## 1dnkea2020009

## Source details

**Canonical URL:** [1dnkea2020009](https://www.imf.org/-/media/files/publications/cr/2020/english/1dnkea2020009.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2020/english/1dnkea2020009.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2020/english/1dnkea2020009.pdf.json)

---

### Executive summary: context and overall stability assessment
- FSAP work conducted prior to the COVID-19 pandemic; report includes stability analysis and stress tests under updated illustrative scenarios to quantify possible implications of the COVID-19 shock on bank solvency.
- Elevated pre-pandemic vulnerabilities:
  - High household leverage amid high real estate valuations.
  - Signs of risk taking in some sectors, such as commercial real estate (CRE).
  - Downside risks to bank profitability amid the low-interest-rate environment.
- Caution: an unusually high degree of caution is required in interpreting stress test results due to heightened uncertainty around post COVID central projections and downside risks.

### Macroeconomic scenarios used in solvency tests
- Three illustrative five-year scenarios for top-down solvency tests of largest SIFIs:
  - "COVID central": GDP falls by 8 percent in 2020, rebounds by 6 percent in 2021; focuses on growth and unemployment implications; does not entail abrupt tightening of financial conditions or a real estate downturn.
  - "COVID prolonged": Growth declines by 8 percent in 2020 and economy continues to contract (example path: COVID Prolonged real GDP growth: 1.7, -8.0, -2.1, 3.3, 3.3, 3.3); assumes no sharp tightening in financial conditions or decline in house prices.
  - "Market shocks" (adverse): sharp rise in asset risk premia combined with a prolonged downturn in the domestic real estate market (housing prices path: Market Shocks: 100.0, 91.2, 82.4, 77.5, 79.1, 80.6; long-term funding cost shocks larger).
- Selected scenario parameters (examples preserved exactly):
  - COVID Central real GDP growth (2019–2024): 1.7, -8.0, 6.0, 3.5, 2.0, 1.8.
  - Harmonized unemployment rate (COVID Central 2019–2024): 5.0, 6.8, 6.2, 5.7, 5.6, 5.6.
  - Nominal policy interest rate (All Scenarios, 2019–2024): -0.8, -0.9, -0.9, -0.8, -0.6, -0.5.
  - Housing prices (2019=100): Both COVID Scenarios: 100.0, 101.5, 103.3, 105.4, 107.5, 109.7; Market Shocks: 100.0, 91.2, 82.4, 77.5, 79.1, 80.6.

### Group-level solvency stress test results (banks and MCIs)
- Aggregate capital and leverage outcomes (starting baseline 2019):
  - Baseline (2019): aggregate CET1 ratio = 18.3 percent; aggregate leverage ratio = 4.8 percent.
  - COVID baseline (low point in 2020): aggregate CET1 ratio = 13.4 percent; aggregate leverage ratio = 3.6 percent.
  - COVID prolonged (low point in 2022): aggregate CET1 ratio = 12.4 percent.
  - COVID prolonged (aggregate leverage low point in 2024): aggregate leverage ratio = 3.0 percent.
- Key group outcomes and caveats:
  - All SIFIs would meet their minimum capital requirements (a 4.5 percent CET1 ratio) under the scenarios modeled.
  - Impact differentiated across SIFIs; a few SIFIs need to partially use capital conservation and/or SIFI buffers.
  - Leverage ratios of a few SIFIs fall below three percent.
  - Uncertainty caveat: further deterioration relative to assumed macro outcomes could lead some SIFIs to breach minimum capital requirements.

### Main drivers of losses and decomposition
- Principal drivers across scenarios:
  - Credit risk and funding-cost shocks are the principal drivers.
  - COVID scenarios: losses mostly driven by materialization of credit risk, predominantly influenced by GDP growth and the unemployment rate.
  - Market shocks scenario: cost of funding shocks also contribute materially to losses; market repricing losses significant.
- Decomposition approach:
  - Change in CAR decomposed into contributions from capital changes (pre-impairment profit, ΔNII, Δnet gains on securities, credit losses, other items minus taxes and dividends) and RWA changes (Appendix V formulas preserved).

### Liquidity stress testing (pre-COVID) and MCIs liquidity reforms
- Top-down bank liquidity tests (maturity-ladder and LCR analysis) run prior to the pandemic:
  - Results: all banks would be able to survive extreme funding shocks for 3 months or longer without external support; three of the 5 banks: survival 5 months or longer.
  - LCR outcomes: All-currencies LCR: 3 of the 5 banks post-shock at about 100 percent or above; two others remain above 65 percent. EUR LCR: 3 of 5 close to or above 100 percent; two in 80-100 percent range.
  - Pre-COVID test shocks were "significantly larger than those that have materialized in 2020 Q1."
- MCIs liquidity and new Pillar II add-ons:
  - Structural protections: over-collateralization, matching principle, limited funding liquidity exposure.
  - Pillar II add-ons to replace LCR floor requirement:
    - Pillar II A add-ons: at least two—one for "unknown arrears" and one for "unknown open conversions" (open conversions add-on calibrated to at least 30 pct. of open conversions in an unlikely stress).
    - Pillar II B add-on: for "refinancing risk" (SMCIs assume they will have to buy 30 pct. of their own covered bonds to avoid an auction failing).
  - Observation period: Pillar II add-ons reported from December 31th 2019 until entry into force; LCR floor remains during observation.

### Insurance and pension fund stress testing (pre-COVID)
- Scope and sample:
  - Top-down and bottom-up solvency stress tests covered five life insurers, six occupational pension funds, and five non-life insurers, accounting for more than 70 percent in each sector; ATP included.
  - Reference date: 30 June 2019.
- Key stress outcomes (market shocks adverse scenario):
  - Life insurers: median SCR ratio drops from 188 percent to 127 percent.
  - Non-life sector: median SCR ratio declines from 233 percent to 207 percent.
  - One life company falls below a solvency ratio of 100 percent; capital needed to restore full coverage would be a moderate fraction of the sample’s aggregate capital.
- Drivers and sensitivities:
  - Most solvency impact attributed to lower stock prices and increases in domestic sovereign and covered bond yields; increased corporate bond spreads.
  - Sensitivity tests: longevity risk (permanent 25 percent decrease in mortality) and climate-related catastrophic events assessed.
  - ATP: resilient but first-year RoE after stress for most life insurers and pension funds between -5 and -25 percent; gradual capital rebuild thereafter.

### Non-financial corporate (NFC) sector vulnerability analysis
- Coverage and findings:
  - Micro-data analysis (2003–2017) used to estimate drivers of firm failures and correlated defaults.
  - Total NFC debt in 2017: 1,400 billion DKK.
  - Sector concentration: CRE, manufacturing industries and trade account for 70 percent of total NFC debt; CRE represents 30 percent of total NFC debt.
  - Debt-at-risk analysis: pre-COVID adverse scenario projects debt-at-risk rising from 5.3 percent to 13.1 percent of NFC debt by 2022; commercial real estate shows particularly large sensitivity (six-fold jump projected in adverse scenario).
- Firm-level statistics (selected median series preserved exactly):
  - Return on equity (All firms median): 2007: 15.9; 2009: 4.4; 2011: 10.4; 2013: 9.9; 2015: 11.5; 2017: 11.5.
  - Liquidity ratio (All firms median): 2007: 1.23; 2009: 1.19; 2011: 1.21; 2013: 1.23; 2015: 1.27; 2017: 1.32.
  - Debt-to-equity ratio (All firms median): 2007: 1.97; 2009: 1.70; 2011: 1.71; 2013: 1.73; 2015: 1.70; 2017: 1.59.
- Correlated-default simulations:
  - Simulations (10,000 iterations) demonstrate that stressed default correlations can substantially increase portfolio losses even when PDs remain low.

### Key methodological and modelling features
- Credit loss satellite models:
  - Loan loss rates estimated via dynamic panel/logit regressions with regressors including real GDP growth and unemployment; sector intercepts adjusted.
  - Exact estimated coefficients (logit specification): Impairment rate (logit transform, 1 year lag): 0.193; Unemployment Rate: -0.060; Real GDP growth rate: -0.321; Short-Term deposit rate: -0.050; Unemployment Rate (1 year lag): 0.292; Real GDP growth rate (1 year lag): -0.094; Short-Term deposit rate (1 year lag): 0.118.
- Net interest income (NII) projection:
  - Highly granular GAP analysis using banks’ reported volumes and "new business" effective interest rates for 2019; stress tests assume constant balance sheet with maturing loans renewed at new business rates.
  - Example new business observation: 2019 new business lending rates ≈ 0.1 percent lower than overall average lending rate.
- Market valuation losses:
  - Measured via modified-duration approach using yield curve shifts from scenario inputs; applied across accounting categories excluding floating rate securities.
- RWAs and Basel III output floor sensitivity:
  - Industry estimate: RWAs will increase by approximately 34 percent by fully phased-in output floor.
  - Static assessment: capitalization ratio falls from 20.5 percent to 15.5 percent after a 34 percent increase in RWAs for IRB banks; pre-COVID shock results suggest IRB banks can withstand this without using macro-prudential buffers.
- Group-level consolidation:
  - Three-step approach to project group RWAs using adjusted parent/subsidiary weights (Appendix VII).

### Selected stress-test numerical and calibration details (preserve exact values)
- Credit loss projections for MCIs: projected LLRs peak at 1.0 percent in the last year of the scenario.
- MCIs top-up collateral cost spreads (COVID baseline): 60 bps, 50 bps, 30 bps, 10 bps, and -10 bps (2020–2024).
- MCIs top-up collateral cost spreads (pre-COVID “market shocks”): 270 bps, 240 bps, 230 bps, 15 bps, and 40 bps (2020–2024).
- Cost of funding shocks by liability type (Market shocks scenario, changes vs end-2019, example entries):
  - Deposits - Credit Institutions and other financial corporations - sight: 2.40, 2.90, 3.10, 2.00, 1.00 (2020–2024).
  - Debt securities issued - Certificates of deposits: 2.40, 2.90, 3.10, 2.00, 1.00.
  - Covered bonds: 0.40, 0.80, 1.00, 0.80, 0.40.
  - Note: Under the two COVID scenarios, funding shocks reflect observed market movements at this juncture (30 basis points for interbank lending and riskier debt securities and 15 basis points for covered bonds and repos in 2020).

### Main threats to financial stability and Risk Assessment Matrix highlights
- Main vulnerabilities summarized:
  - Indebtedness of the household sector and high real estate valuations.
  - Low profitability and signs of greater risk taking.
  - Interconnectedness, particularly through the covered bond market.
- Key high-concern risks:
  - Sharp rise in risk premia: overall level of concern = High.
  - Weaker-than-expected global growth: overall level of concern = High.
  - Prolonged domestic real estate downturn: overall level of concern = Medium.

### Selected policy recommendations and supervisory actions (short/medium term)
- Risk analysis and stress testing (selected):
  - Broaden country coverage of macroeconomic scenarios for solvency tests to include foreign countries where Danish banks and MCIs have significant exposures (DN). (MT)
  - Continue applying interest rate gap analysis for SIFI banks’ NII projections; standardize the DFSA-developed methodology and require SIFI banks to submit detailed data templates for TD exercises (DN and DFSA). (ST)
  - Introduce refinements to satellite models of credit risk: develop country-specific credit risk models for internationally active banks; explore collecting longer histories of impairment rates at granular levels; complement impairment-rate models with satellite models based on PDs. (MT)
  - MCIs: explore collecting historical loan impairment charges by loan type and improve information on guarantees provided by distributing banks to MCIs. (MT)
  - Update household-sector analysis using micro-level data and the credit register; assess resilience under protracted unemployment spikes. (MT)
  - Integrate NFC sector analysis into broader stability analysis and with bank-level credit register data; improve industry classification and introduce CRE/non-CRE split. (MT)
  - Further develop and standardize the multi-layer network model for interconnectedness and contagion developed during the FSAP. (ST)
- Insurance and pension fund stress testing (DFSA):
  - Develop a stress testing framework combining top-down and bottom-up perspectives; use stress test results to challenge insurers’ ORSA projections (DFSA, DN). (MT)
  - Require at least annual QRTs to be audited and require audit assurance for the systems and procedures used to complete QRTs; scrutinize QRT submissions systematically and consider administrative fines for repeated misreporting (DFSA). (ST)

*Source: 1dnkea2020009 — FSAP report content provided in the supplied document.*

### EXECUTIVE SUMMARY __________________________________________________________________________ 6

### EXECUTIVE SUMMARY

### Introduction and context
- The Financial Sector Assessment Program (FSAP) work was conducted prior to the COVID-19 pandemic; the report includes stability analysis and stress tests under updated illustrative scenarios to quantify possible implications of the COVID-19 shock on bank solvency.
- An unusually high degree of caution is required in interpreting the stress test results due to heightened uncertainty around post COVID central projections and downside risks.
- Financial vulnerabilities were elevated on the eve of the COVID-19 pandemic:
  - High household leverage amid high real estate valuations following a long period of loose financial conditions.
  - Signs of risk taking in some sectors, such as commercial real estate (CRE).
  - Downside risks to bank profitability amid the low-interest-rate environment.

### Macroeconomic scenarios used in solvency tests
- Three illustrative five-year scenarios were used in top-down (TD) solvency tests of the largest SIFIs:
  - "COVID central": captures abrupt deterioration of short-term economic prospects owing to the COVID-19 pandemic; focuses on growth and unemployment implications; does not entail abrupt tightening of financial conditions or a real estate downturn.
  - "COVID prolonged": reflects realization of downside risks associated with longer containment measures leading to a more prolonged recession; assumes no sharp tightening in financial conditions or decline in house prices.
  - "Market shocks" (adverse): assumes a severe materialization of market risks—a sharp rise in asset risk premia (adversely affecting banks’ funding costs) combined with a prolonged downturn in the domestic real estate market.

### Group-level solvency stress test results (banks and MCIs)
- Aggregate CET1 and leverage ratios under COVID scenarios:
  - Baseline (2019): aggregate CET1 ratio = 18.3 percent; aggregate leverage ratio = 4.8 percent.
  - COVID baseline (low point in 2020): aggregate CET1 ratio = 13.4 percent; aggregate leverage ratio = 3.6 percent.
  - COVID prolonged (low point in 2022): aggregate CET1 ratio = 12.4 percent.
  - COVID prolonged (aggregate leverage low point in 2024): aggregate leverage ratio = 3.0 percent.
- All SIFIs would meet their minimum capital requirements (a 4.5 percent CET1 ratio) under the scenarios modeled.
- Distributional outcomes and vulnerabilities:
  - Impact is differentiated across SIFIs; in some cases a few SIFIs need to partially use their capital conservation and/or SIFI buffers.
  - The leverage ratios of a few SIFIs fall below three percent.
- Uncertainty caveat:
  - Given the unprecedented nature of the pandemic, a further deterioration of macrofinancial conditions relative to those assumed could lead some SIFIs to breach minimum capital requirements.

