezci et al. (2020) indicated coastal flooding could potentially compromise 20 percent of the global GDP within the century (upper bound).
- NGFS/NiGEM hot house world scenario (June 2021 publication):
  - Peak of 4.8 percent deviation of the output level below the baseline by 2050.
- Historical extreme-weather event examples (first-year shock on GDP):
  - Flash floods in Spain in 1983: 1.2 percent of GDP damage.
  - River floods in France in 1966: 0.6 percent of GDP damage.
  - 2021 floods in Germany: current estimates around 0.1-0.2 percent of GDP.
- Combined suggested range for the physical shock in Ireland:
  - 2 to 20 percent of GDP, equivalent to 0.5 to 5 percent losses of capital stocks and 0.4 to 4 percent of output losses.

### Banking sector stress testing framework and key parameters (Top-down by IMF)
- Institutional perimeter and coverage:
  - Top-Down by FSAP team covering 12 banks: SIs (7 banks) and LSIs (5 banks).
  - 5 domestically focused retail banks and 7 internationally oriented banks (subsidiaries of foreign parents).
  - LSIs subject to sensitivity analysis only.
  - Total coverage about 80 percent of the banking sector, with 73 percent for SIs and 7 percent for LSIs.
- Data and baseline dates:
  - Multiple data vintages: December 2019, December 2020, June 2021.
  - Supervisory data (FINREP, COREP), IRRBB, liquidity and market risk sensitivities (including STE templates).
  - Expected Default Frequency sourced from Moody’s.
  - Market and publicly available data (ECB statistical data warehouse, etc.).
  - Data on policy mitigation impact on banking sectors for COVID-19 (moratoria, public guarantees, liquidity support).
  - Scope: consolidated banking group for banks headquartered in Ireland; foreign subsidiaries assessed unconsolidated covering domestic activities only.
  - Coverage of sovereign and non-sovereign securities exposures: FVPL, FVOCI, and amortized cost (AC).
  - Coverage of lending exposures by ten asset classes including geographic breakdowns and sectoral differentiation for domestic nonfinancial corporations.
- Channels of risk propagation and methodology:
  - FSAP team satellite models and methodologies; balance-sheet regulatory approach.
  - Market risk treated as an add-on, affecting capital resources (P&L or OCI) and capital requirements (RWA).
  - Traded risk impact via revaluation of FVPL and FVOCI securities by counterparty and interpolated credit spreads by residual maturity and issuer.
  - Losses for securities portfolios based on modified duration approach; equity losses based on specified stock market moves.
  - For IRB exposures: projection of PiT and TTC PDs, LGD, EAD, and RWA.
  - For standardized exposures: projection of new flows of defaulted exposures, risk weight downgrades, and coverage ratios for defaulted loans.
  - Provisioning modeled using IFRS9 transition matrix approach (using CBI submissions and COREP data).
  - Funding costs projected at portfolio level using funding structure by product and maturity bucket; lending rates projected at system level and attached to bank-specific effective interest rates.
- Stress test horizon:
  - 2021 Q2–2026 Q2 (5 years)
- Scenarios:
  - Baseline scenario: March 2022 WEO macroeconomic projections.
  - Adverse scenario: captures key risks in the RAM; relies on Global Macro-financial Model (GFM) disaggregated into forty national economies (documented in Vitek (2018)); foreign country scenarios extracted from GFM.
- Sensitivity analyses:
  - Single-factor sensitivity tests for other international banks (4 banks) imposing:
    - Lower bound (10 percentile) of historical distribution of net interest margin, net trading income ratio, net fees and commission income ratio.
    - Upper bound (90 percentile) of net loan loss ratio and loss ratio from off-balance sheet exposure with 50 percent conversion rate to on-balance-sheet exposure.
  - Single-factor sensitivity tests for concentration risk where banks’ top 3 to 5 exposures are assumed to fail.
  - Sensitivity analysis on effect of policy mitigation under Covid-19 (e.g., unwinding of payment breaks).
- Risks and buffers covered:
  - Credit (loans and debt securities), market (valuation and credit spread risk, P&L impact of net open positions), and interest rate risk (IRRBB) on the banking book.
  - Concentration risk via sensitivity analysis.
  - Solvency and liquidity risk interactions (mainly through asset haircut).
- Behavioral and modeling assumptions:
  - Quasi-static approach for balance sheet growth: asset allocation and funding composition remain the same; balance sheet grows in line with nominal GDP paths of major geographical exposures; reduced credit demand in material jurisdictions and FX revaluation effects considered.
  - Floor on balance sheet growth set at zero percent to prevent deleveraging (constraint binding in adverse scenario).
  - RWAs projection:
    - Standardized portfolios: RWAs change due to balance sheet growth, new inflows of NPLs, new provisions, exchange rate movements, and conversion of off-balance sheet items.
    - IRB portfolios: through-the-cycle PDs, downturn LGDs, and EAD used to project risk weights.
  - Interest income from non-performing loan is not accrued.
  - Banks assumed not to issue new shares or make repurchases during stress test horizon.
  - Dividends assumed paid at 25 percent of current period net income after taxes (only if net income is positive) by banks in compliance with supervisory capital requirements.

