## 1norea2020007

## Source details

**Canonical URL:** [1norea2020007](https://www.imf.org/-/media/files/publications/cr/2020/english/1norea2020007.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2020/english/1norea2020007.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2020/english/1norea2020007.pdf.json)

---

### Executive summary — institutional framework and willingness to act
- The Ministry of Finance (MoF) is the sole macroprudential decision-maker in Norway; this arrangement is described as “almost unique” in international comparison.
- Norges Bank and the Finanstilsynet (FSA) provide systemic risk analysis and recommendations to the MoF and play important advisory and analytical roles.
- The macroprudential policy toolkit is “well stocked and actively used.”
- A dominant MoF role could risk delaying macroprudential action because of inaction bias; nevertheless, Norway has shown strong willingness to prioritize long-term benefits and has taken decisive macroprudential actions recently.
- Institutional features supporting willingness to act include:
  - longstanding MoF mandate to safeguard financial stability;
  - well-defined objectives at Norges Bank and the FSA;
  - dedicated financial stability units;
  - established accountability and communication frameworks.

### Systemic vulnerabilities and monitoring
- Key vulnerabilities (exact indicators preserved):
  - Total credit-to-GDP, household debt-to-income, and gross corporate credit-to-GDP are “at or close to an all-time high.”
  - Residential and commercial real estate prices (residential measured against household disposable income) are “at or close to an all-time high.”
  - Share of households with debt exceeding five times their income is at a “record level” and so is their share in overall household debt.
  - Heavy reliance on covered bonds for funding of real estate loans implies significant reliance on wholesale funding.
- Systemic risk monitoring capacity:
  - Norges Bank and the FSA have well-developed monitoring frameworks and publish financial stability reports.
  - A credit registry has recently been established, improving data availability.
  - Remaining important data gaps:
    - micro-data on NPLs and financial distress for households linked to existing micro-data sets;
    - micro-data on the CRE sector (including DRE companies);
    - these gaps limit tool calibration and risk assessment.

### Macroprudential toolkit assessment and stance
- Borrower-based tools and capital requirements:
  - Borrower-based tools have been tightened since the 2015 FSAP for mortgages and consumer loans.
  - Banks hold large capital buffers; Norway continued on the forefront of introducing Basel III requirements after the GFC.
  - Bank profitability has remained strong, serving as a first buffer to absorb losses.
- Effectiveness and design considerations:
  - Recent macroprudential actions “seem to have been effective in slowing down some of the riskier trends.”
  - Given persistent housing debt risks, temporary household measures (such as the debt-to-income cap) should be considered for permanence, with durations matching the structural nature of the risks addressed.
  - Speed limits are a “useful feature” and should remain in place.
  - Consider broadening the toolkit to mitigate CRE vulnerabilities, including sectoral capital tools.
  - Clarify that the liquidity coverage ratio (LCR) in significant currencies is a buffer that could be relaxed during periods of foreign currency liquidity stress.

### Key policy recommendations (selected and preserved timing legend)
- Institutional arrangements
  - 1. Develop a macroprudential policy strategy in cooperation with Norges Bank and the FSA. — Agency: MoF — Timing: ST
  - 2. Use triparty meetings more effectively to reach agreements on risks and policy actions. — Agency: MoF, NB, FSA — Timing: I
  - 3. Give Norges Bank recommendation powers over macroprudential tools that can be relaxed under stress with a ‘comply-or-explain’ mechanism. — Agency: MoF — Timing: I
  - 4. Publish a list of all recommendations made by Norges Bank and the FSA to the MoF in the Annual Financial Market Report to Parliament. — Agency: MoF — Timing: I
- Systemic risk monitoring
  - 5. Collect micro-data on NPLs and financial distress for households and link to existing micro-data sets. — Agency: NB, MoF — Timing: ST
  - 6. Collect micro-data on the CRE sector to facilitate risk assessment and guide tool calibration. — Agency: NB — Timing: ST
- Toolkit
  - 7. Make key parts of the household sector measures permanent, to be adjusted infrequently and subject to speed limits. — Agency: MoF — Timing: ST
  - 8. Reduce tax preferences and relax supply constraints on housing. — Agency: MoF — Timing: ST
  - 9. Consider options to broaden the toolkit to mitigate CRE vulnerabilities, including sectoral capital tools. — Agency: MoF, NB, FSA — Timing: MT
  - 10. Clarify that the LCR in significant currencies is a buffer that could be relaxed during periods of foreign currency liquidity stress. — Agency: MoF — Timing: ST
- Timing legend: I Immediate (within 1 year); ST Short term (within 1-2 years); MT Medium Term (within 3-5 years)

### Systemic risk assessment frameworks and data priorities
- Norges Bank frameworks and data:
  - Traditional four indicators for CCyB advice: (i) credit-to-GDP ratio for Mainland Norway; (ii) ratio of house prices to disposable income; (iii) estimated CRE prices; (iv) banks’ wholesale funding ratio.
  - New four-area framework: (1) financial imbalances; (2) access to credit; (3) banks’ capacity to absorb losses; (4) effects of buffer changes.
  - Micro-data access: household-level database (income, debt, assets, age) and corporate balance sheet data exist but with lags close to two years.
  - Heatmap, composite indicator of systemic risk (CISS), crisis probability models, and macroprudential stress tests are used.
- FSA tools:
  - Three quantitative models in top-down stress tests: NAM-FT, Sebra, and Bank model.
  - Yearly mortgage survey covering approximately 90 percent of total lending market; survey covers approximately 12 percent of banks' new lending in the current quarter.
- MoF monitoring:
  - MoF is developing its own quantitative models and a heatmap; risk that duplication reduces efficiency.
- Data gaps and priorities (explicit items):
  - Closed gaps: micro-data on unsecured household debt; cross-holdings of securities.
  - Ongoing/remaining gaps:
    - Micro-data on households’ NPLs and payment history are missing from new credit registries; credit reporting companies possess payment history data.
    - Norges Bank purchased payment-history data from the largest credit reporting company only for corporates.
    - CRE data gaps across: (i) prices, transactions, rentals, yields; (ii) bank exposure by size and risk distribution; (iii) CRE companies’ financial conditions including DSCR, LTV, profitability; (iv) terms of financing (yield spread, size) through bond markets.
  - Priority recommendations:
    - Collect micro-data on NPLs and link to existing micro-data sets to guide calibration of borrower-based tools (e.g., DTI limit of 5).
    - Collect micro-data on the CRE sector: quarterly bank reporting for largest banks, centralized market data collection, bank exposure detail by region/segment, CRE firms’ financial-condition reporting, CRE bond financing data.
    - Implement at-risk-analyses beyond GDP (e.g., wholesale funding and house prices).

### Household sector vulnerabilities, borrower-based tools, and calibration
- Household sector indicators (preserved figures):
  - Household debt is at 230 percent of disposable income, an all-time high.
  - Share of households with debt exceeding five times gross income has continued to rise; their share of total household debt is also at an all-time high (to year-end 2017).
  - Close to 20 percent of new loans in 2018 exceeded the 4.5 DTI level suggested by the FSA.
- Timeline of borrower-based measures (selected):
  - 2010: Guidelines on Mortgage Lending introduced (LTV 90 percent; home equity credit lines 75 percent; amortization if LTV>70 percent; 5 percent interest rate stress test).
  - 2015: Mortgage regulations (LTV 85 percent; LTV home equity 70 percent; amortization 2.5 percent annually if LTV>70 percent; 5 percent stress test; speed limit 10 percent).
  - 2017: DTI limit introduced at 5 times gross annual income; LTV limit on secondary homes in Oslo at 60 percent; speed limit 8 percent in Oslo and 10 percent elsewhere; consumer credit DTI limit 5 and amortization requirement 5 years.
  - 2019: CCyB increased to 2.5 percent.
- Effectiveness and design considerations:
  - FSA surveys show borrower-based tools have been effective; clustering of new loans close to DTI and LTV caps observed.
  - Speed limits are effective and provide flexibility; Norway’s 10 percent speed limit is broadly in line with other countries.
  - Recommendation: make key household measures permanent, adjusted infrequently, and include speed limits; match duration of measures to structural nature of risks.