### Main drivers of losses
- Credit risk and funding-cost shocks are the principal drivers across scenarios:
  - In the COVID scenarios, bank losses are mostly driven by materialization of credit risk, predominantly influenced by GDP growth and the unemployment rate.
  - In the "market shocks" scenario, cost of funding shocks also contribute materially to losses.
  - MCI results indicate capitalization impact is driven mostly by credit losses; tests assume within-group guarantees and those provided by banks outside the stress perimeter are available to cushion loan-default impact on MCI capitalization ratios.

### Liquidity stress testing (pre-COVID)
- TD liquidity stress tests for banks, implemented prior to the pandemic, used cash flow and LCR analysis.
- Results: all banks would be able to survive extreme funding shocks for 3 months or longer without external support.

### Insurance and pension fund stress testing
- Scope and coverage:
  - Top-down (TD) and bottom-up (BU) solvency stress tests run at solo entity level prior to the pandemic.
  - Covered five life insurers, six occupational pension funds, and five non-life insurers, accounting for more than 70 percent in each sector.
- Results under the "market shocks" adverse scenario:
  - Life insurers: median SCR ratio drops from 188 percent to 127 percent.
  - Non-life sector: median SCR ratio declines from 233 percent to 207 percent.
  - One life company falls below a solvency ratio of 100 percent; capital needed to restore full coverage would be a moderate fraction of the sample’s aggregate capital.
- Sensitivity findings:
  - Sensitivity tests point to longevity risks and assess resilience of non-life companies to climate-related catastrophic events.
- Key contributing shocks:
  - Most of the solvency impact can be attributed to lower stock prices and increases in domestic sovereign and covered bond yields.

### Non-financial corporate (NFC) sector vulnerability analysis
- Method and focus:
  - Used micro-data to estimate drivers of firm-level failures and how sectoral risks are exacerbated by correlated defaults.
- Key findings:
  - NFC debt has grown rapidly in highly cyclical sectors, with a particularly strong buildup of risks in the real estate sector.
  - Debt-at-risk analysis indicates a significant jump in default probabilities, particularly in the commercial real estate sector, under the "market shocks" adverse scenario.

### Recommendations to enhance stability analysis and stress testing (selected, by time horizon)
- Risk analysis and stress testing of banks and MCIs (responsibilities indicated in table; ST = Short term (1-3 years); MT = Medium Term (3-5 years)):
  - Broaden country coverage of macroeconomic scenarios for solvency tests to include foreign countries where Danish banks and MCIs have significant exposures (DN). (MT)
  - Continue applying interest rate gap analysis for projection of SIFI banks’ net interest income; standardize the DFSA-developed methodology and require SIFI banks to submit detailed data templates for TD exercises (DN and DFSA). (ST)
  - Introduce refinements to satellite models of credit risk:
    - Develop country-specific credit risk models for internationally active banks, requiring collection of country-specific historical impairment and default rates for large banks. (MT)
    - Explore collecting longer histories of impairment rates at granular levels for individual banks and sectors. (MT)
    - Complement impairment-rate models with satellite models based on default probabilities (PDs), requiring historical default data by bank and industry/sector exposure. (MT)
  - MCIs:
    - Explore feasibility of collecting historical loan impairment charges by loan type and estimate specific credit-risk models at aggregate and individual MCI levels. (MT)
    - Improve information availability on guarantees provided by distributing banks to individual MCIs. (MT)
  - Update and further refine household-sector analysis using micro-level data and the credit register to integrate into broader stability analysis; assess resilience under more protracted unemployment spikes and uneven household impacts. (DN). (MT)
  - Integrate NFC sector analysis into broader stability analysis and with bank-level credit register data; improve industry classification and introduce sub-industry breakdowns (e.g., split "real estate" into CRE and non-CRE). (DN). (MT)
  - Further develop and standardize use of the multi-layer network model for interconnectedness and contagion developed during the FSAP to continuously assess systemic risks. (DN). (ST)
- Insurance and pension fund stress testing (DFSA):
  - Develop a stress testing framework combining top-down and bottom-up perspectives; use stress test results to challenge insurers’ ORSA projections and undertake horizontal reviews of ORSA projections (DFSA, DN). (MT)
  - Require at least annual QRTs to be audited and require audit assurance for systems and procedures used to complete QRTs; scrutinize QRT submissions more systematically, request re-submissions if necessary, and consider administrative fines for repeated misreporting (DFSA). (ST)

*Source: EXECUTIVE SUMMARY, FSAP report content provided in the supplied document.*

### INTRODUCTION

### INTRODUCTION

### Recent macroeconomic developments and COVID-19 impact
- Steady growth in recent years was abruptly interrupted by the COVID-19 crisis.
- During the 2008-2009 global financial crisis the economy experienced a sharp downturn exacerbated by the burst of the domestic real estate bubble; recovery trailed regional peers.
- In the years immediately before the pandemic: negative policy rates, unemployment at a 10-year low, a positive output gap, and the krone’s tight peg to the euro kept inflation low and stable.
- The domestic outbreak of the pandemic caused a sudden stop in economic activity, a sharp deterioration of short-term economic prospects due in part to containment measures, and a sharp tightening of financial conditions in March 2020.
- Timing and shape of future recovery are highly uncertain and risks are to the downside; concerns include further tightening of financial conditions or a second pandemic wave exposing financial vulnerabilities.

### Pre-pandemic vulnerabilities: households, housing, and mortgages
- Household debt:
  - Denmark’s high level of household (gross) debt remains an important vulnerability despite a gradual downtrend.
  - Household net wealth is relatively high, but a large share of assets is illiquid (housing and pension savings), limiting buffers to shocks.
  - Large household debt increases susceptibility of consumption to adverse interest rate, asset price, and income shocks and can reinforce macrofinancial feedback loops.
- Real estate prices:
  - Nationwide house prices had stabilized in response to prudential measures and appeared broadly in line with fundamentals.
  - House prices in urban areas remained elevated through 2019; downside risks and misalignments appear greater in Copenhagen relative to the national market.
- Mortgage loans:
  - 63 percent of the mortgage stock are variable-rate mortgages, of which 59 percent are non-amortizing, leaving households sensitive to interest rate hikes.
  - Note: more than 50 percent of variable-rate mortgages have fixed-interest-rate periods that are greater than one year.

### Other financial stability concerns: risk taking, corporate sector, covered bonds
- Risk taking in financial intermediaries:
  - Evidence of greater risk taking across banking and insurance sectors.
  - Some medium-sized banks were aggressively extending credit to households with high loan-to-income ratios to increase market share in major urban areas (Copenhagen, Aarhus).
  - Insurers increasingly invested in riskier assets; shift toward unit-linked products supports insurer profitability but transfers more risk to policyholders and raises consumer protection concerns.
- Nonfinancial corporate (NFC) sector and commercial real estate (CRE):
  - Rising leverage in cyclically-oriented firms despite overall decline before the COVID shock; loosening credit standards for the NFC sector.
  - CRE market characterized by rising prices, a large share of foreign investors, concerns regarding income-generating capacity; CRE accounted for about 30 percent of total corporate debt and was especially vulnerable.
  - CRE was among the most leveraged corporate sectors and where debt had grown fastest in recent years; banks have sizeable exposures to CRE, amplifying system interconnections.
  - The pandemic has likely aggravated these vulnerabilities; highly indebted and/or lower-income households and several corporate sectors are relatively more susceptible to shocks.
- Covered bond market:
  - Strong reliance on market-based funding and increasing share of foreign investors provides diversification benefits but increases sensitivity to global investor sentiment.
  - Danish covered bond framework mitigates funding risks, but there are substantial holdings of covered bonds across domestic financial institutions.
- Interconnectedness and contagion:
  - Disruption in the covered bond market is a very low probability but high impact event; a spike in covered bond yields could cause mark-to-market losses, reduced market liquidity, funding pressures, and capital outflows from foreign investor selling.

### FSAP stress testing objectives, scope, and caveats
- Objective: assess capacity of the financial system to withstand extreme but plausible macroeconomic shocks; identify weaknesses and channels of transmission; determine priorities for policy action including micro- and macro-prudential measures.
- Tests differ from central bank stress tests; FSAP tests carried out in close cooperation with DN and DFSA with access to detailed supervisory data.
- Key caveats and limitations:
  - Results should be interpreted with caution due to large uncertainty in scenarios.
  - Analysis presumes pre-crisis relationships between macroeconomic and financial variables remain unchanged.
  - Relationship between growth and capitalization is non-linear; deeper recessions can have disproportionate impacts on capital, making results sensitive to scenario severity.
  - Some second-round and non-linear effects are not fully captured and could imply underestimation of losses.
  - Choices made regarding hypothetical shocks; scenarios are not forecasts and do not necessarily coincide with IMF WEO projections.
  - Under all scenarios, bank capital impact is subject to downside risks if actual macro outcomes are worse than stipulated.

### Stress testing design, coverage, and methods
- Scope and coverage:
  - Top-down solvency stress tests of banks and MCIs cover all seven SIFIs representing 77 percent of the banking system’s total assets; tests performed on largest banks and MCIs individually and at group (consolidated) level.
  - Liquidity stress tests assessed banks’ capacity to confront large funding withdrawals based on cash flow and LCR analysis before the COVID-19 pandemic; implications of COVID-19 for banking liquidity or system-wide contagion were not assessed.
  - Insurance and pension sector tests include bottom-up and top-down exercises and risk analysis for ATP; sensitivity analysis assessed resilience of non-life companies to climate-related catastrophic events such as severe windstorms.
  - Corporate vulnerability analysis uses micro-data to estimate drivers of corporate failures and sectoral correlated defaults, then quantifies loss implications for banks’ corporate exposures.
  - Systemic liquidity and interconnectedness/contagion risks are developed in separate Technical Notes.
- Scenarios and exercises:
  - Three principal scenarios: “COVID central”, “COVID prolonged”, and “market shocks” adverse scenario (illustrative five-year scenarios and sensitivity analyses).
  - Tests include solvency, liquidity, and contagion dimensions; contagion analysis uses a novel multi-layer network model based on bilateral exposures across institutions and sectors with 8 types of exposures (loans, deposits, reverse repos, covered bonds, other debt securities, equities, unlisted shares, and other claims).
  - Liquidity tests include cash flow-based maturity-ladder analysis and LCR-based tests in domestic and foreign currency.
  - Tests cover 21 largest credit institutions (86 percent of assets): 5 systemic banks, 9 non-systemic banks, 7 mortgage credit institutions; and include institutional investors, households, and corporates.
  - Corporate sector analysis includes debt-at-risk and correlated default analyses using micro-level data.

### Innovative features of the 2020 FSAP stress testing exercise
- Detailed interest rates GAP analysis using granular FSAP-collected data to refine projections of net interest income (NII) and profitability, accounting for cross-bank differences in maturity transformation, time-to-repricing, and other asset/liability characteristics.
- Top-down solvency tests of Mortgage Credit Institutions (MCIs) conducted for the first time in an FSAP, including calculation of credit and market losses and impact of collateral top-ups to maintain LTV ratios on MCI profitability.
- Use of default correlation models to separate effects of shocks to firm-specific PDs and shocks to default correlation on portfolio credit losses—capturing how higher default correlation reduces diversification benefits.

### Main threats to financial stability (Box 1 summary)
- Vulnerabilities:
  - The indebtedness of the household sector and high real estate valuations.
  - Low profitability due to continued decline in interest rates and signs of greater risk taking.
  - Interconnectedness, particularly through the covered bond market.
- Associated risks:
  - Mortgage defaults could be triggered by a surge in unemployment associated with an economic slowdown; longer-than-average unemployment spells could particularly undermine repayment capacity of younger and more highly indebted households.
  - Net interest income may continue declining in a low interest rate environment; banks may be unable to expand deposit bases earning negative rates or fully offset NII erosion via fee income; intensified risk taking may deteriorate credit quality.
  - A disruption in the covered bond market—a very low probability but high impact event—could be triggered by an extreme surge in mortgage defaults that depletes MCIs’ buffers, leading to higher covered bond yields, mark-to-market losses, temporary liquidity reductions, funding pressures, and capital outflows.

### Macroeconomic scenarios (selected parameters)
- COVID central scenario:
  - GDP falls by 8 percent in 2020, rebounds by 6 percent in 2021.
  - Unemployment rate rises to 6.8 percent in 2020 before gradually declining.
  - Scenario focuses on growth and unemployment implications and does not entail abrupt tightening of financial conditions or a real estate downturn.
  - Note: the GDP growth rate for 2020 in this scenario (-8 percent) is lower than IMF WEO April projection (-6.5 percent) and Danish Economic Council June projection (-4.5 percent).
- COVID prolonged scenario:
  - Growth declines by 8 percent in 2020 and, instead of a rebound in 2021, the economy continues to contract (details continue beyond the supplied excerpt).

*Source: 1dnkea2020009 - INTRODUCTION*

### 2.1 percent in 2021) before the recovery ensues in 2022.

### 1dnkea2020009 - 2.1 percent in 2021) before the recovery ensues in 2022.

### Macroeconomic scenarios (overview)
- Three scenarios are modeled for Denmark: COVID Central, COVID Prolonged, and Market Shocks.
- Assumptions and key characteristics:
  - COVID Prolonged: cumulative real GDP decline of 8.1 percent points relative to the COVID central scenario over two years (equivalent to 2.4 standard deviations). A standard deviation of (two-year cumulative) real GDP growth, based on 1966–2018, is equal to 3.4 percentage points.
  - Market Shocks: sharp rise in asset risk premia (increasing banks’ funding costs) and a prolonged downturn in the domestic real estate market; incorporates a combination of credit, market, and cost of funding shocks. Designed using input from the IMF’s Flexible System of Global Models.
- Scenario caveats:
  - The scenarios are not meant to be forecasts and are associated with considerable uncertainty and downside risks; the COVID scenarios do not necessarily coincide with IMF WEO projections.