*Source: Ireland: IMF FSAP chapter extract as provided.*

### 5. Regulatory and

### 5. Regulatory and Market-Based Standards and Parameters

### National regulatory framework for capital adequacy
- Baseline regulatory minima and benchmarks assessed in stress tests:
  - CET1: 4.5 percent.
  - Total capital adequacy ratio: 8 percent.
  - Tier 1 capital ratio: 6 percent.
  - Leverage ratio: 3 percent Basel III minimum requirement.
- Baseline scenario is subject to bank specific Pillar 2 requirement.
- Hurdle rates for CET1, T1 and total capital adequacy include any requirements due to systemic buffers for other systemically important institution (O-SII), and do not include capital conservation and capital countercyclical buffers.
- Banks that end the stress test horizon with a capital level or a leverage ratio below the relevant hurdle rates are considered to have failed the test.

### Reporting format and outputs for solvency stress tests
- Outputs potentially include:
  - Evolution of capital ratios for the system as a whole and as groups of retail banks and large international banks.
  - Impact decomposition by result drivers (profit components, losses due to realization of different risk factors).
  - Capital shortfall as sum of individual shortfalls; reported in euros and in percent of nominal annual GDP.
  - Number of banks and corresponding percentage of assets below the regulatory minimum (or below the minimum leverage ratio).

### Banking sector: Liquidity stress test — regulatory standards and parameters
- Institutions included: 12 banks subcategorized as SIs (7 banks) and LSIs (5 banks). Among the total, 5 are domestically focused retail banks and 7 are internationally oriented banks which are subsidiaries of foreign parents.
- Market share: Total coverage is about 80 percent of the banking sector, with 73 percent for SIs and 7 percent for LSIs.
- Data and baseline date:
  - Latest data: June 2021.
  - Source: supervisory data (LCR, NSFR and ALMM Maturity Ladder template).
  - Scope of consolidation: banking activities of the consolidated banking group for banks having their headquarters in Ireland. Foreign subsidiaries are assessed on the unconsolidated level covering domestic activities only.
- Methodology:
  - Basel III LCR and cash-flow based liquidity stress test using maturity buckets by banks, incorporating both contractual and behavioral (where available) with assumption about combined interaction of funding and market liquidity and different level of central bank support.
  - Liquidity test in total currency, USD, and Sterling.
- Risks and buffers:
  - Risks: Funding liquidity; Market liquidity.
  - Buffers: The counterbalancing capacity, including liquidity obtained from markets and/or the central bank’s facilities. Expected cash inflows are included.
- Tail shocks — run-off calibrations and assumptions:
  - Run-off calibration reflects system-wide deposit runs and dry-up of unsecured wholesale and retail funding, with additional run-off for non-resident deposits calibrated following historical events, recent international experience and IMF expert judgment.
  - Retail scenario key assumptions:
    - (i) 10 percent run-off rates for stable retail deposits and 20 percent for less stable retail;
    - (ii) 5-25 percent for operational deposits and 20-40 percent for non-operational deposits;
    - (iii) no changes in liquid assets weights.
  - Wholesale scenario key assumptions:
    - (i) 5 percent run-off rates for stable retail deposits and 15 percent for less stable retail;
    - (ii) 15-3 5 percent for operational deposits and 30-50 percent for non-operational deposits;
    - ((iii) no changes in liquid assets weights.
  - Combined run-off and price shock scenario key assumptions:
    - (i) 10 percent run-off rates for stable retail deposits and 20 percent for less stable retail;
    - (ii) 15-35 percent for operational deposits and 30-50 percent for non-operational deposits;
    - ((iii) liquid assets weights reduction of 0-5 percent for level 1 assets, 3-20 for level 1 covered bonds, 5-15 percent for level 2A assets and 5-25 for level 2B assets.
  - Simulation horizons:
    - 1–month for LCR.
    - 5 days, 1 month, 3 months and 1 year for cash-flow based approach.
  - Haircuts of HQLA calibrated against ECB haircuts, past Euro Area FSAPs, and market shock for investment securities and money market instruments in the solvency stress test.
- Regulatory standards:
  - Consistent with Basel III regulatory framework (LCR).
  - Liquidity shortfall by bank.
- Reporting format for liquidity stress test results:
  - Liquidity ratio or shortfall by groups of banks and aggregated (system wide).
  - Number of banks that still can meet or fail their obligations.

