## 1dnkea2020004

## Source details

**Canonical URL:** [1dnkea2020004](https://www.imf.org/-/media/files/publications/cr/2020/english/1dnkea2020004.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2020/english/1dnkea2020004.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2020/english/1dnkea2020004.pdf.json)

---

### EXECUTIVE SUMMARY — INTRODUCTION
- FSAP work conducted prior to the COVID-19 pandemic; this Technical Note (TN) does not assess the impact of the crisis or recent crisis-related policy measures.
- Institutional arrangement:
  - Macroprudential authority: Systemic Risk Council (SRC) is advisory; Minister for Industry, Business and Financial Affairs (MIBFA) holds ultimate decision-making power over most macroprudential tools (Paragraph 1).
  - Atypical arrangement has rendered the system vulnerable to inaction bias in some cases (Paragraph 1).
- Recent policy and buffers:
  - Banks hold large capital buffers.
  - Several demand-side measures have limited risky lending to highly indebted households (Paragraph 1).
  - Countercyclical capital buffer (CCyB) has been fully released in response to the Covid-19 crisis.

### INSTITUTIONAL FRAMEWORK, INACTION BIAS, AND RECOMMENDATIONS
- SRC design and operation:
  - SRC established in 2013; consists of 10 members: one representative from MIBFA, two from MoF, two from DFSA, two from DN and three independent experts.
  - At least four meetings a year; decisions taken by majority voting; SRC strives for consensus and the secretariat drafts recommendations only after receiving a mandate from the SRC (Paragraphs 11−13).
  - SRC issues observations (soft), warnings, and recommendations with a “comply-or-explain” mechanism (semi-hard).
  - MIBFA is the designated authority with ultimate decision-making power; DN chairs SRC and heads SRC secretariat but lacks decision-making powers on macroprudential instruments.
- Observed inaction bias and examples:
  - Consensus-building procedural requirement has led to delayed action (e.g., SRC noted risk in September 2014 but recommended CCyB activation to 0.5 percent only in December 2017).
  - MIBFA did not comply with SRC recommendation on deferred amortization loans in 2014 and only partially complied in 2017 despite SRC and ESRB warnings.
- Institutional recommendations (selected):
  - Recommendation 1: Give the chair of the SRC ability, enshrined in law, to make proposals for a recommendation after due consultation with other SRC members without the need to strive for consensus (MIBFA; MT) (Paragraphs 11−13; 21).
  - Include a comprehensive overview of SRC recommendations in the MIBFA’s annual report to Parliament to increase accountability (MIBFA; ST) (Paragraph 22).
  - Over time consider giving SRC hard powers over capital tools under CRD/CRR legislation; borrower-based tools could remain with MIBFA while SRC retains recommendation powers with comply-or-explain.

### KEY FINANCIAL VULNERABILITIES
- Macro-financial context and summary stats:
  - Policy rates entered negative territory in mid-2012; krone pegged to euro implying medium-term loose monetary conditions.
  - Credit growth: 4.8 percent in 2019 Q3 versus 3.5 percent in 2018.
  - Aggregate credit-to-GDP ratio: 160 percent of GDP.
  - CCyB increased to 2 percent (effective in December 2020) with forward guidance to 2.5 percent if risks continue to build up.
- Elevated vulnerabilities:
  - High household leverage amid high real estate valuations following a long period of loose financial conditions.
  - Signs of increased risk taking in some sectors, most notably commercial real estate (CRE).
  - Downside risks to bank profitability in a low-interest-rate environment.
  - Mortgage credit institutions (MCIs) central to domestic interbank system; significant contagion via covered bond exposures.
- Household sector specifics:
  - Household debt among highest in advanced economies.
  - Mortgage loans account for about 90 percent of total household debt from banks and MCIs.
  - Household loans share in total loans from MCIs and banks is around 65 percent.
  - 60 percent of mortgage stock on variable terms, of which 36 percent are non-amortizing.
  - A 20 percent reduction in house prices would increase share of residential real estate loans with current LTV > 100 from 2 to 12 percent of MCI mortgage portfolio (excluding bank loans).
  - Non-linear increase in household NPLs with LTVs with break after LTV of 85 percent and a break at LTI of 500.
- Corporate and CRE specifics:
  - Corporate debt-to-GDP at around 90 percent.
  - CRE accounts for about 30 percent of total corporate debt and is the most leveraged corporate sector.
  - Firm-level analysis indicates a six-fold jump in debt-at-risk for the CRE sector in an adverse scenario (debt-at-risk aggregates debt of firms with projected PDs above 1.5 percent).
- Systemic institutions and buffers:
  - Seven institutions assessed as systemically important and subject to systemic risk buffer (SRB).
  - O-SII allocation and buffer rates:
    - Subcategory 1: Score <=5 — Buffer Rate 1 percent — Sydbank; DLR Kredit; Spar Nord Bank A/S
    - Subcategory 2: Score [5-15[ — Buffer Rate 1.5 percent — Jyske Bank; Nordea Kredit Realkreditaktieselskab
    - Subcategory 3: Score [15-25[ — Buffer Rate 2 percent — Nykredit Realkredit
    - Subcategory 4: Score [25-35[ — Buffer Rate 2.5 percent — (no institution listed)
    - Subcategory 5: Score >=35 — Buffer Rate 3 percent — Danske Bank
  - Note: Spar Nord Bank A/S SRB set to 0.5 per cent at end-2019 and 1 per cent at end-2020 per Danish Financial Business Act.
  - Under CRD V, SRB can no longer be used as a substitute for O-SII buffer; Denmark will have to shift SRB to the O-SII buffer.