### Real GDP and labor market paths (selected figures)
- Real GDP growth (2019–2024) by scenario (In percent):
  - COVID Central: 1.7, -8.0, 6.0, 3.5, 2.0, 1.8
  - COVID Prolonged: 1.7, -8.0, -2.1, 3.3, 3.3, 3.3
  - Market Shocks: 1.7, -1.4, -3.6, -1.0, 1.1, 1.3
- Harmonized unemployment rate (2019–2024) by scenario:
  - COVID Central: 5.0, 6.8, 6.2, 5.7, 5.6, 5.6
  - COVID Prolonged: 5.0, 6.8, 10.4, 11.8, 11.0, 10.0
  - Market Shocks: 5.0, 6.2, 8.6, 10.2, 10.5, 10.1
- Consumer price inflation rate (CPI year average) (2019–2024) by scenario:
  - COVID Central: 1.3, 0.6, 0.9, 1.2, 1.4, 1.5
  - COVID Prolonged: 1.3, 0.6, 0.9, 1.2, 1.4, 1.5
  - Market Shocks: 1.3, -0.7, -0.9, -0.2, 2.1, 2.2

### Interest rates, exchange rate, and housing prices
- Nominal policy interest rate (All Scenarios, 2019–2024): -0.8, -0.9, -0.9, -0.8, -0.6, -0.5
- Short term interbank nominal interest rate (3 month) (All Scenarios): -0.8, 1.7, 2.3, 2.6, 1.4, 0.3
- Bank overall funding cost rate (shocks) 1/:
  - Both COVID Scenarios: 0.1, 0.0, 0.0, 0.0, 0.0
  - Market Shocks: 0.6, 0.9, 1.0, 0.8, 0.4
- Long term nominal interest rate (10 year sovereign bond yield):
  - Both COVID Scenarios: -0.3, -0.6, -0.4, -0.1, 0.2, 0.6
  - Market Shocks: -0.3, -0.1, 0.8, 1.3, 1.0, 1.1
- Nominal exchange rate change (DK/EUR; + is DK depreciation, in percent) (All Scenarios): 0.0, 0.0, 0.0, 0.0, 0.0, 0.0
- Housing prices (2019=100):
  - Both COVID Scenarios: 100.0, 101.5, 103.3, 105.4, 107.5, 109.7
  - Market Shocks: 100.0, 91.2, 82.4, 77.5, 79.1, 80.6
- Nominal growth of assets in the banking sector (for stress tests) 2/:
  - COVID Central: 3.5, 0.0, 6.9, 4.9, 3.7, 3.6
  - COVID Prolonged: 3.5, 0.0, 0.0, 4.7, 5.0, 5.0
  - Market Shocks: 3.5, 0.0, 0.0, 0.0, 3.6, 3.9

### Market shocks: external and country-specific layers
- External layer:
  - Escalating global trade tensions and tightening global financial conditions depress business and consumer confidence across advanced and EM economies.
  - Abrupt decompression of risk asset premia amplified by low secondary market liquidity.
  - Within the euro area, higher money market interest rates, banks’ cost of short-term funding, and long-term government bond yields—“high spread economies” hit particularly hard.
  - Monetary normalization in systemic advanced economies fails to materialize in the stress testing period (2020–24).
  - Nordic region: Sweden, Norway and Finland are assumed to be hit by extreme shocks.
- Country-specific layer:
  - Housing prices decline by 22 percent over three years, adversely affecting domestic consumption and residential investment.
  - Exchange rate regime assumed to remain credible; DN able to maintain negative policy rates (at 25 basis points below the projected ECB rates).
  - Fiscal policy remains neutral—only automatic stabilizers play a counter-cyclical role.
- Cross-country GDP deviations under Market Shocks (two-year cumulative deviations from baseline):
  - Sweden: 7.7 percentage points (equivalent to 2.1 standard deviations)
  - Finland: 10.5 percentage points (equivalent to 1.9 standard deviations)
  - Norway: 8.8 percentage points (equivalent to 2.7 standard deviations)

### Cost of funding shocks by liability type (Market shocks scenario, changes vs end-2019, in percent)
- Deposits (excl. repo) - Central banks - ELA: 0.10, 0.20, 0.20, 0.10, 0.10 (2020–2024)
- Deposits (excl. repo) - General governments - sight: 0.10, 0.20, 0.20, 0.10, 0.10
- Deposits (excl. repo) - General governments - term: 0.10, 0.20, 0.20, 0.10, 0.10
- Deposits (excl. repo) - Credit Institutions and other financial corporations - sight: 2.40, 2.90, 3.10, 2.00, 1.00
- Deposits (excl. repo) - Credit Institutions and other financial corporations - term: 2.40, 2.90, 3.10, 2.00, 1.00
- Deposits (excl. repo) - Non-financial corporations - sight: 0.30, 0.60, 0.70, 0.60, 0.30
- Deposits (excl. repo) - Non-financial corporations - term: 0.50, 1.00, 1.10, 1.00, 0.50
- Deposits (excl. repo) - Households - sight: 0.15, 0.30, 0.40, 0.30, 0.15
- Deposits (excl. repo) - Households - term: 0.25, 0.50, 0.60, 0.50, 0.25
- Deposits - Repo: 0.40, 0.80, 0.90, 0.80, 0.40
- Debt securities issued - Certificates of deposits: 2.40, 2.90, 3.10, 2.00, 1.00
- Debt securities issued - Asset-backed securities: 2.40, 2.90, 3.10, 2.00, 1.00
- Debt securities issued - Covered bonds: 0.40, 0.80, 1.00, 0.80, 0.40
- Debt securities issued - Other debt securities issued: 2.40, 2.90, 3.10, 2.00, 1.00
- Derivatives: 0.00, 0.00, 0.00, 0.00, 0.00
- Other liabilities: 0.50, 1.00, 1.10, 1.00, 0.50
- Note: Under the two COVID scenarios, funding shocks reflect observed market movements at this juncture (30 basis points for interbank lending and riskier debt securities and 15 basis points for covered bonds and repos in 2020).

### Macro-financial transmission to banks and Mortgage Credit Institutions (MCIs)
- Transmission channels (adverse scenario):
  - Credit losses driven by economic slowdown and rising unemployment.
  - Net interest income (NII) stressed via partial pass-through of higher funding costs to lending rates:
    - Funding shocks cannot be passed to borrowers immediately because of maturity transformation (loans have longer maturity/time to repricing than deposits).
    - A 50 percent pass-through of funding shocks to lending rates on new loans is assumed.
  - Market losses triggered by spikes in sovereign, corporate, and covered bond yields.
- MCIs:
  - Losses mainly from mortgage default risk (credit losses) and the need to top-up collateral when house prices decline to maintain LTV ratios—affecting their NII.

### Solvency stress testing methodology and sector heterogeneity
- Tests use 2020 Q1 capital ratios as starting point across scenarios.
- Heterogeneity drivers:
  - Sectoral composition of loan books leads to different loan loss rates across banks.
    - System exposures: financial sector 32 percent, households 27 percent, real estate 12 percent, manufacturing industries 9 percent.
    - Individual banks show significant heterogeneity—some concentrated in financial sector, others in cyclical sectors (agriculture, construction, trade).
  - Funding structure and maturity profiles create differences in NII declines across banks.
    - Some banks rely more on unstable funding (deposits from credit institutions and financial corporations, debt securities), others on stable funding (households, non-financial corporates).
    - Larger funding shocks applied to unstable funding sources make those banks suffer larger NII declines.
  - RWA density ranges from 28 percent to 61 percent across the five banks tested—lower risk weights amplify capital ratio impact of losses.

### Credit losses: model and key findings
- Loan loss rates (LLRs) estimated for 9 sectors using a dynamic panel regression with sector-specific fixed effects.
  - Independent variables: real GDP growth rate, unemployment rate, short-term interest rate.
  - Dependent variable: logistic transformation of LLR to allow non-linear effects (larger increases for historically more cyclical sectors and sectors with higher starting LLRs).
- Starting point: system-wide LLRs at record lows and virtually zero.
  - Agriculture LLR: 1.1 percent
  - Trade LLR: 0.4 percent
- Projections:
  - LLRs rise sharply in the “COVID-prolonged” and “market shocks” scenarios from the unusually low starting levels.
  - Agricultural, real estate and construction sectors are historically more cyclical and yield larger projected losses in the stress tests due to lower sector fixed effects and the logit transformation.

### Net interest income (NII)
- NII expected to decline as lending rates on recent new business have been declining while funding rates have been somewhat stable.
  - New business lending rates for 2019 are approximately 0.1 percent lower than the overall average lending rate.
  - 2019 new business funding rates are approximately the same as overall average funding rates.
  - Implication: limited room for further funding cost reductions and continued compression of net interest margins.

*Sources: IMF staff calculations.*

### 25. A detailed and highly granular approach was used to project bank net interest income.

### 25. A detailed and highly granular approach was used to project bank net interest income.

### Projection methodology for net interest income (NII)
- Data collection:
  - Loans classified into 6 categories; securities portfolios into 3 categories.
  - Deposits reported in 10 categories; issued debt securities in 4 groups.
  - For each item, banks reported:
    - total volume at the end of starting quarter,
    - average effective interest rates on the entire portfolio,
    - effective interest rates corresponding to “new business” activity during 2019.
  - Asset and liability amounts distributed into time-to-repricing intervals, which are in turn classified by original maturity.
- Repricing and new business rates:
  - Interest income and expense calculations follow schedules of time-to-repricing and assign new business rates as prices reset.
  - New business rates during the projection horizon = new business rates of 2019 adjusted by idiosyncratic funding shocks.
  - Stress tests performed under the “constant balance sheet assumption” (maturing loans always renewed at “new business” rates; lending not interrupted).
- Implementation detail:
  - Calculations track repricing times and rates for each asset/liability item via “trees” that follow repricing events and rate resets.
  - Stylized example corresponds to a 3-year horizon; actual tests applied over a 5-year horizon.

### Numerical example (Figure 7) — loan portfolio repricing (stylized)
- Portfolio valued at 30,000 DKK at end-2019.
  - 10,000 DKK matures (M) or reprices in 2020.
  - Remaining 20,000 DKK does not mature in 2020; of this:
    - 6,000 DKK reprices in 2021 (original time to repricing >2 years, so does not reprice again in 2022).
    - Remaining 14,000 DKK has 4,000 DKK reprice in 2022.
  - The 10,000 DKK repriced in 2020 is re-invested with an “original maturity” distribution:
    - 5,500 DKK will reprice the following year,
    - 3,500 DKK will reprice after two years,
    - 1,000 DKK will not reprice until a later date.
- Rate resetting convention:
  - Typical loan yields the starting rate for ½ year and the new (re-set) rate for the rest of the year.

### Market valuation losses (securities)
- Measurement approach:
  - Market valuation losses on sovereign and non-sovereign debt securities measured through changes in yields leading to re-pricing using a modified duration approach.
  - For every country and type of issuer, yield curves constructed by linear interpolation of short- and long-term interest rates, as specified in the macroeconomic scenarios.
  - Changes in yields for a given modified duration applied to calculate haircuts and re-price bond portfolios in all accounts (held for trading, available for sale, fair-valued, and held-to-maturity), excluding floating rate securities.
- Formula components (as described):
  - MD = modified duration of the portfolio.
  - MD y = change in the yield caused by the shift in the yield curve (vis-à-vis the value prevailing in the previous year), measured at a point in time that matches the modified duration of the portfolio.

### Mortgage Credit Institutions (MCIs): solvency tests based on macro scenarios
- Role and structure:
  - MCIs originate mortgage loans (via bank branches) and finance them by issuance of covered bonds.
  - Mortgages pooled and placed in bankruptcy-remote capital centers managed by MCIs; each MCI has a number of capital centers.
  - Capital center balance sheet components include mortgage loans, covered bonds, junior covered bonds, unsecured bonds, liquid assets, capital, assets for pillar 2 liquidity requirement, assets for LCR, assets for over-collateralization requirement, assets for supplementary capital.
- Institutional/regulatory features limiting MCI risks:
  - Over-collateralization requirements.
  - Guarantees provided by banks against loan defaults.
  - Matching principle ensuring MCIs pass interest rate risk from covered bond investors to borrowers completely.
  - Very limited exposure to funding liquidity risk.
- Stress-test channels for MCIs:
  - Credit losses:
    - Loan loss rates (LLRs) under the adverse scenario estimated using a dynamic time series regression model with independent variables: real GDP growth rate, unemployment rate, a short-term interest rate, and inflation.
    - Dependent variable is logistic transformation of the LLR to allow non-linear effects.
    - Projected LLRs rise gradually from virtually zero at the starting point and peak at 1.0 percent in the last year of the scenario.
    - Main driver of mortgage defaults = unemployment rate, which peaks toward the end of the horizon.
    - Projected LLRs exhibit strong persistence due to a significant coefficient on the lagged dependent variable.
  - Market losses:
    - Repricing methodology same as banking sector stress tests; analysis conducted using yield curves provided by IMF staff.
  - Loss of net interest income:
    - Occurs when MCIs must issue junior bonds to top-up collateral in capital centers due to LTV ratios rising above 80/60 percent limits when housing prices decline; issuance used to buy and place sovereign securities in capital centers to reduce LTV to permissible levels.
- Guarantees accounted for:
  - Tests assume within-group guarantees and those provided by banks outside stress testing perimeter are available to cushion impact of loan defaults on MCIs’ capitalization ratios.
  - After accounting for guarantees, all MCIs can fully absorb losses through release of the CCyB and drawing down excess capital buffers.