### Corporate sector and CRE vulnerabilities; toolkit for CRE
- Corporate sector facts:
  - Net corporate debt around 80 percent of GDP.
  - CRE accounts for the largest part of corporate loan book: from 25 percent (2000) to 40 percent of corporate bank loans.
  - Among advanced European countries, Norway has one of the highest bank exposures to CRE at 16 percent of total lending.
  - Non-bank financing via bond issuance rose from 2 percent in early 2000 to 8 percent in 2017.
  - Insurance companies’ financing to CRE accounts for 3 percent of insurance companies’ total portfolio in Norway.
- Banks’ exposure and heterogeneity:
  - Banks’ share of loans to commercial property as share of all granted loans to nonfinancial firms varies from 15 to 93 percent.
  - Seven largest banks account for just over three-quarters of Norwegian banks’ loans to the property industries; exposures vary from 26 to 68 percent among those banks.
  - DNB constitutes over 40 percent of the banking sector by size; MoF identified DNB and Kommunalbanken as systematically important.
- Funding and liquidity vulnerabilities:
  - Wholesale funding provides close to half of all funding for banks and mortgage companies; deposits are around 40 percent relative share.
  - About two-thirds of wholesale funding are covered bonds collateralized by mortgages; half of covered bonds are foreign-currency denominated and swapped to domestic currency.
  - LCR and NSFR: minimum LCR of 100 percent for all currencies in total and for each significant currency; banks with EUR and/or USD as significant currencies must maintain minimum LCR in NOK of 50 percent; NSFR monitored but not enforced.
- Policy options for CRE vulnerabilities (preserved recommendations):
  - Broaden toolkit beyond broad-based capital tools; consider sectoral capital measures (e.g., risk weight floors on IRB banks), sectoral CCyB/SRB, borrower-based measures for CRE (LTV, DSCR/ICR limits), intensified microprudential oversight (Pillar II add-ons), and tax measures (interest deduction limitations).
  - Acknowledge operational and calibration challenges for CRE borrower-side measures: valuation complexity, sectoral heterogeneity, and potential leakages to foreign investors.
  - Use data improvements and targeted bank reporting to support sectoral/regionally calibrated tools.

### Capital buffers, stress tests, regulatory changes, and chronology highlights
- Capital buffer and stress-test outcomes:
  - CCyB will increase to 2.5 percent by end-2019.
  - Norway has some of the largest capital buffers internationally.
  - FSA 2019 stress test: six (three) major banks would have CET1 (leverage) ratio below overall requirements if only the CCyB is relaxed to zero; with use of the CCB a maximum of three banks would fail; releasing part of the SRB further reduces number of banks failing.
- Regulatory changes announced by MoF (preserved specifics):
  - Raise the SRB from 3 percent of total exposures to 4.5 percent of domestic exposures, effective from end-2020 for IRB banks and phased-in gradually for standardized approach banks.
  - Introduce risk weight floors for mortgages and CRE loans at 20 and 35 percent, respectively.
  - Transposition of European regulations via the EEA agreement will reduce capital requirements compared to existing Norwegian regulations because of removal of Basel I floor for IRB banks and introduction of SME discount factor.
- Chronology (selected exact entries from 2010–2019):
  - 2010: LTV mortgages 90 percent; home equity credit lines 75 percent; 5 percent interest rate stress test.
  - 2013: SRB introduced at 2 percent.
  - 2014: SRB increased to 3 percent; LGD floor increased from 10 to 20 percent on mortgage exposure for IRB banks.
  - 2015: CCyB introduced at 1 percent; OSII-B introduced at 1 percent; mortgage rules moved from guidelines to regulation (LTV 85 percent; speed limit 10 percent).
  - 2016: CCyB increased to 1.5 percent; OSII-B to 2 percent; LCR all currencies introduced at 100 percent.
  - 2017: CCyB increased to 2 percent; DTI limit introduced at 5; LTV for secondary homes in Oslo introduced at 60 percent; speed limit 8 percent in Oslo, 10 percent elsewhere; LCR for significant currencies introduced at 50 percent.
  - 2018: LCR Requirement for all currencies increased to 100 percent.
  - 2019: CCyB increased to 2.5 percent; guidelines on consumer credit changed to regulation.

### Liquidity policy, LCR flexibility, and recommendations
- Liquidity and covered bond risks:
  - Covered bonds serve as a large part of HQLA; banks hold each other’s covered bonds to fulfill LCR.
  - Liquidity in covered bond markets could fall in systemic events, raising refinancing and collateral posting risks; mark-downs could hurt solvency of banks and insurance/pension funds.
- LCR as releasable buffer:
  - Recommendation: clarify that the LCR in significant currencies is a buffer that could be relaxed during foreign currency liquidity stress, subject to careful judgement and three conditions:
    - (i) buffers are in place;
    - (ii) foreign currency liquidity pressures are generating financial stress;
    - (iii) relaxation is expected to relieve stress and contain procyclical dynamics.
  - Illustration cited: Croatia relaxed foreign currency liquidity requirements and released liquid assets amounting to over 14 percent of GDP.

*Source: IMF staff compilation from EXECUTIVE SUMMARY and chapter excerpts, Norway country report (1norea2020007).*

### EXECUTIVE SUMMARY ____________________________________________________________________________ 6

### EXECUTIVE SUMMARY

### Key findings on institutional framework and willingness to act
- The Ministry of Finance (MoF) is the sole macroprudential decision-maker in Norway; this arrangement is “almost unique” in international comparison.  
- Norges Bank and the Finanstilsynet (FSA) play important advisory and analytical roles, providing systemic risk analysis and recommendations to the MoF.  
- The authorities have taken substantive and wide-ranging macroprudential policy actions in recent years; the macroprudential policy toolkit is “well stocked and actively used.”  
- A dominant role of the MoF could risk delaying macroprudential action because political agents may be more vulnerable to inaction bias; however, Norway has exhibited strong willingness to prioritize long-term benefits and has taken decisive macroprudential actions recently.  
- Institutional features supporting willingness to act in Norway include: a longstanding MoF mandate to safeguard financial stability, well-defined objectives at Norges Bank and the FSA, dedicated financial stability units, and established accountability and communication frameworks.

### Systemic vulnerabilities and systemic risk monitoring
- Financial sector vulnerabilities remain large:
  - Total credit-to-GDP, household debt-to-income, and gross corporate credit-to-GDP are “at or close to an all-time high.”  
  - Commercial and residential real estate prices (residential measured against household disposable income) are “at or close to an all-time high.”  
  - The share of households with debt that exceeds five times their income is at a “record level” and so is their share in overall household debt.  
  - Covered bonds play a key role in banks’ funding of real estate loans, implying a heavy reliance of banks on wholesale funding.
- Systemic risk monitoring capacity:
  - Norges Bank and the FSA have well-developed systemic risk monitoring frameworks and publish financial stability reports.  
  - A credit registry has recently been established, improving data availability.  
  - Important data gaps remain: notably micro-data on NPLs and financial distress for households linked to existing micro-data sets, and micro-data on the CRE sector (including DRE companies), which would support tool calibration and risk assessment.

### Macroprudential toolkit assessment and stance
- Borrower-based tools and capital requirements:
  - Borrower-based tools have been tightened since the 2015 FSAP for mortgages and consumer loans.  
  - Banks hold large capital buffers; Norway continued on the forefront of introducing Basel III requirements after the GFC.  
  - Bank profitability has remained strong, serving as a first buffer to absorb losses.
- Effectiveness and design considerations:
  - Recent macroprudential actions “seem to have been effective in slowing down some of the riskier trends.”  
  - Given persistent housing debt risks, temporary household measures (such as the debt-to-income cap) should be considered for permanence, with durations matching the structural nature of the risks addressed.  
  - Speed limits are a “useful feature” that reduce the cost of mortgage restrictions on banks and households and should remain in place.  
  - There are merits in broadening the toolkit to mitigate CRE vulnerabilities, including considering sectoral capital tools to achieve a more targeted macroprudential stance.  
  - Clarification is needed that the liquidity coverage ratio (LCR) in significant currencies is a buffer that could be relaxed during periods of foreign currency liquidity stress.