### MACROPRUDENTIAL TOOLKIT — HOUSEHOLDS
- Existing borrower-based and housing measures:
  - Supervisory diamond for MCIs: 5 benchmarks (announced 2014, amended 2016).
  - Mandatory downpayment: 5 percent downpayment required (effective 95 percent LTV); single-loan restrictions: 80 percent LTV per loan for MCI-financed loans; remaining 15 percent via additional loan with secondary lien.
  - Guidelines (“Seven Best Practices”) for growth areas: new borrowers with DTI above 4 or 5 require sufficient wealth to maintain positive net wealth under a 10% or 25% house price drop, respectively.
  - Good Business Practice for Mortgage Lending (2018): new borrowers with DTI above 4 and LTV above 60% must have interest rate fixation of at least five years; deferred amortization only if interest rate fixation is 30 years.
- Observed effects of measures:
  - Reduced share of interest-only and variable rate loans in new lending; shift from variable to fixed rates.
  - Large share of highly indebted households (LTI > 4) still receive new loans; significant share continue to opt for deferred amortization loans.
- Recommendations (timing legend: ST: Short term (1-3 years); MT: Medium Term (3-5 years)):
  - Recommendation 6 (MIBFA; ST): Introduce national legislation to include borrower-based tools (limits on LTVs, DTIs, and DSTIs) in the policy toolkit (Paragraphs 42, 44).
  - Recommendation 7 (MIBFA; MT): Introduce a stricter LTV limit to safeguard against large house price shocks (Paragraphs 40, 43, 45).
  - Recommendation 8 (MIBFA; MT): Introduce a binding income-based macroprudential measure that limits lending to households above a certain DTI threshold calibrated using credit registry data (Paragraphs 40, 43, 45).
  - Recommendation 9 (SRC; MT): Issue recommendations to responsible authorities to reduce debt bias, simplify rental regulations, and relax supply constraints on housing (Paragraphs 38, 46).
- Specific policy options if vulnerabilities persist:
  - Introduce stricter LTV limit (micro data suggests NPL jump after LTV 85 percent); earlier IMF work suggests lowering LTV from 95 to 90 percent.
  - Introduce strict DTI limits for all borrowers irrespective of LTV if house prices remain elevated; calibrate using credit registry; consider speed limits to minimize real economy impact.
  - Mandatory amortization for households with debt-to-income above 400 percent.
  - Structural complements: reduce mortgage interest deductibility, rebalance tax incentives on savings products, restore link between taxes and market prices (postponed to 2024), reduce rent control, relax supply constraints.

### MACROPRUDENTIAL TOOLKIT — CRE AND CORPORATES
- Observations:
  - CRE is the most leveraged sector and shows building vulnerabilities; vacancy rates increased in some segments; foreign investor share significant in Copenhagen.
  - Firm-level stress tests show CRE has far greater debt-at-risk increases in adverse scenarios.
- Existing measures:
  - Supervisory Diamond for banks: commercial property lending ≤ 25 percent of total bank lending; lending growth cap of 20 percent.
  - Supervisory Diamond for MCIs: caps lending growth at 15 percent in different segments.
  - 100 percent limit on debt-service-to-income ratio for bank lending.
- Recommended actions if risks intensify:
  - Recommendation 10 (MIBFA; MT): Continue monitoring CRE vulnerabilities and take actions (higher CRE risk weights, sectoral systemic risk buffer) if risks intensify (Paragraphs 49-50, 52).
  - Consider temporary risk weight floors or temporary risk weight add-ons on CRE exposures for IRB banks and MCIs using Art. 458 CRR.
  - Consider sectoral SRB/CCyB; sectoral SRB preferable when available under CRDV/CRRII; capital buffers adjustable over the cycle.

### SYSTEMIC RISK MONITORING, ANALYTICAL CAPACITY, AND DATA GAPS
- Current strengths:
  - Strong analytical capacity; DN produces semi-annual financial stability report; SRC secretariat produces quarterly SRC reports.
  - Central credit registry (CCR) established; allows PD modeling with variables including DTI and financial margins.
  - Use of SRISK, GDP-at-risk approach, multiple models and micro datasets.
- Identified gaps and recommended enhancements:
  - Recommendation 3 (DN; MT): Develop macroprudential stress tests that take feedback loops between financial system and real economy more fully into account while incorporating contagion effects across financial institutions (Paragraph 30).
  - Recommendation 4 (DN; MT): Extend coverage of systemic risk monitoring to the non-bank financial sector (insurers, pension funds, asset management companies) (Paragraph 29).
  - Recommendation 5 (DN; MT): Close data gaps, notably by enhancing coverage and quality of commercial real estate data (Paragraph 31).
  - CRE data needs: centralized market data, indicators on CRE credit exposures (stocks and flows, NPLs, loan-loss provisions), exposure risk characteristics by region/segment, CRE lending standards, regular reporting of CRE companies’ financial condition.
  - DN should consider expanding human resources fully dedicated to macroprudential policy in the SRC secretariat.

### FUNDING, LIQUIDITY, AND INTERCONNECTEDNESS VULNERABILITIES
- Funding and liquidity:
  - Market-based funding focused on domestic issuance of covered bonds by MCIs; covered bonds largely issued domestically and denominated in DKK; FX issuance currently 4.2 percent.
  - Banks and MCIs must maintain minimum LCR of 100 percent for all currencies in total and for each significant currency; banks comply with LCR with a comfortable margin.
  - Net stable funding ratio (NFSR) monitored and expected to be introduced as a requirement in 2021.
  - HQLA calculation: at least 30 percent must be government-guaranteed assets or central bank deposits; mortgage bonds must not exceed 70 percent.
  - LCR in significant currencies is enforced in Denmark.
- Interconnectedness and contagion:
  - Housing and covered bond exposures interlink MCIs, banks, pension funds, insurance companies, foreign investors, and households.
  - MCIs induce high contagion losses through unlisted shares and covered bond exposures; systemic banks vulnerable to inward spillovers via covered bonds.
  - Ensuring health of covered bond market under stress is crucial for financial stability.