### Consolidated (group-level) solvency test results
- Coverage:
  - Groups: Danske Group, Jyske Group, Nykredit Group; plus two solo banks (Sydbank and Spar Nord Bank), and two solo MCIs (DLR Kredit and Nordea Kredit) — total 7 systemic credit institutions (SIFIs).
- Methodological issues in consolidation:
  1. Reported group-level RWAs are lower than sum of bank and MCI RWAs; projected RWAs for bank and MCI cannot be simply added (see Appendix VII).
  2. Intra-group flows of profits/dividends: income from subsidiaries excluded from stress test assumed constant at 2019 level.
  3. Interest on intra-group cross exposures netted out at group level (contrasts with solo-level GAP-based NII projections that include inter-company exposures).
- Illustrative scenario outcomes:
  - COVID central scenario:
    - Aggregate CET1 ratio declines from 18.3 percent in 2019 to a low point of 13.4 percent in 2020.
    - Aggregate leverage ratio declines from 4.8 percent to 3.6 percent over same period.
    - Credit risk (driven by GDP growth and unemployment) is main driver of losses.
    - All SIFIs meet minimum capital requirements (4.5 percent CET1); one of seven SIFIs must partially tap its capital conservation buffer (CCB) to absorb losses.
    - Leverage ratio for one SIFI falls below the three percent threshold.
  - COVID prolonged scenario:
    - Aggregate CET1 ratio falls to a low point of 12.4 percent in 2022.
    - Aggregate leverage ratio declines to 3.0 percent in 2024.
    - All SIFIs meet minimum capital requirements; one SIFI partially uses CCB, another partially uses SIFI buffer.
    - Leverage ratios of a few SIFIs fall below three percent.
  - Market shocks scenario:
    - Aggregate CET1 ratio and leverage ratio decrease to 13.4 percent and 3.4 percent, respectively, in 2024 (troughs at end of horizon).
    - Scenario assumes severe market shocks and lasting domestic real estate downturn; results primarily driven by protracted downturn.
    - A few SIFIs partially use CCBs; a few SIFIs’ leverage ratios dip slightly below three percent.
- Drivers of CET1 reductions:
  - COVID scenarios:
    - CET1 reduction driven by front-loaded credit losses in both parent and subsidiary entities.
    - Bank credit losses estimated to reach unprecedented high levels in 2020.
    - Impact of cost of funding shocks and market risk limited because cost of funding shocks used are those realized so far in 2020 only for this scenario.
  - Market shocks scenario:
    - CET1 reduction driven by credit losses in parent entities during first three years; subsidiary losses become significant later.
    - Pre-COVID adverse scenario envisages a moderate but protracted GDP decline continuing for three years; recovery not robust, resulting in sizable backloaded losses.
    - Reduced net interest income due to large cost of funding shocks and losses from repricing of securities (yield curve shifts) significantly contribute to CET1 declines.
- Cautions and limitations:
  - Relationship between growth and capitalization is non-linear; deeper recessions have potentially disproportionate impacts on capital.
  - Stress test results are very sensitive to scenario severity.
  - Some second-round and non-linear effects not fully captured could imply underestimation of losses.
  - Further deterioration of economic prospects and tighter financial conditions relative to scenarios could lead to one or more SIFIs breaching minimum capital requirements.

### Banks: pre-COVID sensitivity tests (concentration risk and Basel III output floors)
- Concentration risk test:
  - Impact assessed of defaults of largest 3, 5, and 10 non-sovereign exposures (static, instantaneous impact at starting quarter; RWAs assumed constant; pre-impairment income not allowed to offset shocks).
  - Sum of largest ten exposures for all five banks subject to the stress testing exercise = 51 percent of their aggregate CET1 regulatory capital.
  - Results (under 40 percent LGD assumption):
    - Simultaneous default of largest 3 exposures reduces CET1 ratio to 18.8 percent from current 20.5 percent.
    - Simultaneous default of largest 5 exposures reduces CET1 ratio to 17.9 percent from 20.5 percent.
    - Simultaneous default of largest 10 exposures reduces CET1 ratio to 16.3 percent from 20.5 percent.
    - All banks remain with excess capital above minimum requirements and macro-prudential buffers in these tests.
  - More stringent test:
    - Simultaneous failure of largest ten borrowers with an LGD of 60 percent reduces CET1 capital to 14.2 percent, causing one of the five banks to deplete its CCyB buffer and partially use the CCB buffer.
- Sensitivity to Basel III output floors:
  - Output floors will increase RWAs by 34% when fully phased-in.
  - Danish banks are well-capitalized to withstand the impact from Basel-III output floors (per the illustrative results).

*Source: IMF staff calculations and stress-test results presented in the chapter.*

### 41. Before the onset of the pandemic, we also conducted a sensitivity analysis to gauge

### 1dnkea2020009 - 41. Before the onset of the pandemic, we also conducted a sensitivity analysis to gauge

### Basel III output floor sensitivity (IRB banks)
- Basel III output floor: RWAs floored at "72.5 percent of standardized RWAs" for banks using advanced IRB methods.
- Phasing timetable: output floor "rise gradually starting in 2023 and be fully phased-in by 2028."
- Industry and Danish authorities estimate: "RWAs will increase by approximately 34 percent in the Danish banking sector by the time the regulation is fully phased-in in 2028."
- Analysis approach: static assessment that assumes the floors "were fully phased-in today for the four IRB banks."
- Impact on aggregate capitalization:
  - Capitalization ratio falls from "20.5 percent" to "15.5 percent" after a "34 percent" increase in RWAs for the IRB banks.
  - Pre-COVID shock results suggest all four IRB banks can withstand this RWA increase without using macro-prudential buffers.
- Caveats and conditional factors:
  - Analysis is "conducted in a static manner."
  - Results "cannot be directly combined with the insights of section C" because the dynamics of the output floor during an economic downturn have not been analyzed.
  - It is possible the effect of the output floor will be lower than "34 percent" during a recession where IRB risk weights of performing exposures will be higher.
  - The output floor "will only partially be phased-in during the stress testing horizon which ends in 2024."

### Pre-COVID liquidity stress tests of banks (top-down maturity-ladder methodology)
- Methodology highlights:
  - Cash-flow based, top-down (TD) maturity ladder analysis using time structure of contractual cash flows.
  - Tests captured (i) liquidity needs from outflows; (ii) available standby liquidity from inflows; (iii) buffers to counterbalance liquidity gaps.
  - Run-off rate principles:
    - Wholesale funding run-off rates higher than retail funding run-off rates.
    - Secured funding run-off rates lower than unsecured.
    - Sight deposits/insured deposits withdrawn slower than time deposits/uninsured.
    - Cash outflows assumed protracted—"lasting for up to one year" with higher run-off rates in "0–90 days" bucket and lowest after "180 days".
- Roll-off/conversion rates for assets:
  - "100 percent" roll-off rates applied to all maturing asset items except for loans to retail customers: "70 percent".
- Liquidity relief tools allowed in tests:
  - Use of cash and securities (secured instruments, covered bonds, sovereign and non-sovereign securities).
  - Use of assets as collateral to obtain liquidity and ELA from the DN.
  - Pass-fail criterion: a bank that needs ELA to continue operating is considered to have "failed the test."
- Pre-COVID results (based on September 2019 COREP data):
  - Survival periods under assumed shocks:
    - All banks: survival period "longer than 3 months."
    - Three of the 5 banks: survival periods "5 months or longer."
  - Sensitivity LCR tests covered five scenarios for all currencies (and four scenarios for EUR LCR):
    - (i) "LCR baseline"
    - (ii) "freezing of 'non-HQLA' repo market (higher cash outflows)"
    - (iii) "covered bond market stress via higher yields and haircuts (lower HQLA)"
    - (iv) "higher run-off rates on retail liabilities and reduced inflows from retail counter-parties ('Retail shocks', higher outflows and lower inflows)"
    - (v) "higher run-off rates on financing from wholesale sources, reduced inflows from wholesale counterparties, and higher asset haircuts reducing the banks’ counterbalancing capacity ('Wholesale and CBC' higher outflows, lower HQLA)"
  - LCR test outcomes:
    - All-currencies LCR: "3 of the 5 banks would have post-shock LCRs at about 100 percent or above"; the two other banks would remain "above 65 percent" after shocks.
    - EUR LCR: "3 of the 5 banks remain close to or above 100 percent"; the two remaining banks end up in the "80-100 percent" range.
- Specific scenario parameter details:
  - Covered bond market stress:
    - Level 1B covered bonds: liquidation values decline from "93 percent" to "75 percent" (haircuts rise from "7 percent" to "25 percent").
    - Level 2A covered bonds: liquidation values decline from "85 percent" to "65 percent" (haircuts rise from "15 percent" to "35 percent").
  - "Retail shocks (IMF)" scenario: doubles LCR run-off rates on demand deposits; multiplies by "5" the LCR rates on undrawn but committed credit and liquidity lines; reduces by "½" the roll-off rates on inflows from retail counterparties.
  - "Wholesale and CBC (IMF)" scenario: doubles run-off rates on unsecured wholesale funding and on secured funding backed by non-Level 1 or non-Level 2A assets; reduces by "½" the roll-off rates on inflows from institutional counterparties (NFCs and financial institutions); assumes lower liquidation values for Level 2A assets (from "85 percent" to "75 percent") and Level 2B assets (from "50-75 percent" to "40-65 percent").
- Comparative note on stress severity and COVID-19:
  - Shocks assumed in pre-COVID tests were "significantly larger than those that have materialized in 2020 Q1."
  - May 2020 DN Financial Stability Report (FSR) finding: systemic banks, in aggregate, "could cover their liquidity needs under a severe liquidity stress scenario."
  - FSR observation: "overall liquidity situation for systemic banks has not been materially affected by the COVID-19 pandemic as of mid-March."
  - Median LCR across systemic banks increased to "198 percent" in April 2020 from "174 percent" at end-2019.

### Liquidity risk in Mortgage Credit Institutions (MCIs) and Pillar II liquidity requirements
- Structural protections and practices:
  - Covered bonds include clauses that protect MCIs against funding liquidity risk: bond holders can be forced to roll-over positions for "up to thirty years (one year at a time)" in extreme event.
  - Household prepayment rule: household must give "60-day advance notice" to pre-pay a mortgage, obliging full payment; most households lock-in terms of a new mortgage when giving notice.
- Risks from open positions:
  - If borrowers with "open positions" (no new mortgage locked-in) cannot obtain new financing within the "60-day period" because of an extreme covered bond market disruption, MCIs could suffer reduced cash inflows and need to use liquidity buffers—motivating Pillar II add-ons.
- Existing Pillar II framework (background):
  - Danish FSA exemptions: five Danish specialised MCIs (SMCIs) granted exemptions under Article 26 in the LCR Delegated Regulation (LCR DR) allowing cash flows related to mortgage lending to be netted.
  - Existing Pillar II liquidity floor: SMCIs required to hold a liquidity buffer of "2.5 percent of their total lending" (called "LCR floor requirement") composed of assets meeting HQLA definition; caps in Article 17 of LCR DR do not apply to this buffer.
  - The LCR floor requirement is "almost always" the binding requirement.
- New Pillar II model (reform and add-ons):
  - Danish FSA will replace existing LCR floor with a risk-based Pillar II requirement containing at least three specific Pillar II add-ons to better reflect SMCIs' business-model specific liquidity risks.
  - The new model defines the liquidity buffer according to the LCR DR and mirrors practice of setting specific add-ons (as used for Pillar II capital requirements).
  - Types of expected Pillar II add-ons:
    - Pillar II A add-ons: at least two—one for "unknown arrears" and one for "unknown open conversions."
    - Pillar II B add-on: for "refinancing risk."
  - SMCIs must set add-ons based on their own risks, experiences, and historical data; Danish FSA guidance provides benchmarks and methods.
  - Prudential stance: SMCIs must adopt a "prudent approach" in calculating Pillar II liquidity add-ons; Danish FSA may impose an add-on if calculation is not appropriately prudent.
  - Covering requirements: SMCIs must hold liquid assets to cover the LCR requirement, Pillar II A add-ons and potentially a Pillar II B add-on; both add-ons must be covered with "unencumbered HQLA after haircuts and caps (Article 17) according to the LCR DR."
- Pillar II A add-on specifics:
  - Arrears: Pillar II A add-on for arrears determined as a percentage of future interest and principal payments in an unlikely but not inconceivable stress scenario; individual SMCI determines percentages for private and business customers; add-on calculated with deduction for expected arrears (as expected arrears are included in the LCR requirement).
  - Open conversions: open conversions are included as a regular outflow with a weight of "100 pct." in the LCR requirement; open conversions are included in LCR "30 days" before the date of payment until the date of payment; Danish FSA will introduce a Pillar II A add-on for unknown open conversions to ensure sufficient HQLA in stress where many borrowers do not choose fixed-price agreement and thus large volume of open conversions.

*Source: 1dnkea2020009*

### Box 2. Liquidity MCIs (continued)

### Box 2. Liquidity MCIs (continued)

### Open conversions and Pillar II A add-on
- Rationale:
  - Prior to a conversion trend (typically initiated by larger movements in the interest rate level) with potentially a large volume of open conversions, SMCIs need HQLA to cover a certain amount of these open conversions when they enter the LCR 30-day window rather than starting with no liquidity to cover this risk.
  - The Danish FSA also wishes to ensure that the SMCI’s liquidity buffers do not vary dramatically between dates of payment, i.e. in the months when open conversions are not included in the LCR requirement.
- Calibration and application:
  - The Danish FSA’s new guidance will specify that the Pillar II A add-on for open conversions should amount to at least 30 pct. of the open conversions in an unlikely, but not inconceivable, stress scenario.
  - If the amount of open conversions that is included in the LCR requirement 30 days before date of payment is lower than the amount of open conversions that is used in the Pillar II A add-on, the SMCI must take the difference between these two amounts as a Pillar I A add-on in the months where open conversions are included in the LCR calculation.
  - In the months where there are more than 30 days to the date of payment, the SMCI shall take the 30 pct. of the whole amount of open conversions in an unlikely, but not inconceivable, stress scenario as a Pillar II A add-on.
- Intended effect:
  - The two Pillar II A add-ons will ensure a more stable liquidity requirement by creating a minimum amount of unencumbered HQLA which each SMCI must hold.