### Key policy recommendations (Table 1)
- Institutional Arrangements
  - 1. Develop a macroprudential policy strategy in cooperation with Norges Bank and the FSA (Paragraphs 18−20; 22). — Agency: MoF — Timing: ST
  - 2. The triparty meetings should be used more effectively to reach agreements on risks and policy action needed to address them (Paragraphs 15−17, 19, 23). — Agency: MoF, NB, FSA — Timing: I
  - 3. Give Norges Bank recommendations powers over macroprudential policy tools that can be relaxed under stress with a ‘comply-or-explain’ mechanism (Paragraphs 11, 20, 24). — Agency: MoF — Timing: I
  - 4. Publish a list of all recommendations made by Norges Bank and the FSA to the Ministry of Finance in the Annual Financial Market Report to Parliament (Paragraphs 17, 25). — Agency: MoF — Timing: I
- Systemic Risk Monitoring
  - 5. Collect micro-data on NPLs and financial distress for households and link the information to existing micro-data sets to guide calibration of borrower-based tools (Paragraphs 33−34). — Agency: NB, MoF — Timing: ST
  - 6. Collect micro-data on the CRE sector to facilitate risk assessment and guide the development and calibration of macroprudential policy tools (Paragraphs 32, 35). — Agency: NB — Timing: ST
- Toolkit
  - 7. Make key parts of the household sector measures permanent, to be adjusted infrequently and subject to speed limits (Paragraphs 49−50, 53−54). — Agency: MoF — Timing: ST
  - 8. Reduce tax preferences and relax supply constraints on housing (graphs 45, 56). — Agency: MoF — Timing: ST
  - 9. Consider options to broaden the toolkit to mitigate CRE vulnerabilities, including sectoral capital tools (Paragraphs 61, 63). — Agency: MoF, NB, FSA — Timing: MT
  - 10. Clarify that the liquidity coverage ratio in significant currencies is a buffer that could be relaxed during periods of foreign currency liquidity stress (Paragraphs 20, 66−68). — Agency: MoF — Timing: ST
- Timing legend used in Table 1:
  - I Immediate (within 1 year); ST Short term (within 1-2 years); MT Medium Term (within 3-5 years)

### Analytical priorities and next steps
- Strengthen interagency coordination and the policy process:
  - Publication of a macroprudential policy strategy is recommended to reduce inaction bias, foster accountability, strengthen external communications, and prepare markets for possible buffer relaxation in a sharp downturn.  
  - The semiannual triparty meetings should be used more effectively to reach agreement on risks and jointly discuss specific policy actions.  
  - Norges Bank should be given recommendation powers over capital and liquidity tools that can be relaxed, in addition to its current role over the countercyclical capital buffer (CCyB).  
  - The authorities should clarify that the LCR in significant currencies is a buffer that could be relaxed during foreign currency liquidity stress.
- Data and calibration:
  - Prioritize collecting micro-data on household NPLs and financial distress and linking them to existing micro-data sets to guide borrower-based tool calibration.  
  - Prioritize collecting micro-data on the CRE sector (including DRE companies) to facilitate monitoring, analysis, and the potential development of new instruments.

*Source: IMF staff compilation from EXECUTIVE SUMMARY, Norway country report (1norea2020007).*

### 10.      While inaction risks have not materialized in Norway so far, there are no guarantees

### 10.      While inaction risks have not materialized in Norway so far, there are no guarantees

### Principle 2: Ability to Act
- The Norwegian institutional arrangements for macroprudential policy are characterized by strong powers and thereby ability to act.
- The MoF has a broad financial stability mandate; its regulatory perimeter includes non-bank financial institutions (NBFIs) and financial markets.
- The MoF has hard (direct) powers over all macroprudential tools.
- Norges Bank and the FSA have both semi-hard and soft powers as they provide recommendations to the MoF, sometimes with an explicit ‘comply-or-explain mechanism’ (e.g., regarding the countercyclical capital buffer (CCyB) and identifying systematically important institutions, but not for the SRB).
- In practice, the FSA has provided ‘comply-or-explain’ advice on the mortgage regulation.
- The authorities have taken more macroprudential actions in recent years than typically observed in other advanced economies; actions relied in particular on broad-based capital tools and borrower-based household sector tools (see Figure 2).
- Information powers:
  - The FSA can require all supervised institutions to provide any information needed to fulfill its mandate.
  - Under the previous Norges Bank Act, the central bank could, through the MoF, attain any information from the financial sector necessary to fulfill its mandate.
  - Norges Bank has had access to micro-level data from Statistics Norway for research purposes.
  - The new Norges Bank Act (¶5-3) strengthens and extends Norges Bank’s information powers beyond the financial sector: Norges Bank can now attain information directly from the tax authorities, Register of Company Accounts, the Register of Bankruptcies, and the Norwegian Labor and Welfare Administrations—potentially reducing lags in micro-level household and business data.

### Principle 3: Effective Coordination and Cooperation
- Cooperation across the MoF, Norges Bank and the FSA has a long tradition, with regular bilateral meetings predating the late 1980s financial crisis and tri-party meetings on some occasions.
- Since 2006, informal tri-party meetings have been held, usually bi-annual (more frequently if needed), to discuss risks and exchange information—not to take policy actions.
- Meeting practices and frequency:
  - Tri-party meetings: Informal bi-annual (usually June and December) meetings between the Permanent Secretary of the MoF, Norges Bank Governor, and the FSA Director General; venue for discussion of risks and exchange of information and views, not a decision-making body.
  - Bilateral meetings on the CCyB: Quarterly high-level meetings between the Minister of Finance and Norges Bank Governors, taking place a day before the CCyB decision is announced; additional technical-level meetings occur a day earlier and one to two weeks before the decision.
  - Bilateral MoF–FSA meetings: Quarterly meetings between the Head of the Financial Market Department of the MoF and the Director General of the FSA.
  - Bilateral Norges Bank–FSA meetings: Bi-annual high-level meetings (usually August and January) and quarterly technical-level meetings.
- The Minister of Finance does not attend tri-party meetings; as a result, policy deliberations in those meetings are scarce and primarily limited to information exchange.
- Policy discussions often occur through public hearings and formal letters initiated by the MoF; the MoF decides whether to initiate a public hearing on proposals from the FSA or Norges Bank.
  - There is no mechanism that ensures the MoF must take a public position on all recommendations made by the FSA or Norges Bank.
- Public hearings have revealed fundamental differences of opinion among the three institutions on borrower-based tools for the housing sector, including:
  - Whether tools should be temporary (MoF) or permanent (FSA and Norges Bank).
  - Whether tools should be mostly structural (Norges Bank) or adjusted over time (FSA).
  - Whether speed limits should be applied (MoF and Norges Bank) or not (FSA).
  - Whether tools should vary across regions (MoF) or not (Norges Bank and FSA).
  - Whether tighter LTV caps should apply to secondary properties (MoF vs. FSA).
- Exclusive reliance on letter correspondence risks weakening policy coordination and prolonging repeated strategic disagreements.
- Market participant concerns:
  - Uncertainty about if, and under what circumstances, macroprudential policy would be relaxed.
  - Uncertainty whether relaxation of some buffers might be counteracted by stricter requirements from the FSA if banks ‘eat into’ other buffers (e.g., the capital conservation buffer (CCB)).
  - Uncertainty whether authorities would consider relaxing buffers other than the CCyB, in particular the systemic risk buffer (SRB) and the Liquidity Coverage Ratio (LCR) for significant currencies.
- Cross-border dimension:
  - Foreign bank branches constitute around 35 percent of total bank assets.
  - Norwegian banks rely heavily on foreign funding and have operations abroad (in particular DNB).
  - Cross-border MoUs and participation in the ESRB have been used to strengthen cooperation, support enforcement, and reduce leakages.