### GROWTH AT RISK (GaR) — METHODOLOGY, FINDINGS, AND POLICY IMPLICATIONS
- Methodology:
  - GaR framework links current financial conditions to distribution of future growth outcomes; variables grouped as (i) financial conditions/price of risk, (ii) leverage, (iii) external factors.
  - Horizons: short- and long-term refer to 4 and 12 quarters ahead, respectively.
- Key findings:
  - Current loose financial conditions stimulate near-term growth but increase medium-term downside risk.
  - Tail-risk metric: 5th percentile of conditional GDP distribution.
  - In the severely adverse scenario, one-year ahead GDP growth in Denmark will be about 1 percent or less.
  - In the severely adverse scenario, GDP growth could turn negative in a three year ahead period.
  - Higher household leverage boosts one-year ahead growth but increases downside risks at a three-year horizon.
- Policy implications:
  - Caution warranted during continued loose financial conditions due to stretched asset valuations and increased risk taking.
  - Macroprudential buffers recommended to build resilience.
  - SRC already monitors GaR and had increased CCyB to build resilience; release of the buffer with Covid-19 should add resilience to banking system and real economy.

### SUMMARY OF KEY RECOMMENDATIONS (from Table 1)
- Institutional Arrangements:
  - 1. Give the chair of the SRC ability, enshrined in law, to make proposals for a recommendation after due consultation with other SRC members without the need to strive for consensus (MIBFA; MT) (Paragraphs 11−13; 21).
  - 2. Include a comprehensive overview of recommendations made by the SRC in the MIBFA’s annual report to Parliament (MIBFA; ST) (Paragraph 22).
- Systemic Risk Monitoring:
  - 3. Develop macroprudential stress tests that take feedback loops between financial system and real economy more fully into account while incorporating contagion effects across financial institutions (DN; MT) (Paragraph 30).
  - 4. Extend coverage of systemic risk monitoring to the non-bank financial sector (DN; MT) (Paragraph 29).
  - 5. Close data gaps, including by enhancing the coverage and quality of commercial real estate data (DN; MT) (Paragraph 31).
- Toolkit:
  - 6. Introduce national legislation to include borrower-based tools (limits on LTVs, DTIs, and DSTIs) in the policy toolkit (MIBFA; ST) (Paragraphs 42, 44).
  - 7. Introduce a stricter LTV limit to safeguard against large house price shocks (MIBFA; MT) (Paragraphs 40, 43, 45).
  - 8. Introduce a binding income-based macroprudential measure that limits lending to households above a certain DTI threshold calibrated using newly available credit registry data (MIBFA; MT) (Paragraphs 40, 43, 45).
  - 9. Issue recommendations to responsible authorities to reduce debt bias, simplify rental regulations, and relax supply constraints on housing (SRC; MT) (Paragraphs 38, 46).
  - 10. Continue monitoring CRE vulnerabilities and take actions (higher CRE risk weights, sectoral systemic risk buffer) if risks intensify (MIBFA; MT) (Paragraphs 49-50, 52).

*Source: EXECUTIVE SUMMARY and Chapter content, Technical Note prepared by Umang Rawat (IMF).*

### EXECUTIVE SUMMARY __________________________________________________________________________ 6

### EXECUTIVE SUMMARY

### INTRODUCTION
- FSAP work was conducted prior to the COVID-19 pandemic; this Technical Note (TN) does not assess the impact of the crisis or recent crisis-related policy measures. Findings and recommendations of the TN remain pertinent given its focus on vulnerabilities and policy frameworks.
- Denmark’s macroprudential authority is the Systemic Risk Council (SRC), an advisory body; the Minister for Industry, Business and Financial Affairs (MIBFA) holds ultimate decision-making power over most macroprudential tools (Paragraph 1).
- Authorities have undertaken several macroprudential measures since the last FSAP; banks hold large capital buffers and several demand-side measures have limited risky lending to highly indebted households (Paragraph 1).
- The atypical institutional arrangement has rendered the system vulnerable to inaction bias in some cases (Paragraph 1).

### INSTITUTIONAL FRAMEWORK AND INACTION BIAS
- Institutional features:
  - SRC is advisory; MIBFA is decision-maker for most tools (Paragraph 1).
  - SRC strives for consensus in decision-making; the secretariat starts drafting recommendations only after receiving a mandate from the SRC (Paragraphs 11−13 as referenced in Table 1).
- Consequences:
  - The consensus-building procedural requirement has led to delayed action in some cases.
  - Limited powers of the SRC can lead to inaction bias when MIBFA does not comply with SRC recommendations (noted particularly for deferred amortization loans in the household sector).
- Recommended institutional change:
  - "Give the chair of the SRC ability, enshrined in law, to make proposals for a recommendation after due consultation with other SRC members without the need to strive for consensus" (Recommendation 1; MIBFA; MT) (Paragraphs 11−13; 21).

### KEY FINANCIAL VULNERABILITIES
- Elevated vulnerabilities (Paragraphs 2 and summary):
  - High household leverage amid high real estate valuations following a long period of loose financial conditions.
  - Signs of increased risk taking in some sectors, most notably commercial real estate (CRE), which is among the most leveraged corporate sectors and where debt has grown fastest in recent years.
  - Downside risks to bank profitability in a low-interest-rate environment.
  - Mortgage credit institutions (MCIs) play a central role in the domestic interbank system and can generate significant contagion effects across a financial system that is highly interconnected by covered bond exposures.