### Pillar II B add-on and refinancing risk
- Rationale:
  - At refinancing auctions there is a theoretical risk of demand being so low that the SMCI has to buy some of its own covered bonds to avoid the auction failing; this risk is not covered by the LCR requirement as refinancing takes place before the 30-day LCR window.
  - The Danish FSA will therefore introduce standardised liquidity stress testing across the SMCIs to ensure that the SMCIs have enough liquidity to handle such a situation; this stress test may result in a Pillar II B add-on.
  - Until now, this risk has in general been captured by the LCR floor requirement.
- Findings and incentives:
  - SMCIs which use HQLA to meet the requirement for supplementary collateral and the future OC requirement are better equipped to handle refinancing risks than SMCIs which meet these two requirements with assets which are not HQLA, such as their own covered bonds or warranties.
  - The more HQLA the SMCIs have, the better positioned they will be in a stress scenario, as it will be easier to sell HQLA to generate liquidity than own issued covered bonds.
- Calibration and application:
  - The Pillar II B add-on regarding refinancing risks is forward-looking and looks at the budgeted need for refinancing for the coming date of payment, which means it looks three months ahead.
  - SMCIs must generally assume that they will have to buy 30 pct. of their own covered bonds to avoid an auction failing.
  - If the SMCI does not have sufficient HQLA, the SMCI will have to set a Pillar II B add-on to ensure there is a sufficient amount of HQLA to be able to sell HQLA and buy its own covered bonds to avoid the auction failing.
  - SMCIs may use their own covered bonds to meet the future OC requirement and the requirement for supplementary collateral; if they buy back their own covered bonds at auction, they will still have to set a Pillar II B add-on to hold sufficient HQLA.
- Policy judgment:
  - In general, SMCIs will probably not have to set a Pillar II B add-on at the moment, as each SMCI is expected to use a sufficient amount of HQLA as supplementary collateral and to meet the future OC requirement.
  - The Danish FSA wishes to give SMCIs an incentive to use HQLA as supplementary collateral and to meet the future OC requirement, to ensure SMCIs hold a high amount of HQLA.
  - The Danish FSA considers that 30 percent reflects an appropriate risk for the SMCIs in going concern. There is a low risk that an auction will fail completely and that 100 pct. of the auction thus cannot be refinanced; in this case the maturity extension trigger will extend the covered bonds according to Danish law.
  - It is not realistic for SMCIs to have liquidity in advance to handle the full risk of a failed refinancing auction, as this would be too high a requirement and unnecessary given the likelihood of an auction failing.
- Overall:
  - The new Pillar II liquidity add-ons explicitly address risks not covered by the Pillar I LCR requirement in the form of unknown arrears, unknown open conversions and refinancing risks.

### Observation period and implementation timing
- Observation period:
  - There will be an observation period until the OC requirement comes into force.
  - The Pillar II add-ons shall be reported in the observation period from December 31th 2019 and until the entry into force.
  - The LCR floor requirement will remain during the observation period.
  - When the OC requirement comes into force, the Pillar II add-ons are expected to enter into force and the LCR floor requirement will be cancelled.
  - The current calibration of the Pillar II add-ons may be adjusted during the observation period if needed.

---

### Insurance and pension fund stress testing — scope and sample
- Coverage and sample:
  - The resilience of insurance companies was assessed through a top-down (TD) and bottom-up (BU) solvency stress test.
  - The tests covered five life insurers, six occupational pension funds and five non-life insurers, accounting for more than 70 percent in each sector, measured by gross written premiums.
  - Arbejdsmarkedets Tillægspension (ATP) was included; ATP is a compulsory pension scheme within the first pillar of the Danish pension system and is not subject to the solvency requirements of the Solvency II Directive.
  - The reference date for the stress test exercise was set to 30 June 2019.
- Scope of exercise:
  - The pre-COVID stress test was run at the solo entity level and thereby excluded business performed by foreign subsidiaries.
  - By focusing on the Danish entity, a fairly comprehensive view on relevant risks can be achieved given limited foreign subsidiary activity for most Danish groups.

### Insurance sector structure and exposures (findings)
- Business model and risk transfer:
  - Danish life insurers have been shifting from traditional to unit-linked policies, reducing their own market risk exposure; technical provisions for unit-linked account for one third of life insurers’ liabilities.
  - Danish insurers have reduced asset-liability risks and transferred some market risks to policyholders.
- Currency and asset allocation:
  - Between 40 and 50 percent of assets are not denominated in DKK (with FX exposures being mostly EUR); liabilities of life insurers and pension funds are almost exclusively denominated in DKK.
  - Structurally, insurers still have a currency mismatch between assets and liabilities, largely hedged through cross-currency swaps and forwards.
- Asset composition and credit quality:
  - Investment holdings are characterized by a high share in fixed-income securities, particularly in domestic mortgage bonds.
  - Corporate bond investments: 68 percent being rated A or better, and 94 percent being investment grade.
  - Sovereign bond exposures are dominated by domestic exposures (50 percent), but also include a sizable share in bonds issued by high-yield sovereigns in the Euro Area, mostly Italian and Spanish government bonds.
  - Overall credit quality is very high with around two thirds of corporate bonds being rated AAA (text also notes AAA/AA/A etc. in detailed breakdown).
- Property and other exposures:
  - Exposures to stocks and participations are higher than for the average EU firm; real estate exposures, both directly held and indirectly via investment funds, are rather low.

### Scenario specification and shocks
- Scenario design:
  - The scenario builds on the narrative and severity of the “market shocks” adverse scenario; it encompassed several market shocks resulting in a combined stress scenario considerably more severe than what was observed in 2020 Q1.
  - The scenario narrative centers around:
    - escalating global trade tensions and a tightening of global financial conditions which depress business and consumer confidence;
    - weaknesses in economic activity with a knock-on effect on domestic housing markets, funding pressures for the banking system, and the yields of covered bonds moving up in tandem with yields of sovereign bonds.
  - 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 maximum drawdown during the projection horizon is realized right after the reference date.
- Market risk shocks included:
  - Shocks to the risk-free interest rate, equity and property prices, credit spreads of corporate and sovereign bonds, losses on mortgage loans.
  - Given the increase of credit spreads in the scenario, the volatility adjustment (VA) increases following the Solvency II calculation method, offsetting to some degree the negative impact of the credit spread shock.
- Additional sensitivity exercises (reported separately; not added to scenario results):
  - Longevity: permanent 25 percent decrease in mortality rates, applied to all policies.
  - Mortality shock: permanent 15 percent increase in mortality rates, applied to all policies.
  - Pandemic event: temporary 35 percent increase in disability-morbidity rates and temporary 10 percent increase in mortality rates for the next 12 months.
  - Historic catastrophic event sensitivity: windstorm “Anatol” (1999) scaled to current exposures (occurrence once in 80 years).
  - Reinsurance recoverables from the participant’s five largest reinsurers (on a group basis).
  - Single-factor shock: default of the largest banking counterparty with haircuts applied (results not added to scenario outcomes):
    - 100 percent write-off for equity exposures, subordinated bonds and loans, uncollateralized OTC derivative exposures, and guarantees for counterparties’ liabilities.
    - 50 percent write-off for unsecured bonds and loans.
    - 15 percent write-off for mortgage bonds, deposits (beyond deposit guarantee scheme protection), loaned securities, and collateralized OTC derivative exposures.

### Capital standards, modelling assumptions and solvency findings
- Regulatory basis and valuation:
  - Solvency II (implemented in the EU in 2016) forms the basis for the insurance stress test; assets and liabilities are valued mark-to-market with allowed deviations for insurance liabilities (LTG measures and transitional measures).
  - In the BU exercise insurers provided balance sheet and solvency data with and without these measures, both before and after stress.
- Outputs and adjustments:
  - Main output: effect on available own funds, eligible for coverage of the Solvency Capital Requirement (SCR); participating undertakings were requested to recalculate the SCR after stress.
  - Loss-absorbing effects of technical provisions (via profit sharing) and deferred taxes reported separately; management actions only included if non-discretionary rules were in place at the reference date.
- Baseline solvency positions and capital quality:
  - All participants record SCR ratios before stress well above the regulatory threshold of 100 percent.
  - Market risks are the dominant component of insurers’ SCRs; diversification at the BSCR level is modest for life and pension companies but more relevant for non-life firms.
  - Loss-absorbing capacity of technical provisions (LAC TP) is very large for pension funds: of the aggregate diversified losses represented by the BSCR, the LAC TP covers approximately 80 percent, leaving the remaining 20 percent—the SCR—to be covered by own funds.
  - Quality of capital: almost 90 percent of eligible own funds of life companies are unrestricted Tier 1 capital, less than 1 percent is Tier 3; pension companies use only unrestricted Tier 1 capital instruments; non-life firms have a small share of Tier 2/3 but non-Tier 1 instruments account for almost 13 percent on aggregate.

*International Monetary Fund — Denmark (excerpt).*

### 65. The Long-Term Guarantee (LTG) measures and transitionals have only a minor role in

### 65. The Long-Term Guarantee (LTG) measures and transitionals have only a minor role in

### LTG measures and the Volatility Adjustment (VA)
- Danish companies use mainly the Volatility Adjustment (VA), including 13 of the stress test companies.
- As the level of the VA has declined in early 2019 due to a new calculation methodology, its impact on the SCR of the stress test companies was rather limited as of mid-2019:
  - Technical provisions are 0.1 percent lower with these measures.
  - Eligible own funds 0.6 percent higher.
  - The SCR 2.7 percent lower.
- Accordingly, the median SCR ratio of the life sample increases by 12 percentage points (188 instead of 176 percent).
- The VA impact by sector (median SCR increase):
  - Life: 12 percentage points.
  - Pensions: marginal or non-existing.
  - Non-life: marginal or non-existing.

_Source: IMF staff calculations based on company submissions._

### Bottom-up (BU) methodology
- Methodology aligned with recent EIOPA stress tests.
- Participating insurers calculated instantaneous impact of the scenario on their balance sheet and solvency on a best effort basis; simplifications allowed if marginal.
- Insurers reported:
  - Balance sheet, own funds, SCR composition.
  - Impact of LTG measures and transitionals, before and after stress.
- Three-year projections required under baseline and adverse scenarios, including:
  - Gross written premiums, gross claims, lapse rates, investment returns, net earnings, gross technical provisions, own funds, and the SCR.
- Projections had to align with the macrofinancial scenario.
- Market value of investments assumed to stay constant after occurrence of shocks at the reference date (instantaneous shock).
- Recovery in profitability and solvency driven solely by underwriting business and recurring investment income (interest, dividends and rents).

### Top-down (TD) methodology and assumptions
- TD test run by FSAP team to benchmark BU results using QRT inputs:
  - Balance sheet (S.02.01).
  - Asset-by-asset investment holdings (S.06.02).
  - Own funds (S.23.01).
  - Technical provisions (S.12.01) and cash-flow projections for upcoming 50 years (S.13.01).
- Shocks applied to investment assets and insurance liabilities per scenario.
- Haircuts applied to market value of directly held assets; look-through not applied for funds.
- Fixed-income assets re-valued with the stressed term structure (per currency).
- Technical provisions (except unit-linked) after stress approximated with stressed term structure including the VA.
- Unit-linked liabilities decline mirrored market value loss of underlying assets.
- Derivatives not modeled separately due to insufficient granular reporting quality.
- SCR re-calculation limited to selected risk modules:
  - Market risk: capital charges for equity risk, spread risk and property risk proportionately adjusted.
  - Equity risk capital charge corrected for symmetric equity adjustment changing from -1.7 to -10.0 percentage points after fall in equity prices; capital charge used for re-calculation is only 29 percent after stress.
  - Life underwriting risk and health underwriting risk assumed to change proportionately with technical provisions after stressed discount curve.
  - Counterparty default risk, non-life underwriting risk and operational risk assumed unchanged.
- Internal model users’ TD calculations and aggregation performed in simplified approach broadly in line with the standard formula.
- Management actions on discretionary benefits:
  - Assumed insurers would cut future discretionary benefits over a more medium term to limit lapses:
    - Benefits projected for years 5 to 8 after the reference date assumed decreased by 15 percent.
    - Benefits beyond this period assumed decreased by 25 percent.
- Loss-absorbing capacity of deferred taxes assumed to remain constant in adverse scenario (conservative).

### Results of the pre-COVID Scenario Analysis (aggregate and sectoral)
- Asset value declines (aggregate BU numbers):
  - Life sub-sample: assets decline by 11 percent.
  - Pensions sub-sample: assets decline by 7 percent.
  - Non-life sub-sample: assets decline by 5 percent.
- Ratio of assets over liabilities:
  - Life sector: declines marginally from 105 to 104 percent.
  - Non-life sector: declines from 124 to 119 percent.
- Pre-COVID TD calculations broadly confirm balance sheet impact, indicating slightly lower net asset values mainly driven by underestimation of derivative hedging mitigation.
- Median solvency ratios (pre-stress to post-stress):
  - Life sector: from 188 to 127 percent.
  - Non-life sector: from 233 to 207 percent.
- One life company drops below a solvency ratio of 100 percent; capital needed to restore full coverage would be a moderate fraction of the sample’s aggregate capital.
- Majority of pension companies: solvency ratio rises after stress; declines only marginally for others. Drivers:
  - Product-inherent buffers related to unguaranteed products (“conditional” guarantees) allowing management discretion.
  - Conservative asset allocation of pension funds in sample.

### Main drivers of solvency impact and SCR changes
- Most solvency impact attributable to increase in corporate bond spreads, in particular domestic sovereign and covered bonds.
- Partially offset by decreasing insurance liabilities due to slightly higher discount rates and increased VA.
- Life insurers would see capital requirements increase as stress reduces loss-absorbing capacity of technical provisions substantially.

### Medium-term projections and profitability
- First-year projected RoE after stress for most life insurers and pension funds: between -5 and -25 percent.
- Almost all companies expected to be back in positive RoE territory in the following year, though for life and pension sectors mostly in lower single digits.
- All sectors able to slowly rebuild capital after stress, especially non-life firms benefiting from profitable underwriting.
- Eligible own funds projection:
  - Significant drop right after stress, followed by slow rebuild; non-life firms most notable in recovery.

### Insurers’ likely management actions and market impact
- Insurers unlikely to take disruptive, immediate actions; rebalancing spread over weeks or months.
- Expected asset adjustments:
  - Life companies: de-risking, switch into sovereign and mortgage bonds; sell equity and below-investment-grade corporate bonds.
  - Non-life insurers: likely to rebalance and might expand investments in equity and property.
- Overall assessment: sector capable of withstanding simulated losses without materially amplifying market risks or tightening market liquidity.