### Recommendations
- The MoF should develop a macroprudential policy strategy in cooperation with Norges Bank and the FSA. The strategy should:
  - Elaborate ultimate and intermediate objectives of macroprudential policy.
  - Explain the decision-making process leading up to macroprudential action.
  - Set out expected transmission of macroprudential tools and relate them to particular intermediate objectives.
  - Be flexible and periodically updated to reflect new data, analytical insights, and changes in the toolkit.
- Benefits of publishing an ex ante strategy:
  - Continue overcoming the inaction bias by identifying intermediate objectives and indicators to guide use of macroprudential tools while preserving judgment-based decisions.
  - Foster accountability and anchor/strengthen communication by making it easier to hold the MoF accountable and enabling MoF, Norges Bank and the FSA to relate communications to the strategy.
  - Clarify strategic positions (e.g., permanence of borrower-based tools, application of speed limits, regional variation) to reduce repetitive public exchanges of letters defending long-standing positions.
  - Prepare market participants and the public for a relaxation of macroprudential buffers by setting out that buffers may be relaxed to avoid disruption of credit—while communicating carefully to avoid interpretation as an unconditional commitment.
  - Facilitate coordination across MoF, FSA and Norges Bank to reduce the risk of conflicting micro- and macroprudential signals to banks.
- Strengthen tri-party meetings to reach agreement on risks and policy actions:
  - MoF should be represented by the Minister to facilitate meaningful policy discussions.
  - Enhance cooperation on risk assessment ahead of meetings.
  - Seek agreement on a summary of key risks and publish the summary in a public record of the meetings.
  - Norges Bank and the FSA should prepare an overview of key risks ahead of tri-party meetings; discuss and agree a summary to be made public in meeting records.
  - Over time, consider transforming tri-party meetings into a formal macroprudential policy committee.
- Give Norges Bank recommendation powers over macroprudential policy tools that can be relaxed, with a ‘comply-or-explain’ mechanism:
  - The SRB is a key example; its recommendation process is unclear despite the SRB being a releasable buffer.
  - Similar recommendation powers should apply to the LCR in significant currencies and to other releasable tools that might be introduced later (e.g., a sectoral CCyB).
  - Granting Norges Bank semi-hard powers over the SRB is important given high and concentrated bank exposures to the real estate sector and the potential need to relax the SRB in a severe downturn to avoid procyclical feedbacks.
  - Financial supervisors are usually not in charge of releasable tools because relaxation decisions can create conflicts between micro- and macroprudential considerations.
  - The SRB’s role as the flexible capital-based tool in CRR/CRD-IV and considerations to apply it to mitigate sectoral risks further support giving Norges Bank semi-hard powers over it.

*Source: 1norea2020007, IMF Norway country report chapter.*

### 25.      A comprehensive overview of recommendations made by Norges Bank and the FSA

### 25.      A comprehensive overview of recommendations made by Norges Bank and the FSA

### Overview: Recommendation to the Ministry of Finance (MoF)
- Norges Bank and the FSA normally make their recommendations to the MoF publicly available, but there is no comprehensive overview on the extent to which the MoF follows their advice.
- Recommendation: On an annual basis, gather all recommendations from Norges Bank and the FSA in a table in the MoF’s Financial Market Report to Parliament, together with explanations about how the MoF has followed, or deviated from, the advice.
- Purpose: Foster transparency on the advice provided and facilitate accountability with regard to policy actions taken (or not taken) by the MoF.

### Systemic Risk Monitoring — key principles
- Macroprudential policy cannot rely on automatic rules and must be based on continuous assessment of evolving risks.
- Approach: “Guided discretion” — use key indicators to generate signals but rely on judgement that takes into account all relevant information.
- Requirements: Access to data and analytical capacity to assess systemic risks and map risk assessment into policy recommendations and action.
- Note: The strategy could clarify that a relaxation of the SRB would likely only take place in a severe downturn and only after the release of the CCyB has been exhausted.

### A. Systemic Risk Assessment Frameworks — institutional roles and tools
- Norges Bank and the FSA have published financial stability reports since the mid-1990s and have developed sophisticated risk monitoring frameworks.
- The MoF publishes risk assessments (mostly in Norwegian only) but relies heavily on inputs from Norges Bank and the FSA.

Norges Bank — tools, indicators, and ongoing development:
- Organizational resources: Financial Stability Department (FSD), research and market departments contribute to systemic risk monitoring.
- Models and frameworks under development: GDP-at-risk framework, macro-contagion stress tests, structural measures of financial cycles.
- Traditional set of four indicators that played a key role for CCyB advice:
  - (i) credit-to-GDP ratio for Mainland Norway;
  - (ii) ratio of house prices to disposable income;
  - (iii) estimated CRE prices;
  - (iv) banks’ wholesale funding ratio.
  - Method: Compare key indicators with historical trends; the gap between indicators and estimated trends serves as a measure of financial imbalances.
- New framework emphasizing four areas of assessment (reduced emphasis on the four indicators):
  - First: assessment of financial imbalances to assess cyclical systemic risk, comprising three elements:
    - (a) pricing of risk and lending conditions;
    - (b) real estate market vulnerabilities;
    - (c) vulnerabilities in the household and corporate sectors.
  - Second: assessment of access to credit to evaluate whether a reduction in the buffer rate is needed; uses information on:
    - (a) financial market stress;
    - (b) developments in credit and credit practices;
    - (c) banks’ profitability.
  - Third: assessment of banks’ capacity to absorb losses based on banks’ capital adequacy, earnings and loss prospects, including stress tests.
  - Fourth: assessment of the effects of a change in the buffer requirement on banks and the economy (capital raising, dividend policy, loan pricing, and whether reductions increase willingness to lend).
- Micro data access:
  - Norges Bank has access to detailed micro data on households and enterprises for research purposes, but with considerable lag (close to two years).
  - Household-level database: includes information on income, debt, financial and real estate assets, and age.
  - Corporate data: annual balance sheets for all enterprises, quarterly balance sheets and stock information for listed companies, and annual loan level data from banks.
  - Lags remain substantial despite digitalization.
  - Micro data on non-performing loans (NPLs) and credit ratings from the largest credit reporting company (Bisnode) is used only for corporates.
  - Limited CRE data: both Norges Bank and the FSA use information from market participants, statistics on rents and implied yields, and sales prices for some segments.
- Heatmap, composite indicator of systemic risk (CISS), and crisis probabilities:
  - Heatmap tracks indicators across three areas: (i) risk appetite and asset valuations; (ii) non-financial private sector vulnerabilities; (iii) financial sector vulnerabilities. Uses color-coding (green = low, red = high vulnerability).
  - CISS based on five submarkets: money, bond, equity, banking, and commodity and foreign exchange markets; shown promising ability to signal systemic banking crises.
  - Models on crisis probabilities are also used.
- Stress tests:
  - 2018 and 2019 Financial Stability Reports used a macroprudential stress test framework.
  - Primary focus: how macroprudential policy tools can prevent deterioration of macroeconomic developments rather than a binary “pass”/“fail” for banks.

FSA — models, surveys, and use:
- Three quantitative models in yearly top-down stress tests:
  - NAM-FT macro-model for economic projections;
  - Sebra model to allocate banks’ total corporate loan losses to individual banks;
  - Bank model to generate bank-specific profit and capital projections.
- Uses stress test results in risk-based on-site inspections and as part of decision basis when determining Pillar 2 (guidance) requirements.
- Survey data:
  - Yearly mortgage survey evaluates borrower-based macroprudential measures.
  - Sample covers approximately 90 percent of total lending market.
  - Survey covers approximately 12 percent of the banks' new lending in the current quarter.

MoF monitoring capacity:
- MoF is developing its own quantitative models and a heatmap.
- Risk noted: building another assessment framework at the MoF may reduce efficiency of the overall risk monitoring framework given existing capacity at Norges Bank and the FSA.