### MACROPRUDENTIAL TOOLKIT: HOUSEHOLDS
- Recent developments:
  - Recent measures have aided in softening house prices and switching households to loans with higher amortization and lower interest rate risks.
  - Countercyclical capital buffer (CCyB) has been fully released in response to the Covid-19 crisis.
- Policy guidance (short- and medium-term):
  - Short-term (ST) and medium-term (MT) recommendations (from Table 1):
    - Introduce national legislation to include borrower-based tools (limits on loan-to-value (LTV), debt-to-income (DTI), and debt service to income (DSTI)) in the policy toolkit (Recommendation 6; MIBFA; ST) (Paragraphs 42, 44).
    - Introduce a stricter LTV limit to safeguard against large house price shocks (Recommendation 7; MIBFA; MT) (Paragraphs 40, 43, 45).
    - Introduce a binding income-based macroprudential measure that limits lending to households above a certain DTI threshold calibrated using newly available credit registry data (Recommendation 8; MIBFA; MT) (Paragraphs 40, 43, 45).
    - Issue recommendations to responsible authorities to reduce debt bias, simplify rental regulations, and relax supply constraints on housing (Recommendation 9; SRC; MT) (Paragraphs 38, 46).
  - Rationale:
    - Where LTV rules become less binding in an elevated house price economy, binding income-based limits (e.g., DTI) should be used irrespective of LTV ratios.
  - Current juncture:
    - Given COVID-19 uncertainty and economic downturn, further tightening is undesirable now, but authorities should stand ready to act if household vulnerabilities persist.

### MACROPRUDENTIAL TOOLKIT: COMMERCIAL REAL ESTATE (CRE) AND CORPORATES
- Observations:
  - CRE shows building vulnerabilities; firm-level analysis indicates CRE sector is most vulnerable to increased debt-at-risk in an adverse scenario.
- Policy guidance:
  - Do not tighten macroprudential tools at the current juncture.
  - Over time consider measures addressing CRE vulnerabilities, including higher risk weights and sectoral systemic risk buffer (Recommendation 10; MIBFA; MT) (Paragraphs 49-50, 52).

### SYSTEMIC RISK MONITORING AND ANALYTICAL CAPACITY
- Current strengths:
  - Strong analytical capacity and generally good data quality and availability; recent establishment of a credit registry is progress (Paragraphs 29−31).
- Recommended enhancements:
  - Develop macroprudential stress tests that take feedback loops between financial system and real economy more fully into account while incorporating contagion effects across financial institutions (Recommendation 3; DN; MT) (Paragraph 30).
  - Extend coverage of systemic risk monitoring to the non-bank financial sector (insurers, pension funds, asset management companies) (Recommendation 4; DN; MT) (Paragraph 29).
  - Close data gaps, notably by enhancing coverage and quality of commercial real estate data (Recommendation 5; DN; MT) (Paragraph 31).

### SUMMARY OF KEY RECOMMENDATIONS (from Table 1)
- Institutional Arrangements:
  - 1. Give the chair of the SRC ability, enshrined in law, to make proposals for a recommendation after due consultation with other SRC members without the need to strive for consensus (MIBFA; MT) (Paragraphs 11−13; 21).
  - 2. A comprehensive overview of recommendations made by the SRC should be included in the MIBFA’s annual report to Parliament to further increase accountability (MIBFA; ST) (Paragraph 22).
- Systemic Risk Monitoring:
  - 3. Develop macroprudential stress tests that take feedback loops between financial system and real economy more fully into account while incorporating contagion effects across financial institutions (DN; MT) (Paragraph 30).
  - 4. Extend coverage of systemic risk monitoring to the non-bank financial sector (DN; MT) (Paragraph 29).
  - 5. Close data gaps, including by enhancing the coverage and quality of commercial real estate data (DN; MT) (Paragraph 31).
- Toolkit:
  - 6. Introduce national legislation to include borrower-based tools (limits on LTVs, DTIs, and DSTIs) in the policy toolkit (MIBFA; ST) (Paragraphs 42, 44).
  - 7. Introduce a stricter LTV limits to safeguard against large house price shocks (MIBFA; MT) (Paragraphs 40, 43, 45).
  - 8. Introduce a binding income-based macroprudential measure that limits lending to households above a certain DTI threshold calibrated using newly available credit registry data (MIBFA; MT) (Paragraphs 40, 43, 45).
  - 9. Issue recommendations to responsible authorities to reduce debt bias, simplify rental regulations, and relax supply constraints on housing (SRC; MT) (Paragraphs 38, 46).
  - 10. Continue monitoring CRE vulnerabilities and take actions (higher CRE risk weights, sectoral systemic risk buffer) if risks intensify (MIBFA; MT) (Paragraphs 49-50, 52).
- Timing legend: ST: Short term (1-3 years); MT: Medium Term (3-5 years).

*Source: EXECUTIVE SUMMARY, Technical Note prepared by Umang Rawat (IMF).*

### Section V concludes. The assessment is conducted based on the IMF guidance, which is laid out in

### 1dnkea2020004 - Section V concludes. The assessment is conducted based on the IMF guidance, which is laid out in

### Institutional framework: design, roles, and operational constraints
- Macroprudential authority: Systemic Risk Council (SRC), established by MIBFA in 2013.
  - Consist of 10 members: one representative from the MIBFA, two from the MoF, two from the DFSA, two from DN and three independent experts.
  - At least four meetings a year.
  - SRC is an advisory body; MIBFA is the designated authority with ultimate decision-making power.
- Roles of institutions:
  - Danmarks Nationalbank (DN): central bank with explicit mandate for financial stability; chair of SRC and head of SRC secretariat; strong analytical role but no decision-making powers on macroprudential instruments.
  - Danish Financial Supervisory Authority (DFSA): integrated financial supervisor and competent authority for CRD IV/CRR instruments; conducts stress tests, quantitative impact studies, and SIFI designation.
  - Ministry of Industry, Business and Financial Affairs (MIBFA): designated authority for CRD IV/CRR instruments; minister typically decides on other macroprudential instruments implemented by executive order.
  - Ministry of Finance (MoF) and independent experts (including at least one with insurance/pension knowledge) participate in SRC.
- Decision-making features and constraints:
  - Decisions in SRC taken by majority voting; an “abstention rule” prevents government members and DFSA members (when advising government) from voting on recommendations to government.
  - SRC strives for consensus, but consensus-building is internalized prior to formal voting; secretariat generally drafts proposals only after SRC mandate.
  - SRC powers: observations (soft), warnings, and recommendations with a “comply-or-explain” mechanism (semi-hard). SRC does not have hard powers over macroprudential tools; MIBFA controls most tools.
  - Information powers: SRC has access to information from DFSA, relevant ministries, and DN; the CCR data enhances information powers.
- Operational capacity and coordination:
  - SRC secretariat headed by DN; collaborative inputs from DN departments and DFSA provide sufficient quality/quantity of resources, though DN has few staff fully dedicated to macroprudential policy.
  - SRC facilitates cooperation, exchange of information, and participates in ESRB and Nordic Baltic Macroprudential Forum (NBMF).
  - Existing MoUs to strengthen cross-border cooperation (including a MoU among Nordic MoFs and with the ECB).