### ATP (box summary)
- ATP shows considerable resilience in pre-COVID stress test; individual reserve requirement covered by a significant margin.
- Hedging portfolio closely offsets declining value of bond holdings as valuation of liabilities decreases accordingly.
- Bonus potential shrinks, driven mainly by higher spreads and shocks to equity and participations; stress reduces ATP’s profits, turning negative in first year and remaining flat in following two years if asset prices do not recover.
- ATP not subject to Solvency II; required capital calculated using a 3-months expected shortfall with a 99 percent confidence level.
- ATP scale and structure:
  - Over 5.2 million members, one million pensioners.
  - Pension contributions divided: guarantee contributions 80 percent; bonus contributions 20 percent.
- Selected financials and metrics (DKK million, unless stated otherwise):
  - Contributions: 9,049 (2014), 9,055 (2015), 9,572 (2016), 9,703 (2017), 9,871 (2018), 10,061 (2019).
  - Pension benefits: 13,661 (2014), 14,566 (2015), 15,454 (2016), 16,075 (2017), 16,878 (2018), 17,054 (2019).
  - Investment return: 116,752 (2014), 6,584 (2015), 60,149 (2016), 15,879 (2017), 24,108 (2018), 107,340 (2019).
  - Net profit: 2,481 (2014), 5,402 (2015), -785 (2016), 17,239 (2017), -25,585 (2018), 33,893 (2019).
  - Bonus potential: 95,831 (2014), 101,242 (2015), 100,454 (2016), 117,695 (2017), 92,086 (2018), 125,980 (2019).
  - Total pension provisions: 704,423 (2014), 705,214 (2015), 759,251 (2016), 768,576 (2017), 785,459 (2018), 885,608 (2019).
  - Total assets: 812,433 (2014), 781,228 (2015), 869,746 (2016), 893,483 (2017), 906,735 (2018), 1,024,472 (2019).
  - Return before tax on pension savings returns (in percent): 23.3 (2014), 1.1 (2015), 10.1 (2016), 2.5 (2017), 3.7 (2018), 16.2 (2019).
  - Bonus rate (in percent): 15.8 (2014), 16.8 (2015), 15.2 (2016), 18.1 (2017), 13.3 (2018), 16.6 (2019).
- ATP asset allocation (in percent):
  - Government and mortgage bonds: 31 percent.
  - Real estate: 14 percent.
  - Infrastructure: 13 percent.
  - Private equity: 9 percent.
- ATP disclosed sensitivities:
  - A 25 percent decline in equity prices would cause a loss (before tax) of around DKK 55 bn.
  - A 1 percentage point increase in interest rates is relevant; spread increases or commodity/inflation changes less concerning.

### Sensitivity analyses (pre-COVID)
- Longevity risk:
  - Strong increase in life expectancy reduces median SCR ratio of life insurers and pension funds to 210 percent (-37 percentage points).
- Higher mortality rates (including from a severe epidemic anticipated before COVID) reduce annuity payments; coverage for sickness pay and hospitalization only moderate.
- Catastrophic non-life events:
  - Repetition of windstorm Anatol (1999) would result in reduction of non-life insurers’ own funds of only 1 percent due to reinsurance.
- Further TD calculations reveal vulnerabilities from:
  - Lower interest rates (specifically for pension funds).
  - To a lesser degree, higher covered bond spreads.
- Solvency II standard formula insights:
  - Market risk sensitivities shown both before (gross) and after (net) profit sharing with policyholders.
  - Lower future discretionary benefits significantly reduce risks for life insurers and pension funds.

_Source: IMF staff calculations based on company submissions._

### 79. Further vulnerabilities exist related to lower interest rates, higher covered bond

### 79. Further vulnerabilities exist related to lower interest rates, higher covered bond spreads, and stock market declines, but losses would partially be borne by policyholders

### Stress-test findings and sensitivity analyses
- Topline scenario: A severe downward shock to the risk-free interest rate term structure (−100 basis points parallel shift for maturities up to 20 years) would cause a substantial decline in pension funds’ available capital, with the ratio of assets over liabilities moving close to 100 percent (TD calculation).
- Importance of hedging: The TD calculation underestimates the mitigating impact from derivative hedging; comparison with the BU exercise demonstrates the importance of pension funds’ interest rate hedging.
- Covered bond spreads shock: A 200 basis points increase in domestic covered bond spreads would have a much smaller impact because the effective durations of these holdings are relatively short.
- Equity and other market shocks are considered in standard formula sensitivities (value-at-risk with a 1-year horizon and a 99.5 confidence level); some prescribed shocks cause substantial declines in net asset value.
- Loss absorption by policyholders: Reduced future discretionary benefits (profit sharing / loss-absorbing capacity of technical provisions) make substantial shocks more manageable in reported SCR results.

### Low-yield environment and profitability pressures
- Profitability history: Positive returns on equity in both life and non-life insurance throughout 2009-2018 have been maintained, but profitability in the life sector is challenged by low interest rates.
- Legacy guarantees: A large (though constantly declining) legacy portfolio of contracts with high guarantees remains. Contracts offering guarantees of more than 3 percent account for 23 percent of all life technical provisions, and many of these contracts still have remaining maturities of more than ten years.
- Guaranteed-rate structure: For longer maturities, interest rate guarantees on life policies are continuously declining, but a sizable portion of legacy business with guarantees above 3 percent might continue to weigh on life insurers’ profitability well into the 2030s.

### Reinvestment and duration risks
- Current coupons: Bonds expiring over the next five years mostly pay coupons between 1 and 2 percent.
- Average coupon: With an average coupon rate already below 1.9 percent, Danish insurers have only moderate reinvestment risks compared to EU peers.
- Reinvestment projections: The projection lines assume reinvestment at rates of 1.5, 1.0 and 0.5 percent.
- Business mix vulnerability: Insurers more active in non-life and unit-linked life business are less affected and could sustain the low-yield environment for a prolonged period; companies with a high legacy stock of guarantees are likely to experience a drain on profitability.

### Regulatory reporting and supervisory recommendations
- Macroprudential stress-testing:
  - The DFSA, in cooperation with Danmarks Nationalbank, is encouraged to further expand regular macroprudential stress tests.
  - Tests should blend bottom-up and top-down perspectives.
  - Results should be used to challenge companies’ Own Risk and Solvency Assessments (ORSA) and underlying projections for future business, specifically expectations for premium growth and investment returns.
- Data quality and supervisory reporting:
  - The DFSA should intensify data quality checks and enforce high-quality supervisory reporting.
  - Supervisory reporting should be thoroughly scrutinized; companies should be required to re-submit corrections of misreported data without undue delay.
  - The DFSA should require at least the annual QRTs to be audited, and require audit assurance for the systems and procedures used to complete the QRTs.

### Non-Financial Corporate (NFC) sector: recent trends and vulnerabilities
- Coverage and data:
  - Analysis used disaggregated firm-level data on Danish NFCs for the period 2003-2017 from the Danish Central Business Register.
  - Firm size defined as: small (<50), medium (50-249), and large (>250).
  - NFCs categorized into 8 sectors: agriculture, building and construction, information, manufacturing and extraction, real estate, trade, transport, and other corporates; firms in finance and insurance excluded.
- Debt dynamics:
  - The NFC debt-to-GDP ratio in Denmark remained stable with a slight downward trend from 2015 to 2018 and is below debt ratios of peer European countries.
  - The debt service ratio has been declining since 2013, influenced by the low level of interest rates.
  - Total NFC debt in 2017: 1,400 billion DKK.
  - Sector concentration: (Commercial) real estate, manufacturing industries (including extraction) and trade account for 70 percent of total NFC debt.
  - Small firms: The share of debt concentrated in small firms has increased steadily for the past decade and is now above 60 percent of total debt.
  - (Commercial) real estate represents 30 percent of total NFC debt and firms in this sector are highly leveraged and atomistic.
- Firm-level indicators:
  - ROE ratios rebounded across all sectors and firm sizes after the 2009 decline, but have not returned to pre-crisis levels for small and medium firms.
  - Liquidity ratios improved across the board.
  - Debt-to-equity ratios have been declining for all firm sizes and most sectors; (commercial) real estate and transport have the highest leverage ratios.

*Source: IMF staff calculations based on company submissions.*

### 87. Measures of debt-at-risk (DaR) and firms-at-risk (FaR) based on the interest coverage

### Measures of debt-at-risk (DaR) and firms-at-risk (FaR) based on the interest coverage ratio (ICR)

### Debt service capacity and ICR findings
- Interest cover ratio (ICR) defined as the ratio of EBITDA to interest payments due in the same period; higher ICRs indicate stronger debt servicing capacity.
- ICR thresholds of 1.5 or 2 are typically used to identify NFCs that are at risk of not being able to service their debts.
- NFC ICRs in Denmark have gradually recovered from the low levels reached during the GFC, but the recovery has been uneven across firms by sector and size.
- Small firms still exhibit relatively low ICR levels; the 25th percentile of the ICR distribution corresponds to a small negative number (indicating that at least 25 percent of the firms have zero or negative earnings).
- In 2017:
  - 44 percent of outstanding debt was held by firms with ICR below 2.
  - 24 percent of the debt was held by firms with negative earnings.
  - In 2009 the share of debt held by firms with negative earnings was 37 percent.

### Key firm-level statistics (median values by year and firm size)
- Return on equity (median, percent)
  - All firms: 2007: 15.9; 2009: 4.4; 2011: 10.4; 2013: 9.9; 2015: 11.5; 2017: 11.5
  - Large firms: 2007: 17.9; 2009: 12.8; 2011: 15.6; 2013: 12.7; 2015: 16.6; 2017: 17.1
  - Medium firms: 2007: 29.8; 2009: 12.5; 2011: 18.2; 2013: 18.5; 2015: 22.4; 2017: 22.9
  - Small firms: 2007: 15.3; 2009: 4.1; 2011: 10.1; 2013: 9.7; 2015: 11.2; 2017: 11.1
- Liquidity ratio (median, ratio)
  - All firms: 2007: 1.23; 2009: 1.19; 2011: 1.21; 2013: 1.23; 2015: 1.27; 2017: 1.32
  - Large firms: 2007: 1.20; 2009: 1.17; 2011: 1.17; 2013: 1.23; 2015: 1.24; 2017: 1.25
  - Medium firms: 2007: 1.25; 2009: 1.26; 2011: 1.29; 2013: 1.33; 2015: 1.37; 2017: 1.36
  - Small firms: 2007: 1.23; 2009: 1.19; 2011: 1.21; 2013: 1.22; 2015: 1.26; 2017: 1.32
- Debt-to-equity Ratio (median, ratio)
  - All firms: 2007: 1.97; 2009: 1.70; 2011: 1.71; 2013: 1.73; 2015: 1.70; 2017: 1.59
  - Large firms: 2007: 1.75; 2009: 1.79; 2011: 1.63; 2013: 1.50; 2015: 1.66; 2017: 1.58
  - Medium firms: 2007: 2.23; 2009: 1.95; 2011: 1.89; 2013: 1.84; 2015: 1.73; 2017: 1.65
  - Small firms: 2007: 1.95; 2009: 1.68; 2011: 1.71; 2013: 1.72; 2015: 1.70; 2017: 1.59

### Debt growth and sectoral vulnerability
- Debt growth has been substantial in sectors with enhanced vulnerability—characterized by high leverage, small and numerous firms, and below-average interest coverage ratios (example: real estate).
- Mid-sized banks and MCIs have increasingly relaxed lending standards; DN analysis suggests riskiness of credit allocation in the corporate sector has been increasing since 2013.

### Empirical analysis of corporate defaults and default-probability modeling
- Default rates of NFCs reached historic lows after spiking during the GFC; between 2008-2010 there was a sharp and widespread rise in bankruptcies across sectors.
- Post-GFC downward trend in defaults reversed temporarily in 2015-16 due to actions by the Danish Business Authority to identify and close fraudulent/unregistered enterprises.
- Historical data reveal a positive correlation between leverage (defined as the ratio of debt-to-assets) and default rates within sectors.
- The authorities and the IMF team estimated default probability models using granular micro data to address:
  - (A) Relation between corporate debt levels and probabilities of default (PD) of Danish firms.
  - (B) How tightening in macro-financial conditions (shocks) affect PDs.
- Pre-COVID modeling results:
  - Debt-at-risk (defined as debt issued by firms with projected PDs greater or equal to 2 percent) as a percentage of total NFC debt would increase from 5.3 percent to 13.1 percent by 2022 in the pre-COVID adverse scenario.
  - The commercial real estate sector appears especially sensitive; a six-fold jump increase is projected for this sector in the adverse scenario.
- Pre-COVID baseline scenario (based on October 2019 WEO projections) assumed growth rates: 2020: 1.9 percent; 2021: 1.7 percent; 2022: 1.6 percent; 2023-24: 1.5 percent per year.

### Correlated default analysis and simulation results
- Ignoring correlation in defaults can underestimate portfolio-level credit risk; diversification benefits decline as default correlations increase.
- Default correlation models used:
  - Step 1: Compute industry-specific default rates pi and within- and cross-industry default correlations Diiρ and Dijρ using historical data for 2003-2018.
  - Step 2: Generate distributions of firm defaults in portfolios using default rates and asset correlations; calibrate asset correlation parameters to match historical moments (average default rates and correlations).
- Simulations (10,000 iterations) of loan portfolios comprising 50 and 200 firms were run under three parameter sets:
  1. Firm-specific PDs and cross-firm correlation parameters matching historical data.
  2. Stressed PDs (3 times the historical default rate) with historical correlation.
  3. Historical PDs with stressed correlations (3 times the historical correlation).
- Simulations show that default correlation shocks affecting well diversified loan portfolios could trigger significant credit losses even if PDs remain low.

### Banking sector stress testing matrix — key points
- Institutional perimeter includes:
  - 7 systemic credit institutions: Danske Bank Group, Jyske Bank Group, Nykredit Realkredit Group, Sydbank, SparNord Bank, DLR Kredit, Nordea Kredit.
  - 5 commercial banks (unconsolidated): Danske Bank, Nykredit Bank, Jyske Bank, Sydbank, Spar Nord Bank.
  - 6 mortgage credit institutions (unconsolidated): Realkredit Danmark, Nykredit Realkredit, Totalkredit, Jyske Realkredit, DLR Kredit, Nordea Kredit.
- The 7 largest systemic credit institutions cover approximately 77 percent of the total assets for credit institutions.
- Data source and baseline date:
  - Bank-by-bank supervisory data; Baseline Date: 30 September 2019.
- Channels of risk propagation and methodologies:
  - Balance sheet-based approach.
  - Credit risk: Loss calculations based on panel regression model for logit transformed sectoral impairment rates with regressors including unemployment rate and its lag, GDP growth rate and its lag, short term interest rate and its lag; parameter estimates account for sector riskiness and bank portfolio composition.
  - Market risk: Full re-valuation of individual securities incorporating hedges; IMF provided yield curve shocks and DN stress team performed security-level calculations.
  - Net interest income: Gap analysis based on granular asset/liability data; cost of funding shocks defined by liability type and affect banks differently depending on funding composition.
  - Pre-impairment income assumed to stay at level observed for last 4 quarters (2018Q4-2019Q3) with non-performing loans generating no income.
- MCIs affected by: materialization of credit risk; materialization of market risk leading to downward repricing of securities in capital centers; adverse impact on net interest income when MCIs issue junior bonds to top-up collateral at a negative spread vis-a-vis sovereign bonds.
  - MCIs’ credit risk loss calculations based on time series regression for logit transformed aggregate impairment rates with regressors including GDP growth and its lag, unemployment and its lag, short-term interest rate and its lag, inflation and its lag.
  - Top-up collateral needs calculated when housing prices decline and LTV ratios exceed 80/60 percent limits.
- Credit growth in stress tests assumed to grow with the nominal GDP growth rate, floored at 0.
- Stress test horizon: Five years (2020–2024).