### B. Data Gaps — current status and priorities
- General: Data quality is generally good, but important data gaps remain.
- Closed recent gaps: micro-data on unsecured household debt and cross-holdings of securities between financial institutions.
- Ongoing work focuses on:
  - Micro data on households:
    - New debt registries established under the debt information act of 2017 will only cover unsecured household loans.
    - Working group recommended extending information services to include collateralized household debt (e.g., mortgages and car loans).
    - Debt information act contains provision for including collateralized debt later (change in regulation needed).
    - Statistics Norway launched microdata.no providing researchers access to various micro-data, including some financial data.
  - Corporate data, particularly for the commercial real estate (CRE) market:
    - Working group recommended against establishing a full credit registry for corporate loans at this stage; recommended changing reporting of banks’ lending to corporate sector from annual to quarterly for the largest banks.
    - Another working group will consider alternatives to close CRE data gaps in light of ESRB concerns.
    - Identified CRE data gaps cover four dimensions:
      - (i) market information on prices, transactions, rentals, yields, etc.;
      - (ii) bank exposure in terms of size and risk distribution;
      - (iii) CRE companies’ financial conditions including debt service capacity ratio (DSCR), LTV ratio, profitability, etc.;
      - (iv) terms of financing (e.g., yield spread, size) through bond markets.

- Important remaining gap: missing micro data on households’ NPLs that could be linked to existing data to guide policy calibration.
  - New credit registries do not include payment history.
  - Credit reporting companies possess payment history data used for credit scoring.
  - Norges Bank has purchased such data from the largest credit reporting company, but only for corporates.
  - Consequence: despite micro-level household data (including daily data from credit registries), Norges Bank cannot identify households with NPLs and their characteristics, limiting ability to assess appropriateness of the DTI limit of 5.

### C. Recommendations — data collection and analysis
- Collect micro-data on NPLs and financial distress for households and link to existing micro-data sets to guide calibration of borrower-based tools.
  - Norges Bank should use enhanced information powers in the Norges Bank Act and request the MoF to require credit reporting companies to provide micro-level data on household NPLs (to be shared with the FSA).
  - Linked NPL data could support models for probability of default (PDs) using explanatory variables including debt-to-income (DTI) level and financial margins.
  - Use case: guide setting of borrower-based tools such as the DTI limit of 5.
- Collect micro data on the CRE sector to facilitate risk assessment and guide use and calibration of macroprudential tools.
  - Steps recommended:
    - More frequent (e.g., quarterly) reporting by banks on corporate exposures, at least for the largest banks, as a first step.
    - Centralized collection of market data with better coverage to monitor valuation gaps and demand-supply imbalances.
    - Bank exposure reporting to include size and risk characteristics of underlying CRE properties (e.g., by region, segment) to monitor concentrations and guide sectoral tools.
    - Regular reporting of CRE companies’ financial conditions to assess repayment risks.
    - Data on CRE companies’ bond financing to monitor risks outside the banking system from policy leakages.
  - Norges Bank should use new information powers under the new Norges Bank Act to improve data collection with collaboration from the FSA and Statistics Norway.
- Implement at-risk-analyses beyond GDP:
  - In addition to ongoing GDP-at-risk analysis, consider at-risk-analyses for wholesale funding and house prices (following IMF examples applied to foreign wholesale funding and house prices).

### Systemic Risks and Macroprudential Tools — mapping vulnerabilities into policy
- The report maps systemic vulnerabilities into recommendations for the macroprudential toolkit following the 2014 IMF SGN approach.
- Contextual macro-financial setting:
  - Norway is a small resource-rich open economy; macrofinancial developments are strongly affected by external conditions, in particular oil prices.
  - After the oil price decline in 2014, the economy recovered with robust GDP growth, relatively low and stable inflation, and a downward trend in long-term interest rates; recovery in oil prices supported these developments.
- Credit and balance-sheet signals:
  - Total credit continues to gradually grow more rapidly than GDP; the credit ratio stands at its highest ever level.
  - Credit growth has somewhat lost momentum and is now more in line with estimated trend level.
  - Credit gap was around a very high 15 percent a decade ago and has been brought down.
  - Overall, bank assets, profitability and reliance on wholesale funding have remained stable, while leverage has improved.

*Source: IMF staff illustration.*

### 40.      Broad-based capital tools have been actively used to increase resilience and reign in

### 1norea2020007 - 40.      Broad-based capital tools have been actively used to increase resilience and reign in

### Broad-based capital tools: usage and assessment
- The CCyB will increase to 2.5 percent by end-2019.
- Norway has some of the largest capital buffers in international comparison; these buffers provide resilience against shocks given extensive imbalances in the financial system.
- Stress test results:
  - The FSA’s 2019 stress test results indicate that six (three) major banks would have a common equity tier 1 (CET1) capital (leverage) ratio below overall requirements when the hurdle rates are assumed to only relax the CCyB by setting it to zero.
  - If banks are assumed to also be allowed to make use of the CCB (as the framework for the buffer assumes), a maximum of three banks would fail the test.
  - If part of the SRB would also be released, the number of banks failing—and the need for recapitalization—would fall further.
  - These results indicate that the existing buffers would provide valuable resilience in a severe downturn.
- Evidence on costs of broad-based capital tools:
  - Bank profitability has remained strong (even compared to other countries where requirements are lower).
  - Interest margins have stayed relatively stable (even declined somewhat for mortgages).
  - Credit, consumption and output growth have not been subdued.
  - The macroprudential toolkit is dominated by capital tools, which are empirically found to be less costly than borrower-based tools.
  - The 85 percent cap on the loan-to-value (LTV), enforced by the authorities, is in line with median calibrations in a large country sample and is associated with limited costs in the form of reducing consumption growth.

### Regulatory changes announced by the Ministry of Finance (MoF)
- The MoF has announced an increase in the SRB and an introduction of sectoral risk weight floors.
- Transposition of the European regulations into Norwegian law through the EEA agreement will lead to a reduction in capital requirements compared with existing Norwegian regulations due to:
  - Removal of the Basel I floor for banks using the internal rating based (IRB) approach.
  - Introduction of the small-to-medium size enterprise (SME) discount factor in line with the European regulatory framework.
- Specific MoF announcements:
  - Raise the SBR from 3 percent of total exposures to 4.5 percent of domestic exposures, effective from end-2020 for IRB banks and phased-in gradually for banks applying the standardized approach.
  - Introduce risk weight floors for mortgages and CRE loans at 20 and 35 percent, respectively.
  - Plan to request other countries to reciprocate those floors to harmonize risk weights for Norwegian and foreign banks operating in Norway.

### Household sector vulnerabilities: key indicators and dynamics
- Household debt and related indicators:
  - Household debt is at 230 percent of disposable income, an all-time high.
  - Aggregate debt service ratio is at a historic peak, while the interest-rate burden is far below its peak in the 1990s’ crisis.
  - Mortgage loan growth has been relatively strong yet stable; consumer loan growth exhibited strong growth in recent years but has been brought down and remains a small share of overall household debt.
  - Rising share of mortgage loans in overall credit signals continued risks in the housing sector.
- House prices and regional patterns:
  - Residential house prices have been on a strong upward trend since the 1990s, with momentum weakened somewhat recently and regional differences (most pronounced increase in Oslo).
  - When measured against disposable income, house prices have trended upward for close to 30 years.
- Tax and incentive structure:
  - Two recurring taxes on residential properties: net wealth tax and property tax.
  - Discounts on assessed values: 75 percent on primary residences and 10 percent on secondary houses.
  - Only 11 percent of taxpayers pay the net wealth tax.
  - Threshold for the net wealth tax is NOK 1.48 million (€158,000).
  - Interest on mortgages is fully tax deductible.
- Distributional risk features:
  - Share of households with debt exceeding five times gross income has continued to rise, as has their share of total debt (at least to year-end 2017).
  - Household assets are large but unevenly distributed and concentrated in real estate, which is relatively illiquid and would decline in value if house prices fall.
  - The share of households with a DTI exceeding five is at an all-time high and so is their share in total household debt.
  - Close to 20 percent of new loans in 2018 exceeded the 4.5 DTI level suggested by the FSA.