### Willingness to act, ability to act, and observed inaction bias
- Principle I — Willingness to Act:
  - SRC has explicit mandate and six intermediate objectives: mitigate/prevent excessive credit growth and leverage; mitigate/prevent excessive maturity mismatch and market illiquidity; limit direct exposure concentrations; limit systemic risks related to indirect exposure concentrations (interconnectedness); limit systemic risks connected with systemically important financial institutions and reduce misaligned incentives; strengthen resilience of financial structures.
  - Despite DN’s central analytical role, DN lacks decision-making powers.
  - Example of delay: SRC noted “the risk of systemic financial risks building up” from September 2014, but recommended activation of the CCyB to 0.5 percent only in December 2017.
- Principle II — Ability to Act:
  - SRC issues 12 statements to date: two observations, one warning, nine recommendations (see table of interventions in source).
  - SRC’s recommendations can be complied with or met with a public report within three months explaining non-implementation; SRC can publish an assessment of consequences for systemic risks if recommendations not followed.
  - Inaction bias observed on housing measures: MIBFA did not comply with SRC recommendation on deferred amortization loans in 2014 and only partially complied in 2017 despite SRC warnings and ESRB warning in September 2016.
- Principle III — Effective Cooperation and Coordination:
  - DN leads macroprudential analysis and prepares quarterly Joint Macroprudential Report for SRC meetings.
  - Ministries propose legislative amendments and act as bridge between SRC and government.
  - Secretarial arrangements involve staff from DN, DFSA and ministries; DN and DFSA jointly conduct risk assessments prior to SRC meetings.
  - Regional engagement: DN and DFSA participate in ESRB and NBMF; cross-border MoUs exist to arrange mutual recognition of macroprudential measures and host-country regulation coordination.

### Systemic risk monitoring: tools, data, and analytical framework
- Monitoring approach and outputs:
  - DN produces a semi-annual financial stability report; SRC secretariat produces quarterly SRC reports with three parts: risk picture, CCyB and housing market notes, and ad-hoc thematic notes.
  - Guided discretion approach: SRC uses a broad set of indicators for each intermediate objective and relies on judgement informed by models and micro data.
- CCyB methodology and indicators:
  - SRC’s CCyB advice is based on six key indicator categories: Risk perception; Property prices; Credit standards; Credit developments; Risk build-up in credit institutions; Model-based indicators.
  - Example indicators: financial stress indicator, credit spreads, equity volatility; house prices, flat prices, commercial property prices, house price-to-income gap; lending survey, interest spread, housing burden; total credit-to-GDP, credit to households and corporates, credit-to-GDP gap; leverage, earnings, excess capital adequacy; financial cycle estimates.
- Data resources and analytical models:
  - Use of SRISK model to measure systemic risk for banks and contagion via covered bonds; GDP-at-risk approach developed; multiple models, micro datasets, and empirical approaches employed.
  - Central credit registry (CCR) availability: detailed lender and borrower data, linkable to Statistics Denmark registries; allows PD modeling using explanatory variables including DTI and financial margins.
- Identified gaps and required enhancements:
  - Monitoring is bank/MCI-centric; non-bank financial sector coverage (insurers, pension funds, asset managers) should be extended.
  - Macroprudential stress testing needs development to capture spillovers to real economy and second-round effects and to assess impact of macroprudential instruments.
  - CRE data coverage and quality improvements needed: centralized market data, indicators on CRE credit exposures (stocks and flows, NPLs, loan-loss provisions), exposure risk characteristics by region/segment, CRE lending standards, and regular reporting of CRE companies’ financial conditions.