*Source: IMF staff analysis and DN micro data as presented in the chapter "Measures of debt-at-risk (DaR) and firms-at-risk (FaR) based on the interest coverage ratio (ICR)".*

### 3. Tail shocksScenario analysis

### 3. Tail shocksScenario analysis

### Scenario design and purpose
- Scenario-based tests assess the impact of shocks on the entire portfolio of assets, including those placed in the banking and trading books.
- Market risk yield curve shocks and funding shocks are calibrated based on what has happened so far in response to COVID-induced stress and observed market movements at the time.

### COVID central scenario (banking sector)
- GDP: sharp fall in GDP of 8 percent in 2020, rebound of 6 percent in 2021.
- Comparison: contraction in GDP for 2020 is worse than what is projected in the IMF World-Economic-Outlook (WEO) for April.
- Cumulative decline: results in a cumulative decline of real GDP relative to the September 2019 WEO projections equivalent to 1.8 standard deviations (6.1 percentage points) over two years.
- Unemployment: goes up to 6.8 percent in 2020, gradual recovery reaching 5.6 percent in 2023.
- Main driver of losses: materialization of credit risk (predominantly influenced by GDP growth and the unemployment rate).
- Funding cost shocks for 2020: 30 basis points for interbank lending and riskier debt securities; 15 basis points for covered bonds and repos.
- Impact of funding shocks: limited in the COVID scenarios because these shocks were considerably less severe than in the “market shocks” adverse scenario.
- Yield curve shocks: broadly in line with what has happened so far in response to the COVID-induced stress.

### COVID prolonged scenario (banking sector)
- Path: instead of a rebound in 2021, the economy continues to contract by 2.1 percent before recovery ensues in 2022.
- Cumulative decline: results in a cumulative decline of real GDP relative to the September 2019 WEO projections equivalent to 4 standard deviations (13.6 percentage points) over two years.
- Unemployment: peaks at 11.8 percent in 2022.
- Main driver of losses: materialization of credit risk (predominantly influenced by GDP growth and the unemployment rate).
- Funding cost shocks for 2020: 30 basis points for interbank lending and riskier debt securities; 15 basis points for covered bonds and repos.
- Yield curve shocks: broadly in line with observed COVID stress responses.
- Note: COVID central scenario is still subject to downside risks; prolonged scenario acknowledges these looming risks.

### Market shocks adverse scenario (banking sector)
- Generation: triggered by a series of global and domestic shocks using IMF’s Flexible System of Global Models (RES) as input; driven by combination of external and domestic shocks amplified by existing vulnerabilities.
- Main drivers: escalating global trade tensions and a tightening of global financial conditions depressing business and consumer confidence across advanced and EM economies.
- Abrupt decompression of risk premia: particularly impacts “high spread economies” in Europe; reflected in higher money market interest rates, banks’ cost of short-term funding, and long-term government bond yields.
- Housing: housing prices decline by 18 percent over three years, adversely impacting domestic consumption and residential investment.
- Funding pressure: heavy reliance on wholesale funding by Danish banks and MCIs makes them vulnerable to market turmoil; overall cost of funding increases by 100 basis points, with partial pass-through to lending rates.
- Exchange rate: scenario assumes that the exchange rate regime remains credible despite domestic and international stress.
- Cumulative GDP decline: results in a cumulative decline of real GDP relative to the October 2019 WEO projections equivalent to 2.5 standard deviations (8.6 percentage points) over two years.

### Behavioral adjustments and constraints used in tests
- Balance sheet assumptions:
  - Balance sheets are assumed to be quasi-static—the rate of loan growth and funding growth are tied to nominal GDP growth.
  - Asset disposal and acquisition is not permitted.
  - External capital injections are not permitted.
  - Non-performing loans are not permitted to generate interest income.
- Dividend policy:
  - Banks are assumed to make dividend pay-outs of 40 percent for periods with positive net income and no pay-outs in case of negative income.
- Taxation:
  - Banks are assumed to pay the same tax rate calibrated as the average of tax rate for the five large banks during the last four quarters. Same calibration exercise is undertaken for the MCIs as well.

### Sensitivity analysis (banking sector)
- Purpose: evaluates the impact of shocks to single risk factors in a static manner.
- Tests include:
  - Simultaneous default of 3, 5, and 10 largest single exposures with several LGD assumptions.
  - A significant increase to the RWAs as a result of Basel-III output floors.

### Calibration of risk parameters and regulatory standards
- Credit risk – Expected Losses:
  - Credit risk satellite model projects future impairment rates where, using Frye-Jacobs LGD function, impairment rate projections are split into Point-in-time PD and LGD parameters such that PD*LGD = Impairment rate projection.
- Credit risk – RWAs:
  - Convert shocks to the PIT PD parameters to through-the-cycle (TTC) PD parameters for projection of RWAs.
- Market risk:
  - Yields for sovereign and corporate bonds determined based on historical statistical relationships with other variables available in the scenario.
- Regulatory/accounting:
  - Capital definition follows national implementation by DFSA. Data on combined CET1, including the CCB, SIFI, pillar II and CCyB buffers are available from DFSA for 2020 Q1.
  - RWA behave dynamically according to changes in credit risk parameters.

### Reporting format for results (banking sector)
- Output presentation includes:
  - Evolution of CET1 capitalization ratios for the consolidated banking groups.
  - Decomposition of the reduction in CET1 capital ratio in terms of drivers (credit risk, market risk, interest rate risk).

---

### Insurance sector: Tail shocks and sensitivity (Appendix II STeM)
- Institutions included (examples):
  - 5 life insurers: AP Pension, Danica Pension, PFA Pension, TopDanmark Livsforsikring, Velliv
  - 6 occupational pension funds: PenSam Liv, PK Farmakonomer, PK Socialrådgivere, PK Sundhedsfaglige, PK Sygeplejersker, Sampension
  - 5 non-life insurers: Alm. Brand Forsikring, Codan Forsikring, Lærerstandens Brandforsikring, TopDanmark Forsikring, Tryg Forsikring
  - 1 pension fund: ATP
- Market share:
  - Life and occupational pension funds: >70 percent of total balance sheet assets
  - Non-life: >70 percent of total balance sheet assets
- Reference date: June 30, 2019
- Methodology:
  - Investment assets: market value changes after price shocks, affecting the solvency position.
  - Sensitivity analysis: effect on available capital and solvency position.
  - Time horizon: instantaneous shock for bottom-up; 3-year projection (baseline and adverse) for top-down.
- Adverse scenario shock parameters:
  - Risk-free interest rates (without volatility adjustment): +/-0 bps (1y DKK), +25 bps (10y DKK)
  - Sovereign bond spread: +110 bps (Denmark), +40 bps (AAA-rated euro area countries), up to +250 bps (high-yield euro area countries)
  - Stock prices: -26.5 percent (domestic), -20 percent (other advanced economies), -25 percent (emerging and developing economies)
  - Property prices: -22 percent (domestic residential), -26.4 percent (domestic commercial), -7.7 and -9.2 percent (other countries, residential and commercial, respectively)
  - Domestic corporate bond spreads (financials): between +100 bps (AAA) and +300 bps (B and lower)
  - Domestic covered bond spreads: +80 bps (AAA)
  - Foreign corporate bond spreads (financials): between +50 bps (AAA) and +300 bps (B and lower)
- Sensitivity shocks considered:
  - Longevity shock (life and pensions only): permanent 25 percent decline in mortality rates
  - Mortality shock (life and pensions only): permanent 15 increase in mortality rates
  - Pandemic event (life and pensions only): temporary 35 percent increase in disability/morbidity rates, temporary 10 percent increase in mortality rates
  - Mass lapse event (life only): 20 percent lapse rate in contracts where discontinuance results in a loss for the insurer
  - Catastrophic events (non-life only): repetition of Windstorm Anatol (December 1999) based on today’s exposures
  - Default of largest banking counterparty
  - Lower interest rate term structure: -100 bps parallel shift (for all currencies)
  - Spread shock to domestic covered bonds: +250 bps
- Risks/factors assessed:
  - Market risks: interest rates, share prices, property prices, credit spreads
  - Underwriting risks: longevity, mortality, pandemic event, catastrophic events
  - Credit risks (top-down): default of largest financial counterparty
  - Summation of risks with no diversification effects
- Buffers:
  - Product-specific buffers and loss-absorption capacity stemming from policyholder participation
- Behavioral adjustments:
  - Bottom-up: management actions limited to non-discretionary rules in place at the reference date
  - Top-down: none
- Regulatory/accounting standards:
  - Solvency II and National GAAP
- Output presentation for insurance:
  - Impact on solvency ratios (including and excluding long-term guarantee measures)
  - Impact on net income
  - Contribution of individual shocks
  - Dispersion measures of solvency ratios and net income
  - Survey-based reactions and management actions of insurers after stress

---

### Key risks from the Risk Assessment Matrix (Appendix III)
- Sharp rise in risk premia
  - Overall level of concern: High
  - Impact: Abrupt deterioration in market sentiment could trigger protracted correlated decline in asset prices, erode household wealth, weigh on consumption growth, raise funding costs in the Danish covered bond market, and create (re-)financing problems, especially for MCIs.
- Weaker-than-expected global growth
  - Overall level of concern: High
  - Impact: Sustained negative impact on Danish exports, investment, consumption, and growth; amplified by rising trade tensions and a disorderly Brexit; corporate-sector losses would be accompanied by rising unemployment, NPLs, and loan loss impairments.
- Financial shocks originating in the Nordic-Baltic region
  - Overall level of concern: Medium
  - Impact: Losses for major Danish SIFIs due to presence in those countries; amplified by regional interlinkages and potential deleveraging; ML cases and fines could disrupt banking groups’ capacity and profitability.
- Prolonged downturn in the domestic real estate market
  - Overall level of concern: Medium
  - Impact: Decline in real estate prices could drag down other asset classes, dampen consumption and investment, increase unemployment, raise NPLs and impairment losses, and trigger credit rationing.

---

### Banks: satellite models and methodologies for profit and credit loss projections (Appendix IV)
- Impairment rate (LLR) model:
  - Estimated for 9 sectors using historical data from 1992 to 2018.
  - Sectors: (1) agriculture, (2) Manufacturing industries, (3) Building and construction, (4) Trade, (5) Transportation, (6) Finance and insurance, (7) Real estate, (8) Other corporates, (9) Households.
  - Dynamic panel specification (logit of ImpRate) includes sector fixed effects, lagged impairment rate, contemporaneous and one-year-lagged GDP growth rate, unemployment rate, and short-term deposit rate.
- Estimated coefficients (exact):
  - Impairment rate (logit transform, 1 year lag): 0.193
  - Unemployment Rate: -0.060
  - Real GDP growth rate: -0.321
  - Short-Term deposit rate: -0.050
  - Unemployment Rate (1 year lag): 0.292
  - Real GDP growth rate (1 year lag): -0.094
  - Short-Term deposit rate (1 year lag): 0.118
- Model adjustments:
  - Intercept adjustment: sector-specific intercepts recalibrated using bank and industry level data for 2011–2018 to better represent banks included in stress tests.
  - Lagged dependent variable at starting point: use bank-specific last year’s impairment rate for projections to reflect starting point heterogeneity.
- Obtaining stressed PDs and LGDs:
  - Satellite model produces stressed impairment rate projections (ImpRate ≈ PD*LGD) but not separately PD and LGD.
  - Second condition: use Frye-Jacobs (FJ) LGD function to relate conditional LGD (cLGD) to conditional PD (cPD).
  - FJ relation (conceptual): cLGD = Φ[Φ−1[cPD] − k] / cPD, where k = Φ−1[PD] − Φ−1[PD×LGD]; asset correlation parameter set to zero.
  - Solved forms used:
    - cPD = Φ[Φ−1[Imp.rate] + k], where k = Φ−1[PD] − Φ−1[PD×LGD]
    - cLGD = ImpRate / cPD
  - Reference evaluation example: Frye-Jacobs LGD Function evaluated at reference point PD = 2 percent, LGD = 45 percent.

*Source: 1dnkea2020009 - 3. Tail shocksScenario analysis*

### Appendix V. Contributions to Changes in Capital—Adverse

### Appendix V. Contributions to Changes in Capital—Adverse Scenarios

### Decomposition of changes in the CET1 capital adequacy ratio (CAR)
- Define CAR = Capital / RWA.
- Change in CAR over time:
  - 훥CAR푡+1 = CAR푡+1 − CAR푡 = 훥Capital푡+1 / RWA푡 − 훥RWA푡+1 ⋅ CAR푡 / RWA푡+1 − 훥Capital푡+1 ⋅ 훥RWA푡+1 / (RWA푡 ⋅ RWA푡+1).
- Interpretation of terms:
  - First term: (partial) contribution of changes in capital (numerator) to the change in the CAR ratio.
  - Second term: contribution of changes in RWA (denominator) to the change in the CAR ratio.
  - Third term: joint contribution of changes in RWA and capital; generally much smaller and can be added to RWA, capital, or both contributions.
- For chart construction, the joint effect (third term) is added to the contribution of changes in RWA:
  - Contribution of ΔCapital to ΔCAR = 훥Capital푡+1 / RWA푡
  - Contribution of ΔRWA to ΔCAR = − 훥RWA푡+1 ⋅ CAR푡 / RWA푡+1 − 훥Capital푡+1 ⋅ 훥RWA푡+1 / (RWA푡 ⋅ RWA푡+1)