### Borrower-based tools: history, effectiveness, and proposed changes
- Timeline and tools:
  - FSA guidelines on mortgage lending started in 2010.
  - Binding temporary regulations recommended in 2015 covering LTV caps, stressed financial margins, and minimum amortization requirements (with regional differences).
  - Risk weights on mortgages doubled in 2014.
  - In 2017, a DTI cap of five times gross annual income was introduced.
  - A ten percent ‘speed limit’ allows new lending to deviate from one or more requirements each quarter (eight percent in Oslo).
  - Measures related to consumer credit introduced in 2017 to reduce leakages.
- Effectiveness:
  - FSA surveys indicate borrower-based tools have been effective; clustering of new loans close to DTI and LTV caps observed.
  - Use of speed limits has been fairly stable over time and similar across Oslo and the rest of the country.
- FSA September 2019 proposals:
  - Reduce the DTI limit from 5 to 4.5.
  - Lower the general speed limit from 10 to 5 percent.
  - Remove regional differences in speed limits.
  - Remove the lower LTV limit on second homes in Oslo (FSA suggested removal of the lower LTV limit on second homes in Oslo).
  - The reduction in the DTI level represents a considerable tightening given the distribution of new loans in 2018; the tightening of the general speed limit would also represent somewhat of a tightening as banks are reluctant to stay too close to the limit.
- Norges Bank position:
  - Did not see a reason to tighten the DTI cap or the general speed limit.
  - Expressed support for removal of regional differences.
  - Opened the door to extending the lower LTV limit on second homes to the whole country instead of removing it fully.

### Considerations on design features of borrower-based tools
- Key considerations for whether borrower-based tools should be temporary, adjusted, subject to speed limits, distinguish between first and second properties, and be regionally differentiated:
  - Household sector risks have been a prominent and continuous concern in Norway; tools addressing long-standing or semi-permanent concerns should be in place.
  - Borrower-based tools (LTV and DSTI caps) are empirically effective in reducing vulnerabilities over time and provide households with buffers.
  - Given political costs of introducing measures, keeping them in place and adjusting calibration over time appears prudent and less costly than removing and later reintroducing them.
  - Speed limits provide flexibility and a favorable cost-benefit trade-off; Norway’s 10 percent speed limit appears broadly in line with other countries.
  - Tighter restrictions on investments apart from the first property can mitigate speculative investments; making the Oslo-specific LTV limit on second properties a nationwide restriction is an option.
  - Regional differences can optimize benefits vs costs when systemic risk is regional, but they entail implementation costs.

### Recommendations (household sector)
- Key elements of the household sector measures should be made permanent, to be adjusted infrequently, and include speed limits.
  - The duration of measures should match the structural nature of the risks they address—which argues for making them permanent.
  - Keeping tools in place (and adjusting them occasionally if needed) is more prudent and cost effective than removing and later reintroducing them.
  - Speed limits reduce the cost of mortgage restrictions on banks and households and should remain in place.

*Source: IMF staff analysis in 1norea2020007 - Chapter excerpts on macroprudential policy and household vulnerabilities.*

### Box 2. International Comparison of Debt Limits and the Use of Speed Limits

### Box 2. International Comparison of Debt Limits and the Use of Speed Limits

### International experience with DTI/LTI caps and speed limits
- Five high-income countries introduced DTI/LTI caps as a complement to LTV caps: UK in 2014, Ireland in 2015, Norway in 2017, and Czech Republic and Slovakia in 2018.
- Caps reported across countries are calibrated between 3.5 (Ireland) to 9 (Czech Republic).
- Cross-country comparison is complicated by differences in coverage and income definitions:
  - Coverage: overall debt (Norway and Slovakia) vs. mortgage loan (UK, Czech Republic, and Ireland).
  - Income: gross annual income (Norway, UK, Ireland) vs. net annual disposable income (Czech Republic and Slovakia).
- Speed limits on newly approved loans vary from 5 (Czech Republic) to 20 (Ireland, Slovakia) percent.
- Speed limits are also applied to other borrower-side macroprudential measures (e.g., LTV ratios in New Zealand and Malta) ranging from 5 to 20 percent.

### Table highlights: Use of DTI/LTI Caps and Speed Limits
- Measures and calibration (as presented in the source table):
  - 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 the table: IMF Macroprudential Policy Survey.

### Norway: assessment of borrower-based tools and calibration considerations
- Current stance: further tightening of borrower-based tools does not appear urgent; mortgage and consumer lending is losing momentum.
- Recommendation on calibration:
  - A lower DTI for structural reasons could be argued, but recalibration should be grounded in data analysis.
  - Suggested analysis: link micro data on NPLs to household debt characteristics (including DTI levels) to identify an appropriate structural DTI level.
- Operational note: the Ministry of Finance (MoF) did not follow the advice of the FSA at the time of the assessment.

### Tax preferences, supply constraints, and sequencing
- Tax incentives and mortgage interest deductibility:
  - Recent tax reform reduced the valuation discount on houses in the net wealth tax calculation, but the discount remains sizable; further alignment between houses and other wealth is needed.
  - Consider gradually phasing out mortgage interest deductibility, e.g., lowering the cap and applying deduction only on the primary residence.
  - Reforming interest deductibility should be cautious and gradual because house prices can respond rapidly, posing risks to economic stability.
  - The current low interest rate environment provides an opportunity to act before higher rates increase mortgage burdens.
- Supply constraints:
  - Construction and urban planning regulations are perceived as quite strict, especially in Oslo.
  - Several measures have been taken to lower construction costs and time, but more active use of region-specific housing policies and relaxation of excessively stringent local regulations should be considered.

### Corporate sector and CRE vulnerabilities: assessment and facts
- Corporate indebtedness:
  - Gross corporate debt is relatively high, but a significant portion is intercompany debt; net corporate debt in Norway is around 80 percent of GDP.
  - Corporate credit growth has slowed since the GFC and is now more in line with operating revenue; corporate loans have fallen as a share of total bank loans.
  - Market financing (corporate bonds) has risen while reliance on bank loans has reduced.
- Corporate performance:
  - Most corporates are in sound financial condition except for some oil-related industries.
  - Debt servicing capacity is high in most industries; bank losses on overall corporate loans have been low even after the GFC and the 2014 oil downturn.
  - Some highly-leveraged oil service industries (e.g., drilling and supply) show weakening debt servicing capacity.
- CRE sector specifics:
  - CRE accounts for the largest part of the corporate loan book and has grown from 25 percent of overall corporate bank loans in 2000 to 40 percent.
  - CRE prices rose rapidly in the last decade; yields fell to historic-low levels, particularly in Oslo prime office segment.
  - Banks’ large exposure to CRE loans and cyclicality in CRE pose risk of substantial loan losses with sharp price corrections.
  - Exposures vary significantly across banks by size and underlying risk (location, size, segment).
- Quantitative points on CRE and non-bank financing:
  - Among advanced European countries, Norway has one of the highest bank exposures to CRE at 16 percent in total lending.
  - Foreign intra-regional and cross-regional investments averaged 42 percent of CRE investments between 2006 and 2015; in Norway, 18 percent of CRE investments are foreign.
  - Non-bank financing via bond issuance increased from 2 percent in early 2000 to 8 percent in 2017.
  - Insurance companies’ exposure to CRE represents more than 5 percent of total assets in several countries; in Norway, financing to CRE accounts for 3 percent of insurance companies’ total portfolio.