### Contagion analysis — regulatory thresholds and outputs
- Institutional perimeter:
  - Domestic interbank contagion: 12 banks (SIs 7, LSIs 5); 5 domestically focused retail banks; 7 internationally oriented banks (subsidiaries).
  - Cross-border contagion: country-pair bilateral exposure across Ireland, rest of Euro Area countries, United Kingdom, and United States.
  - Cross-sectoral contagion: bilateral exposure across Irish banks and top 10 nonbank financial institutions (NBFIs) in terms of exposure size.
- Data and baseline date:
  - Latest data: Supervisory as of June 2021 (and to the extent possible December 2021).
  - BIS consolidated banking statistics.
- Methodology:
  - Balance-sheet model: Interbank and cross-border network model by Espinosa-Vega and Solé (2010).
- Tail shocks and default threshold:
  - Pure contagion: hypothetical default of institutions.
  - Default threshold: banks would default if their total CET1 ratios falling below 4.5 percent.
- Reporting format for contagion results:
  - Capital shortfall systemwide, by bank and by group: contagion and vulnerability scores.
  - Direction and size of spillovers within the network.

### Banking sector: Climate risk analysis — parameters and reporting
- Institutional perimeter and data:
  - 12 banks subcategorized as SIs (7 banks) and LSIs (5 banks). Market share: total coverage about 80 percent (SIs 73 percent, LSIs 7 percent).
  - Supervisory data as of June 2021; public data from 2003 to 2020 from capital IQ, Moody’s Analytics and Eurostat.
- Methodologies:
  - Transition risk:
    - Single factor sensitivity analysis to assess near-term impact on corporate credit quality from a rising carbon tax, allowing sectoral differentiation.
    - Bank credit impairment generated by applying changes in sectoral PDs from entire firm sample to bank corporate loan PDs.
    - Market losses from bank holdings of mark-to-market debt securities estimated using a duration approach relying on estimated PDs and Merton theory.
  - Physical risk:
    - Scenario-based analysis simulating macroeconomic impact of a severe flooding event, translated into bank losses through credit, market, and interest rate risk channels.
- Firm behavioral response:
  - Firms allowed to pass through partial or full cost of carbon tax to consumers through increase in prices; corresponding drop in demand incorporated based on pre-determined price elasticity.
- Tail shock:
  - Increase of carbon tax from €33.50 to €100 per ton, based on CBI national targets and NGFS scenarios.
  - Impact assessed from 1-year to 5-year horizon, assuming shock materializes immediately in the first year.
- Regulatory and market-based standards:
  - No capital thresholds are applied.
- Reporting format:
  - Change in corporate PDs by sector with and without firm behavioral response from 1-year to 5-year horizon.
  - Bank credit impairment by sectors due to shock on PDs on corporate loan, from 1-year to 5-year horizon.
  - Bank market losses on holdings of debt securities due to shock on credit spread induced from corporate PDs, from 1-year to 5-year horizon.
  - Bank capital ratio impact from 1-year to 5-year horizon.