### Systemic vulnerabilities, macro-financial context, and tool recommendations
- Macro-financial context and broad vulnerabilities:
  - Policy rates entered negative territory in mid-2012; sovereign yields negative up to 10 years; krone pegged to euro implying medium-term loose monetary conditions.
  - Credit growth: 4.8 percent in 2019 Q3 versus 3.5 percent in 2018.
  - Aggregate credit-to-GDP ratio: 160 percent of GDP.
  - CCyB developments: CCyB increased to 2 percent (effective in December 2020) with forward guidance to 2.5 percent if risks continue to build up.
  - IMF stress test results: in a severe adverse scenario, three SIFIs would have CET1 below overall requirements when hurdle rates assumed to relax CCyB to zero but not allow use of CCB and SRB; if CCB use allowed, all seven SIFIs meet regulatory requirements, some by partially depleting CCB.
  - CCyB release during Covid-19: CCyB released and set to zero to limit economic impact; authorities temporarily relaxed systemic risk buffer in Faroe Islands from 3 to 2 percent.
  - Leverage ratio considerations: minimum leverage ratio requirement, once binding, may overrule risk-based capital requirements and make part of current buffers unusable; SIFIs with low risk weight density may face higher Tier 1 requirement from leverage rule; consideration could be given to introduce a leverage ratio buffer once leverage requirements become binding.
- Household vulnerabilities:
  - Household debt among highest in advanced economies; mortgage loans account for about 90 percent of total household debt from banks and MCIs; household loans share in total loans from MCIs and banks is around 65 percent.
  - Household mortgage characteristics: 60 percent of mortgage stock on variable terms, of which 36 percent are non-amortizing.
  - Structural contributors to high household debt: mandatory savings schemes, rental market regulation limiting supply in major cities, favorable tax treatment of owner-occupied housing, high mortgage interest deductibility, and housing property taxation decoupled from market prices (new system passed in 2017 was due to take effect in 2021 but postponed to 2024).
  - House prices: elevated nationally and especially in Copenhagen; house prices in urban areas rising faster than fundamentals; house-price-at-risk analysis indicates greater downside risks in Copenhagen relative to national market.
  - Vulnerable household groups:
    - Buyers in highly appreciating urban areas (e.g., Copenhagen) with higher LTI and credit growth.
    - High DTI households, especially low-income households with high debt servicing burdens.
    - Households with variable rate and/or deferred amortization mortgages.
- Housing finance measures to date:
  - Supervisory diamond for MCIs: regulatory framework with 5 benchmarks (announced 2014, amended 2016).
  - Mandatory downpayment: 5 percent downpayment required for residential real estate purchases (effective 95 percent LTV); single-loan restrictions: 80 percent LTV per loan for MCI-financed loans; remaining 15 percent financed by an additional loan with secondary lien.
  - Guidelines on Good Mortgage Lending in Growth Areas (“Seven Best Practices”): for new borrowers in growth areas with DTI above 4 or 5 require sufficient wealth to maintain positive net wealth under a 10% or 25% house price drop, respectively.
  - Good Business Practice for Mortgage Lending (2018): new borrowers with DTI above 4 and LTV above 60% must have interest rate fixation of at least five years and deferred amortization only if interest rate fixation is 30 years.
- Observed outcomes:
  - Recent measures reduced share of interest-only and variable rate loans in new lending, and shifted borrowers from variable to fixed rates.
  - Large share of highly indebted households (LTI > 4) still receive new loans; significant share continue to opt for deferred amortization loans.
  - Sensitivity analysis: a 20 percent reduction in house prices would increase share of residential real estate loans with current LTV > 100 from 2 to 12 percent of MCI mortgage portfolio (excluding bank loans).
  - Non-linear increase in household NPLs with LTVs with break after LTV of 85 percent and a break at LTI of 500.

### Key policy recommendations (from the assessment)
- Institutional recommendations:
  - SRC chair should be given the ability, enshrined in law, to make proposals for a recommendation—after due consultation with other SRC members—without the need to strive for consensus, to limit consensus-building delays before secretariat starts drafting proposals.
  - Over time, consider giving SRC hard powers over capital tools under CRD/CRR legislation (i.e., make SRC the designated authority for macroprudential policy). Borrower-based tools, given distributional implications, can remain with MIBFA while SRC retains recommendation powers with comply-or-explain.
  - Include a comprehensive overview of SRC recommendations in MIBFA’s annual report to Parliament (Folketing) to increase transparency and accountability; gather all SRC recommendations in a table with explanations of how MIBFA followed or deviated from advice.
- Systemic risk monitoring and data:
  - Further extend systemic risk monitoring coverage to non-bank financial sector (insurers, pension funds, asset management companies) in both regular surveillance and in-depth analysis.
  - Develop macroprudential stress test frameworks that capture feedback loops between financial system and real economy and allow analysis of macroprudential instrument effects; focus on how tools can prevent macroeconomic deterioration rather than banks merely “passing” tests.
  - Enhance CRE sector data: centralized market data collection, indicators on CRE credit exposures (stocks and flows, NPLs, loan-loss provisions), exposure risk characteristics (by region, segment), CRE lending standards, and regular reporting of CRE companies’ financial condition.
- Housing-sector and borrower-based tools:
  - Introduce national legislation to include borrower-based tools in the policy toolkit for real estate sector: limits on LTV ratio, DTI/LTI ratio, and debt service-to-income ratio (DSTI); these would allow binding measures rather than consumer-protection-based guidelines and enhance SRC legitimacy in using them for macroprudential purposes.
  - Stand ready to take further action if housing-related risks persist, given incomplete toolkit and observed inaction bias on borrower-based measures.
- Operational and cross-border coordination:
  - Strengthen regional cooperation arrangements given strong interlinkages in the Nordic/Baltic region, including continued use and development of MoUs and mutual recognition arrangements to reduce leakages and facilitate enforcement.
  - DN should consider expanding human resources fully dedicated to macroprudential policy in the SRC secretariat to support continuous monitoring and faster proposal drafting.

*IMF Staff assessment based on the Staff Guidance Note (IMF, 2014a), background note (IMF, 2014b), and related IMF policy papers as presented in the source chapter.*

### Box 1. International Comparison of Debt Limits and the Use of Speed Limits

### Box 1. International Comparison of Debt Limits and the Use of Speed Limits

### International adoption of DTI/LTI caps
- As of now, five high-income countries have introduced DTI/LTI caps as a complement to their LTV caps: UK in 2014, Ireland in 2015, Norway in 2017, and Czech Republic and Slovakia in 2018.
- Table. Use of DTI/LTI Caps and Speed Limits:
  - Norway
    - Measure: DTI
    - Numerator: All debt
    - Denominator: Gross annual income
    - Limit: 5
    - Exemptions: 10 percent of new loans per quarter
  - UK
    - Measure: LTI
    - Numerator: Mortgage loan
    - Denominator: Gross annual income
    - Limit: 4.5
    - Exemptions: Up to GBP 100 million per annum or extending fewer than 300 mortgages
  - Czech Republic
    - Measure: LTI
    - Numerator: Mortgage loan
    - Denominator: Net annual disposable income
    - Limit: 9
    - Exemptions: 5 percent
  - Slovakia
    - Measure: DTI
    - Numerator: All debt
    - Denominator: Net annual disposable income
    - Limit: 8
    - Exemptions: 10 percent
  - Ireland
    - Measure: LTI
    - Numerator: Mortgage loan
    - Denominator: Gross annual income
    - Limit: 3.5
    - Exemptions: 20 percent
- Source for table: IMF Macroprudential Policy Survey.