### Prudential measures applied and recent actions
- Capital-based measures and oversight:
  - Norwegian banks adopted a risk weight floor of 100 percent on CRE exposures for SA banks in 2014 under the European framework.
  - Intensified oversight and Pillar II capital add-ons for banks with concentrated exposures were introduced in 2018.
  - A temporary risk weight floor of 35 percent for IRB banks was announced to become effective end-2020.
  - Countercyclical buffer (CCyB) increase from 2 to 2.5 percent effective end-2019 reflected CRE sector risks.
- Tax and corporate-debt policy changes:
  - Corporate income tax rate reduced from 28 percent in 2013 to 22 percent in 2019, mitigating corporate debt bias for firms with foreign shareholders.
  - Interest deduction limitation now applies to all debt instead of only related parties; the rule is 25 percent of taxable EBITDA if a threshold amount of MNOK 25 in net interest expenses is exceeded.

### Recommendations and policy avenues
- Broaden the macroprudential toolkit for CRE vulnerabilities; reliance solely on broad-based capital measures may be insufficient to contain price and credit booms.
- Specific avenues to explore:
  - Sectoral capital measures:
    - Target banks with higher exposures to CRE.
    - Example: temporary risk weight floors on IRB banks.
    - Consideration of regional/regional-risk calibrated risk weights (as in the UK) could be useful if justified.
    - Sectoral CCyB/SRB could be considered despite potential tensions with the European framework.
  - Borrower-based measures:
    - Consider if sectoral capital measures prove insufficient to contain CRE risks.
    - Borrower-based measures directly reduce excessive credit growth and valuations and increase borrower resilience.
    - Advantages: apply to flow of credit and can be implemented faster; can cover bank and non-bank domestic borrowers.
    - Challenges: calibration difficulties due to sector heterogeneity and data limitations; do not address foreign investor risks.
  - Other measures:
    - Microprudential measures (intensified oversight, Pillar II add-ons).
    - Tax measures (e.g., interest deduction limitation) to influence incentives.
- Box 4 synthesis on instruments:
  - Capital measures: raise risk weights (SA), set higher LGD values for CRE (Article 164(5) CRR), CCyB/SRB; limitations include leakages to non-bank and foreign exposures.
  - Borrower-side measures: LTV limits, DSCR/ICR limits, mortgage lending value requirements; directly target borrowers and cover bank and non-bank domestic borrowers but face calibration and heterogeneity challenges.
  - Tax and capital flow management measures: can disincentivize speculative short-term investments in some contexts; stamp duties effective in Hong Kong SAR and Singapore but constrained in EEA; interest deduction limits effective in reducing debt bias for highly leveraged CRE firms.

*Source: IMF Macroprudential Policy Survey.*

### Box 4. Use of Macroprudential Tools Geared Towards the CRE Sector (cont’d)

### Box 4. Use of Macroprudential Tools Geared Towards the CRE Sector (cont’d)

### Capital-based measures: implementation and examples
- A risk weight floor of 100 percent has been applied on CRE exposure among banks applying the standardized approach in many European countries (including Norway) in line with national discretion in the CRR (ESRB Macroprudential Policy Survey).
- Only a few countries apply risk weight floors on IRB banks.
- UK: slotting system with risk weights ranging from 50 to 250 percent based on the risk level for the IRB banks.
- Sweden and Norway are proposing to impose the risk weights on CRE exposures for IRB banks.
- Hong Kong SAR: differential risk weights for IRB banks depending on the property characteristics.
- SRB activations cited to partly address CRE-related vulnerabilities: Croatia and Hungary.
- CCyB activations cited to partly address CRE-related vulnerabilities: Norway, the Netherlands, Australia, and Ireland.

### Borrower-side measures: usage and limits
- Very few advanced economies have used borrower-side measures for CRE-related risks: Denmark, Hong Kong SAR, and Cyprus (based on IMF and ESRB surveys).
- Hong Kong SAR applies the tightest limits among listed examples:
  - LTV: 30-40 percent (values depending on size of loan and borrower characteristics: first-time, foreign or not).
  - DSTI limits: 20-30 percent (values depending on size of loan and borrower characteristics).
- Historical tightening in Hong Kong SAR since 2012: prior DSTI limits of 50-60 percent and LTV limits of 50 percent were tightened on a number of occasions.
- Most LTV limits in other countries range between 55-65 percent.
- Most DSTI limits in other countries range between 60-80 percent.

### Evidence on effectiveness and policy leakages
- Limited evidence on effectiveness of these measures in managing CRE cycles, given the short history of use and policy leakages (for example to international investors in Hong Kong SAR and Singapore).

### Operational challenges with borrower-side measures for CRE
- CRE valuation is more challenging than residential property valuation, requiring more due diligence and frequent updating.
  - Valuations can be conducted internally by banks and/or by external agencies with differing methodologies and wide-ranging results.
  - In Hong Kong SAR, if valuations are done by banks, period checks (preferred quarterly) are required by external agencies to ensure prudence.
  - If valuations are done by external agencies, banks must establish policies and procedures to ensure valuation reliability.
- Sectoral heterogeneity complicates calibration of limits; one universal limit is unlikely to be appropriate.
  - Despite heterogeneity, Hong Kong SAR and many other places applied no differential limits for different sectors.

### Other measures used to address CRE risks
- Denmark: limits on lending growth to each individual sector — 15 percent (mortgage banks) and 20 percent (deposit banks); banks’ CRE exposure limited to 25 percent of their total lending.
- Stamp duties introduced to limit speculative foreign investments in Singapore, Hong Kong SAR, Australia, Canada, and the UK.
- Other fiscal/tax measures mentioned:
  - Increased property taxation.
  - Reduced tax deductibility of interest payments.
  - Higher capital gains taxes.
  - Many countries, including Norway, have limits on interest deductions.
- Structural policy: reforming land and urban planning policies to improve elasticity of supply and contain CRE price growth (ESRB report 2018).

### Key characteristics of the Norwegian CRE market (Box 5 summary excerpts)
- CRE market heterogenous by business activities and property types.
- Three main business types: rentals and management, purchases and sales, and development of construction projects.
- Rental management companies account for about 83 percent of net debt and 86 percent of earnings.
- Office segment: around half of total value of transactions in the past ten years; retail segment: accounts for 1/3.
- Close to 60 percent of all offices (in square meters) built in Norway in the past decade were in Oslo.
- Investor base mostly domestic: as of 2018, foreign investors account for less than 20 percent of total investments.
  - Large foreign investor inflow in 2014/15 subsided afterwards; largest foreign investors from Finland, Sweden, the UK and the US.
- CRE financing structure:
  - Equity: one third of total funding to CRE companies.
  - Bank debt: one third.
  - Remainder: other debt and a small but rising share of bond debt.
  - Smaller CRE companies rely almost entirely on bank financing via SPVs; larger groups increasingly rely on bond financing and banks with loans backed by pools of properties (multiple SPVs).

### Banks’ exposure to CRE and heterogeneity across banks
- Banks’ share of loans to commercial property as share of all granted loans to nonfinancial firms varies from 15 to 93 per cent.
- Seven largest banks combined account for just over three-quarters of Norwegian banks’ loans granted to the property industries; their exposures vary from 26 to 68 percent.
- Handelsbanken has the largest relative CRE exposure among all the banks.
- Small banks have highest relative (but often low absolute) CRE exposure, lending mostly to hotels, retail, and logistics where prices are broadly stable.
- Larger banks more exposed to Oslo’s prime CRE market where valuations are most stretched.
- Banks also have indirect exposure to the CRE sector through lending to other sectors.

### Funding and liquidity vulnerabilities
- Banks rely heavily on wholesale funding; wholesale funding provides close to half of all funding for banks and mortgage companies and is a large share relative to deposits (at 40 percent).
- About two-thirds of wholesale funding is comprised of covered bonds collateralized by mortgages; half of covered bonds are foreign-currency denominated and swapped into domestic currency through the foreign exchange swap market.
- Unsecured senior bonds provide a third of wholesale funding, 40 percent of which are issued in foreign currency.
- Covered bonds are the largest category of outstanding HQLA as government bond issuance is relatively small.
- Banks invest in each other’s covered bonds to fulfill LCR requirements.
- Banks hold foreign currency deposits at central banks against short-term foreign currency funding.
- LCR and NSFR:
  - Banks and mortgage companies must maintain a minimum LCR of 100 percent for all currencies in total and for each significant currency.
  - Banks with EUR and/or USD as significant currencies must maintain a minimum LCR in NOK of 50 percent.
  - Banks fulfill LCR and the NSFR is monitored (NSFR still only monitored, not enforced).