### Insurance sector: Solvency risk — regulatory standards, scenarios, and outputs
- Institutional perimeter and market share:
  - 10 life insurers, 8 non-life insurers, 7 reinsurers listed.
  - Market share: Life: >70 percent; Non-life: >70 percent; Reinsurance: >70 percent (gross premiums written, total and domestic business, 2020 market shares).
  - Consolidation: Solo-entity level.
  - Data: Regulatory reporting. Reference date: June 30, 2021.
- Methodology and time horizon:
  - Investment assets: market value changes after price shocks affecting solvency.
  - Insurance liabilities: impact on best estimate by changing discount rate; proportionate change for the risk margin.
  - Required capital after stress approximated by the Solvency II standard formula also for internal model users.
  - Time horizon: Instantaneous shock.
- Adverse scenario (aligned with banking sector stress severity) — key shocks:
  - Risk-free interest rates (without volatility adjustment):
    - -17 bps (1yr EUR), -49 bps (10yr EUR);
    - -17 bps (1yr USD), -48 bps (10yr USD);
    - -17 bps (1yr GBP), -49 bps (10yr GBP).
  - Sovereign bond spread: +160 bps (domestic), +25 bps for other low-yield advanced economies, up to +180 bps for emerging and developing economies.
  - Stock prices:
    - -58.0 percent (domestic),
    - -18.0 percent (Euro Area and United States),
    - -20.0 percent (other advanced economies),
    - -35.0 percent (emerging and developing economies).
  - Property prices:
    - -19.9 percent (domestic, residential),
    - -34.1 percent (domestic, commercial),
    - -5.0 percent (foreign, residential),
    - -18.0 percent (foreign, commercial).
  - Corporate bond spreads:
    - between +60 bps (AAA, non-financials) and +420 bps (CCC and lower, non-financials),
    - between +75 bps (AAA, financials) and +450 bps (CCC and lower, financials).
  - Rating downgrades: one category (3 notches) for one third of the corporate bond portfolio.
  - EUR external value: -11.8 percent.
- Sensitivity analyses:
  - Risk-free interest rates +/-100 bps (all currencies).
  - EUR external value +/-10 percent.
  - Stock prices -40 percent.
  - Default of largest banking counterparty.
- Risks and buffers assessed:
  - Market risks: interest rates, share prices, property prices, credit spreads, currency.
  - Credit risks: default of largest financial counterparty.
  - Summation of risks, no diversification effects.
  - Buffers: Solvency II long-term guarantee measures and transitionals including Volatility Adjustment (VA); unit-linked life insurance losses borne by policyholders.
- Regulatory/accounting standards:
  - Solvency II.
  - National GAAP.
- Reporting format for solvency results:
  - Impact on valuation of assets and liabilities.
  - Impact on solvency ratios (including and excluding long-term guarantee measures and transitionals).
  - Contribution of individual shocks to changes of eligible own funds.
  - Dispersion measures of solvency ratios.
  - Capital shortfall and possible de-risking of investment assets to re-establish full coverage of solvency requirements.

### Insurance sector: Liquidity risk — methods and outputs
- Institutional perimeter:
  - Bottom-up by CBI and top-down by IMF; overlapping samples of insurers and reinsurers.
  - Market share: Life: 20 percent (gross premiums written); Non-life: 45 percent (gross premiums written); >70 percent (gross premiums written, total and domestic business, 2020 market shares).
  - Data: Regulatory reporting. Reference dates: December 31, 2020 and June 30, 2021.
- Methodology:
  - Stock/flow assessment of liquidity sources and needs.
  - Shock to cash flows based on EIOPA’s 2021 adverse scenario (market and insurance risks).
  - Reduction in value of liquid assets based on EIOPA’s 2021 adverse scenario.
  - Revaluation of derivative positions after interest rate shock.
- Time horizon:
  - Bottom-up: 90 days.
  - Top-down: Instantaneous (1 day, 5 days).
- Tail shocks and sensitivity:
  - Scenario analysis: EIOPA 2021 adverse scenario.
  - Sensitivity: Parallel shift of the interest rate term structure (for all currencies): +25 bps, +50 bps, +100 bps.
- Risks and buffers:
  - Liquidity risk channels assessed: shock to market value of assets, mass lapse shock, mortality shock, pandemic morbidity shock, increase of non-life cost of claims, shock to reinsurance inflows, reduction in written premiums, and margin calls for interest rate swaps.
  - Buffers: None reported.
- Reporting format for liquidity results:
  - “Sustainability indicator”: Net flows divided by liquid assets.
  - Total amount of variation margin calls.
  - Variation margin as percent of cash holdings.
  - Variation margin as percent of high-quality liquid assets.

### Investment fund sector: Liquidity risk — parameters and outputs
- Institutional perimeter:
  - Fixed-income bond funds.
  - Data source: Morningstar.
  - Reference date: Portfolio reporting date: January 1, 2021, or later.
- Methodology:
  - Various levels of redemption shocks compared to level of highly liquid assets at the fund level.
  - Redemption shocks calculated based on historical data on redemptions using VaR and Expected Shortfall methodologies with multiple thresholds.
  - Historical time series with monthly frequency.
- Time horizon: Instantaneous shocks.
- Scenario analysis:
  - Pure redemption shock: severe outflows based on historical distribution.
- Sensitivity analyses:
  - Risk-free interest rates +/-100 bps (all currencies).
  - EUR external value +/-10 percent.
  - Stock prices -40 percent.
  - Default of largest banking counterparty.
- Risks and buffers:
  - Liquidity risk: severe redemption shock.
  - Buffer: Stock of highly liquid assets.
- Reporting format for fund liquidity results:
  - Number of funds with a redemption coverage ratio (ratio of highly liquid assets to redemptions) below one.
  - Liquidity shortfall amount for individual funds after redemptions.

*Source: IMF staff summary of stress testing matrices, supervisory data and scenario parameters as presented in the document.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1irlea2022013.pdf_