### Borrower-based measures and housing market recommendations
- Overall assessment:
  - Further tightening of borrower-based measures would help address residual risks related to the housing market.
  - Due to uncertainty and the economic downturn caused by the COVID-19 crisis, further tightening is undesirable at the current juncture; authorities should continue monitoring developments and stand ready to act in case of an abrupt correction in house prices.
- Specific recommendations if vulnerabilities persist:
  - Introduce a stricter LTV limit if house price valuation remains elevated.
    - Micro data suggests households’ NPLs respond non-linearly with LTVs with an observed jump after the LTV level of 85 percent.
    - Earlier IMF work notes that the LTV limit should be lowered from 95 to 90 percent to better protect households from house price declines.
    - The selected issues paper finds that reducing LTVs from 95 to 90 percent would lower aggregate consumption by about 1.5 percentage points one year after introduction but increase it by 0.2 percentage points in a new steady-state because of lower debt-servicing costs.
  - Introduce strict DTI limits for all borrowers irrespective of LTV considerations if the uptrend in house price continues.
    - In an economy with elevated house prices, rules targeting LTVs become less binding; income-based measures (DTI, LTI, DSTI) may be more effective.
    - Authorities should introduce a binding DTI limit (prudently calibrated using, for example, the credit registry database) such that households above this limit are not granted loans.
    - Speed limits can be considered to minimize negative impact on the real economy.
  - Mandatory amortization for highly-leveraged households:
    - Households with debt-to-income above 400 percent should be subject to mandatory amortization, irrespective of amortization periods.
- Structural policy complements:
  - Limit debt-bias in mortgage finance by:
    - (i) further reducing mortgage interest deductibility, taking advantage of the current low rate environment;
    - (ii) balancing tax-incentives on savings products that incentivize large mortgage borrowing;
    - (iii) fastening the process of re-establishing the link between taxes payable and current residential real estate market prices (currently postponed to 2024).
  - Reduce rent control to stimulate the rental market.
  - Relax supply constraints, e.g., relax restrictions on the size of new apartments in urban areas and streamline zoning and planning procedures across municipalities.

### Corporate sector and commercial real estate (CRE) vulnerabilities
- Aggregate corporate debt:
  - Corporate debt-to-GDP at around 90 percent is stable and benign compared to peers, but riskiness of credit allocation has increased since 2013 (leverage-based indicator).
  - Corporate lending standards have eased; nominal corporate credit growth has picked up after a long post-GFC slump.
- Corporate financial condition:
  - Debt service to income for corporates fell from a peak of 50 percent during the GFC to 35 percent in 2019.
  - NFC default rate fell from 5.3 percent in 2009 to 1.5 percent in 2017.
- CRE-specific vulnerabilities:
  - CRE is the most leveraged sector in Denmark and accounts for about 30 percent of the total corporate debt.
  - CRE prices have continued to increase while required yields trend down, making the sector vulnerable to repricing of risks.
  - Vacancy rates have increased in some segments, raising concerns about income-generating capacity.
  - Foreign investors account for a considerable share of transactions, particularly in Copenhagen, which can amplify boom-bust cycles.
- Stress test and adverse scenario:
  - Firm-level analysis indicates a six-fold jump in debt-at-risk for the CRE sector in an adverse scenario—far greater than other sectors.
  - Debt-at-risk is calculated by aggregating debt of firms with projected PDs above 1.5 percent.
  - Given large bank exposures and interconnectedness, specialized monitoring of CRE is needed.

### CRE-related measures and recommended macroprudential actions
- Existing Danish measures addressing CRE risks:
  - Supervisory Diamond for banks: limits the sum of commercial property lending to no more than 25 percent of total bank lending and a lending growth cap of 20 percent.
  - Supervisory Diamond for MCIs: caps lending growth at 15 percent in different segments.
  - 100 percent limit on debt-service-to-income ratio for lending by banks.
- Recommendations if risks intensify:
  - Risk weights:
    - Consider introducing temporary risk weight floors or adding temporary risk weight add-on on CRE exposures for IRB banks and MCIs using Art. 458 CRR.
    - Examples: Belgium has used Art. 458 to increase risk weights on residential real estate exposures; Sweden and Norway are proposing risk weights on CRE exposures for IRB banks; Hong Kong SAR has differential risk weights for IRB banks; UK has a slotting system with risk weights ranging from 50 to 250 percent for IRB banks.
    - Note: future introduction of Basel risk-weight floors for IRB-model banks should be part of consideration; timeline uncertain, so a temporary add-on via Art. 458 CRR might be easier in the short run.
  - Capital buffers:
    - Countries have activated systemic buffers partly on grounds of addressing CRE-related vulnerabilities: SRB activations (Croatia and Hungary); CCyB activations (Norway, the Netherlands, Australia, and Ireland).
    - CRE prices are one of the indicators used in Denmark for deciding the CCyB buffer rate.
    - Sectoral SRB/CCyB would be preferable as they are better targeted; sectoral SRB would be available within the European framework once CRDV/CRRII comes into effect.
    - Advantage: capital buffers are adjustable over the cycle and can be relaxed when needed, allowing losses to be absorbed and credit to flow.

### Funding and liquidity vulnerabilities
- Market-based funding structure:
  - Market-based funding in Denmark is focused on domestic issuance of covered bonds by MCIs.
  - Danish commercial banks mainly rely on deposits to finance retail lending and mortgages; MCIs issue covered bonds to finance mortgage portfolios.
  - Covered bonds are largely issued domestically and denominated in DKK; amount of covered bonds issued in FX is at a historic minimum of currently 4.2 percent.
- Liquidity regulation compliance:
  - Banks and MCIs must maintain a minimum LCR of 100 percent for all currencies in total and for each significant currency.
  - Banks comply with the short-term Liquidity Coverage Ratio (LCR) with a comfortable margin.
  - Net stable funding ratio (NFSR) is currently monitored and expected to be introduced as a requirement in 2021.
- HQLA concentration and rules:
  - When calculating HQLA, at least 30 percent must be government-guaranteed assets or central bank deposits and mortgage bonds must not exceed 70 percent.
  - Covered bonds play a dominant role in banks’ HQLA portfolios, complemented by government bonds and central bank deposits.
  - Concentration issues at asset class and issuer level are being closely monitored by the DFSA.
  - Significant currencies: currencies (other than Swedish kronor and Norwegian kroner) that make up more than 5 percent of liabilities are considered significant currencies for the bank/MCI.
- LCR in significant currencies:
  - The LCR in significant currencies provides an important buffer against foreign currency liquidity risks and is enforced in Denmark, not just monitored.