### Risks from covered bond market and foreign currency liquidity
- Liquidity in the covered bond market could be reduced in a systemic event, compounding banks’ risks from real estate exposures:
  - Higher refinancing risks; constrained ability to post additional collateral to maintain bonds’ cover ratios; mark-downs of covered bonds could hurt solvency of banks, life insurance companies and pension funds.
- The LCR in significant currencies provides an important buffer against foreign currency liquidity risks and can be relaxed in periods of financial distress.
  - Three conditions for relaxation (application of existing principles; IMF, 2014; IMF-FSB-BIS, 2016):
    - (i) buffers are in place;
    - (ii) foreign currency liquidity pressures are generating financial stress;
    - (iii) relaxation is expected to relieve stress and thereby contribute to containing adverse procyclical dynamics.
  - Illustration: Croatia relaxed foreign currency liquidity asset requirements and released liquid assets amounting to over 14 percent of GDP (Box 4 in IMF, 2017).

### Recommendations (selected)
- Clarify that the LCR in significant currencies is a buffer the authorities could consider relaxing in periods of foreign currency liquidity stress.
  - Rationale: a relaxation of macroprudential policies can help counter financial stresses from foreign currency liquidity pressures and maintain provision of financial services to the real economy.
  - Caution: liquidity pressures do not mechanically call for relaxation; careful judgement required and the three conditions above should be satisfied before relaxation.

### Structural vulnerabilities and systemic institutions
- Financial system size: about 270 percent of GDP; concentrated in banking.
- Banks represent over half of the financial system (in assets); largest commercial bank, DNB, constitutes over 40 percent of the banking sector.
- Bank exposures highly concentrated in real estate: loans to residential and commercial real estate sectors comprise over two-thirds of private sector loans.
- Direct interconnectedness through cross-holdings of covered bonds is extensive: banks and mortgage companies hold over half of issued covered bonds in NOK.
- Systemic buffers:
  - High SRB rate is explicitly mitigating these risks; stress test results suggest buffers are sufficiently high.
  - Authorities recently announced the buffer rate will be increased.
- Identification of systematically important institutions:
  - MoF has identified DNB and Kommunalbanken as systematically important.
  - FSA criteria for identification: (i) total assets exceeding 10 percent of GDP; (ii) lending exceeding 5 percent of domestic loans (including to municipalities); consideration of EBA guidelines.
  - In October 2018, FSA proposed an additional criterion of at least 10 percent market share in corporate lending in one or more regions; MoF sent proposal to public hearing but decided not to follow FSA’s advice.
  - Systemically important institutions (DNB and Kommunalbanken) are subject to a 2 percent O-SII buffer and an additional leverage ratio requirement of 1 percent.

*Source: Box 4 (concluded) and Box 5 excerpts from the provided IMF Norway FSAP content.*

### Appendix II . Macroprudential Policy Measures, 2010–19

### Appendix II . Macroprudential Policy Measures, 2010–19

### Time Descriptions — Chronology of measures (2010–2019)
- 2010: Guidelines on Mortgage Lending introduced
  - LTV limit on mortgages at 90 percent
  - LTV limit on home equity credit lines at 75 percent
  - Amortization required if LTV exceeds 70 percent
  - 5 percent interest rate stress test
- 2011: Guidelines on Mortgage Lending
  - LTV limit mortgages lowered to 85 percent
  - LTV limit on home equity credit lines lowered to 70 percent
- 2013: Systemic Risk Buffer (SRB)
  - Introduced for the first time and set at 2 percent
- 2014:
  - SRB: Increased to 3 percent
  - Risk weight for IRB banks: LGD floor increased from 10 to 20 percent on mortgage exposure
- 2015:
  - CCyB: Introduced for the first time and set at 1 percent
  - Other systemically important banks buffer (OSII-B): Introduced for the first time and set at 1 percent
  - Guidelines on mortgage lending changed to regulation:
    - LTV limit at 85 percent on mortgages
    - LTV limit at 70 percent on home equity credit lines
    - Amortization of 2.5 percent annually required if LTV exceeds 70 percent
    - 5 percent interest rate stress test
    - Speed limit of 10 percent
- 2016:
  - CCyB: Increased to 1.5 percent
  - OSII-B: Increased to 2 percent
  - Liquidity coverage ratio (LCR) all currencies: Introduced and set at 100 percent
- 2017:
  - CCyB: Increased to 2 percent
  - Guidelines on Mortgage Lending changed to Regulation:
    - LTV limit on mortgages kept at 85 percent
    - LTV limit on equity credit lines lowered to 60 percent
    - Amortization of 2.5 percent annually required if LTV exceeds 60 percent (lowered from 70 percent)
    - DTI limit introduced and set at 5 times gross annual income
    - LTV limit on secondary homes in Oslo introduced at 60 percent
    - Speed limit lowered to 8 percent in Oslo and kept at 10 percent for other regions
  - Guidelines on Consumer Credit:
    - DTI limit of 5 times gross annual income
    - 5 percent interest rate stress test
    - Amortization requirement of 5 years
  - LCR Requirements:
    - For all currencies, increased to 80 percent
    - For significant currencies, introduced and set at 50 percent
- 2018:
  - Mortgage Regulations renewed with no changes
  - LCR Requirement: For all currencies, increased to 100 percent
- 2019:
  - Guidelines on consumer credit changed to regulation
  - CCyB: Increased buffer to 2.5 percent

### Mortgage lending measures and limits
- LTV limits timeline:
  - Mortgages: 90 percent (2010) → 85 percent (2011, maintained in 2015 and 2017)
  - Home equity credit lines: 75 percent (2010) → 70 percent (2011, maintained in 2015) → 60 percent (2017)
  - Secondary homes in Oslo: LTV limit introduced at 60 percent (2017)
- Amortization requirements:
  - 2010: Amortization required if LTV exceeds 70 percent (unspecified rate)
  - 2015 regulation: Amortization of 2.5 percent annually required if LTV exceeds 70 percent
  - 2017 regulation: Amortization of 2.5 percent annually required if LTV exceeds 60 percent (lowered from 70 percent)
  - Consumer credit amortization requirement: 5 years (2017 guidelines)
- Stress testing:
  - 5 percent interest rate stress test specified in 2010, reiterated in 2015, and applied to consumer credit in 2017

### Debt service and borrower limits
- DTI (Debt-to-Income) measures:
  - 2017: DTI limit introduced and set at 5 times gross annual income (mortgage lending)
  - 2017 Guidelines on Consumer Credit: DTI limit of 5 times gross annual income
- Speed limits (supply/credit growth constraints):
  - 2015 regulation: Speed limit of 10 percent
  - 2017: Speed limit lowered to 8 percent in Oslo and kept at 10 percent for other regions

### Capital buffers and risk-weight changes
- Systemic Risk Buffer (SRB):
  - 2013: Introduced and set at 2 percent
  - 2014: Increased to 3 percent
- Countercyclical Capital Buffer (CCyB):
  - 2015: Introduced and set at 1 percent
  - 2016: Increased to 1.5 percent
  - 2017: Increased to 2 percent
  - 2019: Increased to 2.5 percent
- Other systemically important institutions buffer (OSII-B):
  - 2015: Introduced and set at 1 percent
  - 2016: Increased to 2 percent
- Risk-weight for IRB banks:
  - 2014: LGD floor increased from 10 to 20 percent on mortgage exposure

### Liquidity coverage ratio (LCR) requirements
- 2016: LCR (all currencies) introduced and set at 100 percent
- 2017 adjustments:
  - For all currencies, increased to 80 percent
  - For significant currencies, introduced and set at 50 percent
- 2018: LCR Requirement for all currencies increased to 100 percent

*Source: Appendix II . Macroprudential Policy Measures, 2010–19*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1norea2020007.pdf_