### Decomposition of capital evolution (assuming no capital injections)
- Capital change identity:
  - Capital푡+1 − Capital푡 = 훥Capital푡+1 = Profit푡+1 − Dividends푡+1
  - Profit푡+1 = Net interest income푡+1 + Credit losses푡+1 + Net gains on securities푡+1 + Other items푡+1 − Taxes푡+1
- Alternative expression:
  - Profit푡+1 = Pre-impairment profit푡+1 + 훥Net interest income푡+1 + 훥Net gains on securities푡+1 + Credit losses푡+1 + 훥Other items푡+1 − Taxes푡+1
- Final decomposition used:
  - 훥Capital푡+1 = Pre-impairment profit푡+1 + 훥Net interest income푡+1 + 훥Net gains on securities푡+1 + Credit losses푡+1 + (훥Other items푡+1 − Taxes푡+1 − Dividend푡+1)
- Sources of change in capital (numerator) and their contribution to CAR:
  - Pre-impairment profits (Pre-impairment profit푡+1)
  - Losses or gains due to change in net interest income (훥Net interest income푡+1)
  - Losses or gain due to market risk (securities) (훥Net gains on securities푡+1)
  - Credit losses (Credit losses푡+1)
  - Taxes and other items (훥Other items푡+1 − Taxes푡+1 − Dividend푡+1)
- Change in RWAs contribution reiterated:
  - Change in RWAs = − 훥RWA푡+1 ⋅ CAR푡 / RWA푡+1 − 훥Capital푡+1 ⋅ 훥RWA푡+1 / (RWA푡 ⋅ RWA푡+1)

### RWA calculation inputs and through-the-cycle adjustments
- For RWAs, Basel Advanced Approaches regulatory capital formulas are used.
- Through-the-cycle adjustments applied to all PDs/LGDs:
  - 훥푃퐷푅퐸퐺 = 0.2 훥푃퐷푃퐼푇
  - 퐿퐺퐷푅퐸퐺 = 퐿퐺퐷푑표𝑤𝑛𝑡𝑢𝑟𝑛

---

### Appendix VI. MCIs: Satellite Models and Methodologies for Stress Testing

### Credit loss projections — impairment rate model for Danish mortgage sector
- Sample and period:
  - Historical data from 1980 to 2018.
- Regression specification:
  - logit(ImpRate푡) = 훼0 + 훾 × logit(ImpRate푡−1) + 훽 × M푡 + 휀푖,푡
  - t indicates the year.
- Definition of ImpRate:
  - ImpRate measured as the ratio of impairment charges (flow) to total loans and accumulated losses.
  - Note: In few instances observed flows of impairment charges can be negative; negative observations were floored at a small positive value because the logit transformation doesn’t exist for negative values.
- Explanatory variables (M푡):
  - Both contemporaneous and first lag of GDP growth rate, unemployment rate, short-term interest rate and inflation.
  - Inclusion of lagged dependent variable captures starting-point effects (e.g., an MCI entering stress with a high impairment rate suffers bigger losses).
- Implementation detail:
  - Use average impairment rate for each MCI calculated using data for 2017, 2018 and 2019Q3.

### Top-up collateral calculations for MCIs
- LTV thresholds by loan type:
  - 60 percent for Manufacturing/ Office and Business/ Agricultural properties.
  - 80 percent for all other sectors.
- If LTV limit breached due to house price shock, MCIs must provide top-up collateral to return portfolios to permissible limit.
- Calculation steps:
  - Assign mortgages of each MCI in 2019Q3 into 80/60 LTV limit using HTT data.
  - Distribute loans equally in 1 unit LTV buckets 0-1, 1-2, 2-3, ..., 99-100, 100-101, and apply mid-points 0.5, 1.5, 2.5, ..., 99.5, 100.5 to compute initial house values.
  - Apply house-price shock to house values and compute new LTV = loan amount / shocked house value.
  - Account for loan growth using scenario credit growth rates, floored at 0.
  - Top-up collateral amount = loan amount − (LTV limit) * shocked house values.
  - Repeat for years 2020-2024.
  - Assume MCIs already hold required amount of top-up collateral at starting point (2019Q3).
  - Top-up requirement = calculated requirement for each year − initial (starting point) collateral.
  - Adjust for excess collateral available at MCIs; floor top-up collateral requirement at 0.
- Cost adjustment to NII via spreads (applied to additional requirement each year):
  - COVID baseline spreads for years 2020 through 2024: 60 bps, 50 bps, 30 bps, 10 bps, and -10 bps, respectively.
  - Pre-COVID “market shocks” scenario spreads for years 2020 through 2024: 270 bps, 240 bps, 230 bps, 15 bps, and 40 bps, respectively.

---

### Appendix VII. Group-Level Consolidation

### Projection of RWAs for banking groups — three-step approach
- Step 1:
  - At starting point, maintain subsidiary RWAs as reported and adjust parent RWAs so that sum matches reported group-level RWAs.
  - Note: parent entity definitions — the parent entity is the bank in Danske and Jyske groups; the MCI is the parent entity in the Nykredit Group.
- Step 2:
  - Calculate weights for parent and subsidiary using adjusted RWAs:
    - 푊푠푢푏 = 푅푊퐴푠푢푏 / 푅푊퐴퐺푟표푢푝
    - 푊푝푎푟푒푛푡 = 푅푊퐴푝푎푟푒푛푡,푎푑푗푢푠푡푒푑 / 푅푊퐴퐺푟표푢푝
- Step 3:
  - Calculate group RWA growth rates as weighted averages of entity growth rates:
    - (푅푊퐴푔푟표푢푝,푡+1 / 푅푊퐴푔푟표푢푝,푡) = 푊푠푢푏 × (푅푊퐴푠푢푏,푡+1 / 푅푊퐴푠푢푏,푡) + 푊푝푎푟푒푛푡 × (푅푊퐴푝푎푟푒푛푡,푡+1 / 푅푊퐴푝푎푟푒푛푡,푡)
  - Apply sequence of growth rates to starting group-level RWAs (as reported) to generate future RWA path.

---

### Liquidity roll-off and roll-in rates (selected notes and highlights)
- Definitions:
  - Run-off rate: fraction of liability amount maturing in a period that is withdrawn by claim holders; sight deposits and contingent liabilities are exceptions.
  - Roll-off rate: fraction of asset amount maturing in a period that is converted into a cash inflow by the bank.
  - All run-off rates highlighted in blue are defined with respect to the initial outstanding balance.
- Selected entries from maturity-ladder inflows table (roll-off rates, In percent):
  - Level 1 tradable assets: 100 across all time buckets shown (Overnight through Greater than 9 months up to 12 months).
  - Level 1 (CQS1): 100 across all time buckets shown.
  - Level 1 (CQS2, CQS3): 100 across all time buckets shown.
  - Level 1 (CQS4+): 100 across all time buckets shown.
  - Level 1 covered bonds (CQS1): 100 across all time buckets shown.
  - Level 2A corporate bonds (CQS1): 100 across all time buckets shown.
  - Level 2A covered bonds (CQS1, CQS2): 100 across all time buckets shown.
  - Level 2A public sector (CQS1, CQS2): 100 across all time buckets shown.
  - Level 2B tradable assets and subcategories (ABS, covered bonds, corporate bonds, shares, public sector): 100 across all time buckets shown.
  - Other tradable assets: 100 across all time buckets shown.
  - Other assets: 100 across all time buckets shown.
  - Monies due not reported in 2.1 resulting from loans and advances granted to Retail customers: 70 across all time buckets shown.
  - Monies due to Credit institutions: 100 across all time buckets shown.
  - Monies due to Other financial customers: 100 across all time buckets shown.
  - Paper in own portfolio maturing: 100 across all time buckets shown.
  - Other inflows: 100 across all time buckets shown.

*Source: Appendix V–VII, “Contributions to Changes in Capital—Adverse Scenarios”; “MCIs: Satellite Models and Methodologies for Stress Testing”; “Group-Level Consolidation”; maturity ladder inflows (selected entries).*

### Appendix VIII.

### Appendix VIII. Banks: Liquidity Stress Tests

### Run-off Rates for Liquidity Outflows (In percent)
- Time buckets (columns implied): Overnight; Greater than overnight up to 2 days; Greater than 2 days up to 3 days; Greater than 3 days up to 4 days; Greater than 4 days up to 5 days; Greater than 5 days up to 6 days; Greater than 6 days up to 7 days; Greater than 7 days up to 2 weeks; Greater than 2 weeks up to 3 weeks; Greater than 3 weeks up to 30 days; Greater than 30 days up to 5 weeks; Greater than 5 weeks up to 2 months; Greater than 2 months up to 3 months; Greater than 3 months up to 4 months; Greater than 4 months up to 5 months; Greater than 5 months up to 6 months; Greater than 6 months up to 9 months; Greater than 9 months up to 12 months.
- Liabilities resulting from securities issued (if not treated as retail deposits)
  - Unsecured bonds due (0010): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - Regulated covered bonds (0030): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - Securitisations due (0040): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - Other (0050): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
- Liabilities resulting from secured lending and capital market driven transactions collateralised by:
  - Level 1 tradable assets (0070): - - - - - - - - - - - - - - - - - -
  - Level 1 excluding covered bonds (0080): - - - - - - - - - - - - - - - - - -
  - Level 1 central bank (0090): - - - - - - - - - - - - - - - - - -
  - Level 1 (CQS 1) (0100): - - - - - - - - - - - - - - - - - -
  - Level 1 (CQS2, CQS3) (0110): - - - - - - - - - - - - - - - - - -
  - Level 1 (CQS4+) (0120): - - - - - - - - - - - - - - - - - -
  - Level 1 covered bonds (CQS1) (0130): 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
  - Level 2A tradable assets (0140): (no explicit row values)
  - Level 2A corporate bonds (CQS1) (0150): 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30
  - Level 2A covered bonds (CQS1, CQS2) (0160): 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30
  - Level 2A public sector (CQS1, CQS2) (0170): 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30
  - Level 2B tradable assets (0180): (no explicit row values)
  - Level 2B ABS (CQS1) (0190): 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
  - Level 2B covered bonds (CQS1-6) (0200): 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
  - Level 2B: corporate bonds (CQ1-3) (0210): 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
  - Level 2B shares (0220): 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
  - Level 2B public sector (CQS 3-5) (0230): 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
  - Other tradable assets (0240): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - Other assets (0250): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
- Liabilities not reported in 1.2, resulting from deposits received (excluding deposits received as collateral)
  - Stable retail deposits (sight) (0260): 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.3 1.0 1.0 0.9 2.2 3.3 2.7 2.7 2.7 - -
  - Stable retail deposits (term) (0270): -7.0 7.0 7.0 7.0 7.0 7.0 10.0 10.0 10.0 10.0 10.0 10.0 7.0 7.0 7.0 5.0 5.0
  - Other retail deposits (sight): 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.9 2.5 2.5 2.5 6.0 7.0 6.5 6.5 5.5 - -
  - Other retail deposits (term) (0280): -20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 15.0 15.0 15.0 10.0 10.0 10.0 7.0 7.0
  - Operational deposits (sight) (0290): 0.7 0.7 0.7 0.7 0.7 0.7 0.7 1.5 4.0 4.0 4.0 8.0 10.0 9.0 7.0 6.0 - -
  - Non-operational deposits from credit institutions (0300): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - Non-operational deposits from other financial customers (0310): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - Non-operational deposits from central banks (0320): - - - - - - - - - - - - - - - - - -
  - Non-operational deposits from non-financial corporates (sight): 1.1 1.1 1.1 1.1 1.1 1.1 1.0 1.8 5.5 5.5 5.5 10.0 13.0 10.0 8.5 6.5 - -
  - Non-operational deposits from non-financial corporates (term) (0330): -50 50 50 50 50 50 50 50 50 40 40 40 30 30 30 20 20
  - Non-operational deposits from other counterparties (0340): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
  - FX-swaps maturing (0350): (no explicit row values)
  - Derivatives amount payables other than those reported in 1.4 (0360): (no explicit row values)
  - Other outflows (0370): 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
- Outflows from committed facilities (1090)
  - Committed credit facilities (1100)
    - Considered as Level 2B by the receiver (1110): 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.5 1.5 1.5 2.5 2.5 5.0 - - - -
    - Other (1120): 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.5 1.5 1.5 2.5 2.5 5.0 - - - -
  - Liquidity facilities (1130): 2.1 2.1 2.1 2.1 2.0 2.0 2.0 3.6 4.6 14.0 6.0 18.0 14.0 - - - -
  - Outflows due to downgrade triggers (1140): (no explicit row values)

### Counterbalancing Capacity — Haircuts (In percent)
- Coins and bank notes (0730): -
- Withdrawable central bank reserves (0740): -
- Level 1 tradable assets (0750): (no explicit row values)
- Level 1 excluding covered bonds (0760): (no explicit row values)
- Level 1 central bank (0770): 7
- Level 1 (CQS 1) (0780): 7
- Level 1 (CQS2, CQS3) (0790): 7
- Level 1 (CQS4+) (0800): 7
- Level 1 covered bonds (CQS1) (0810): 7
- Level 2A tradable assets (0820): (no explicit row values)
- Level 2A corporate bonds (CQS1) (0830): 20
- Level 2A covered bonds (CQS 1, CQS2) (0840): 12
- Level 2A public sector (CQS1, CQS2) (0850): 12
- Level 2B tradable assets (0860): (no explicit row values)
- Level 2B ABS (CQS1) (0870): 40
- Level 2B covered bonds (CQS1-6) (0880): 25
- Level 2B corporate bonds (CQ1-3) (0890): 40
- Level 2B shares (0900): 70
- Level 2B public sector (CQS 3-5) (0910): 30
- Other tradable assets (0920): (no explicit row values)
- Central government (CQS1) (0930): 50
- Central government (CQS 2 & 3) (0940): 60
- Shares (0950): 80
- Covered bonds (0960): 35
- ABS (0970): 60
- Other tradable assets (0980): 80
- Non tradable assets eligible for central banks (0990): 80

### Notes and metadata included in the appendix
- The appendix includes template references: Template: C 66.01.a - Maturity ladder (0005 OUTFLOWS; 1089 CONTINGENCIES).
- Country context label: DENMARK.
- Sources cited within tables: Authorities; and IMF staff.

### Key points on reported numeric patterns (as presented)
- Many marketable liability categories (unsecured bonds, covered bonds, securitisations, other market liabilities) show uniform run-off rates of 100 across all time buckets.
- Several Level 1 and Level 2 marketable asset classes have reduced run-off assumptions (for example, Level 1 covered bonds (CQS1) at 10 across all buckets; Level 2A corporate/public/covered bonds at 30; Level 2B categories mostly at 50).
- Retail and non-financial corporate deposit categories show heterogeneous short-term run-off assumptions including small fractional rates (e.g., 0.2, 0.3, 0.4, 0.7, 1.1) and large discrete jumps or negative/large values for term deposit conversion/run-off assumptions (e.g., -7.0, -20.0, -50).
- Haircuts on collateral for counterbalancing capacity range from 7 percent for many Level 1 items up to 80 percent for certain shares/other tradable or non-tradable assets.

*Source: Appendix VIII, 1dnkea2020009 - Appendix VIII.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1dnkea2020009.pdf_