### Structural vulnerabilities, interconnectedness, and systemic institutions
- Interconnectedness via housing and covered bonds:
  - Housing asset exposures interlink MCIs, banks, pension funds, insurance companies, foreign investors, and the household sector.
  - High mandatory pension contributions and household savings facilitated development of the world’s largest covered bond market in percent of GDP.
  - Shocks to real estate can affect household consumption via wealth and collateral effects and via financial wealth effects through large pension savings invested in financial assets.
- Contagion and network analysis:
  - Danish credit institutions are mostly exposed to shocks from within the banking system.
  - MCIs play a central role in the domestic interbank system, linking systemic and non-systemic banks through covered bond exposures and inducing highest levels of contagion losses through unlisted shares and covered bond exposures.
  - Systemic banks are more vulnerable to inward spillovers owing to their covered bond holdings.
  - Ensuring the health of the covered bond market under stressed conditions is crucial for financial stability.
- Systemically important institutions and buffers:
  - Seven institutions are assessed by the DFSA to be systemically important and therefore subject to systemic risk buffer (SRB).
  - Criteria for O-SII identification (only one needed):
    - total assets in per cent of domestic GDP > 6.5 per cent
    - loans in per cent of the total lending by the domestic sector > 5 per cent
    - deposits in per cent of the total deposits of the domestic sector > 3 per cent
  - Allocation and buffer rates (Table 5):
    - Subcategory 1: Score <=5 — Buffer Rate 1 percent — Sydbank; DLR Kredit; Spar Nord Bank A/S
    - Subcategory 2: Score [5-15[ — Buffer Rate 1.5 percent — Jyske Bank; Nordea Kredit Realkreditaktieselskab
    - Subcategory 3: Score [15-25[ — Buffer Rate 2 percent — Nykredit Realkredit
    - Subcategory 4: Score [25-35[ — Buffer Rate 2.5 percent — (no institution listed)
    - Subcategory 5: Score >=35 — Buffer Rate 3 percent — Danske Bank
  - Note: For Spar Nord Bank A/S the systemic risk buffer is set to 0.5 per cent at the end of 2019 and 1 per cent at the end of 2020 according to the Danish Financial Business Act.
  - Under CRD V, SRB can no longer be used as a substitute for systemically important institutions buffer; Denmark will have to shift its SRB to the O-SII buffer and SRB can be used to build buffers for other risks in the economy.

*Source: IMF Macroprudential Policy Survey and Denmark chapter content provided.*

### Appendix I. Growth at Risk

### Appendix I. Growth at Risk

### Methodology
- Framework: Growth-at-Risk (GaR) approach developed in October 2017 GFSR.
- Purpose: Links current financial conditions to the distribution of future growth outcomes to assess whether a tightening or an easing of financial conditions is macro-critical for financial stability and future growth.
- Variables used (three groups):
  - (i) financial conditions or price of risk: real short term interest rate, interbank spread, sovereign local debt spread, corporate local debt spread, equity prices, and real house prices;
  - (ii) leverage: household debt-to-GDP gap, credit growth (y-o-y change);
  - (iii) external factors: quarterly euro area GDP growth, US 10yr yield, VIX, global liquidity indicator and oil prices.
- Definition: Financial conditions are tight (seen from a historical perspective) when above zero and loose otherwise.
- Note on horizons: Short- and long-term refer to 4 and 12 quarters ahead, respectively.

### Key Findings
- Current loose financial conditions:
  - Stimulate growth in the near term.
  - Increase risks of a medium-term downturn.
- Tail-risk metric:
  - The 5th percentile of the conditional GDP distribution is used as the metric for growth-at-risk.
  - In the severely adverse scenario, one-year ahead GDP growth in Denmark will be about 1 percent or less.
  - In the severely adverse scenario, GDP growth could turn negative in a three year ahead period.
- Impact of sudden tightening of financial conditions:
  - A sudden tightening of financial conditions is associated with increased downside risks to growth.
  - Tightening of both domestic and global financial conditions will increase growth-at-risk.
- Leverage effects:
  - Higher household leverage increases growth outlook in the short run (one-year ahead) by boosting consumption and output.
  - Higher household leverage increases downside risks in the three-year horizon by making household balance sheets more susceptible to shocks.

### Scenarios and Simulations (as described)
- Conditional distributions of future real GDP growth simulated for:
  - 4 quarters ahead (short-term) and 12 quarters ahead (long-term).
  - One standard deviation shocks are simulated.
- Comparative scenario results described:
  - Before shock vs After shock distributions shown for:
    - Short-term leverage shock
    - Long-term leverage shock
    - Global financial conditions shock
    - Domestic financial conditions shock

### Policy Implications and Recommendations
- Caution warranted in a period of continued loose financial conditions due to risk build-up from stretched asset valuations, risk illusion, and increased risk taking.
- Macroprudential buffers recommended to build resilience in an adverse scenario.
- Systemic Risk Council (SRC) monitoring:
  - The SRC is already monitoring GaR as part of its macrofinancial surveillance framework.
  - The SRC has been closely monitoring risks associated with prolonged periods of loose financial conditions and had increased the CCyB to build resilience in periods of stress.
  - The release of the buffer with the Covid-19 pandemic should add much needed resilience to the banking system and the real economy.

*Prepared by Umang Rawat and Elizabeth Mahoney (both IMF). Sources: IMF staff calculations.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1dnkea2020004.pdf_