## 1norea2020008 ## Source details **Canonical URL:** [1norea2020008](https://www.imf.org/-/media/files/publications/cr/2020/english/1norea2020008.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2020/english/1norea2020008.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2020/english/1norea2020008.pdf.json) --- ### Executive summary — context and system structure - Much of the FSAP work was conducted prior to the COVID-19 outbreak; the risk and vulnerability analysis integrates original work with quantification of COVID-19 impacts on bank solvency under two separate scenarios. The original “market shock” scenario explores additional risks that feature less prominently in the COVID scenarios. - Financial sector size and composition: - Financial sector assets, excluding the Government Pension Fund Global (GPF-G), total 290 percent of GDP. - The GPF-G represents another 230 percent of GDP. - Banks represent 54 percent of the financial system (in terms of total assets). - Mortgage companies represent 23 percent of the financial system (in terms of total assets). - At the end of 2018 the banking sector comprised 141 domestic and foreign commercial banks. - Monetary policy context: - Norges Bank had raised its key policy rate four times between September 2018 and September 2019, to 1.5 percent, and during March–May 2020 it cut rates to zero percent in three quick steps. - Lending and balance-sheet features: - Approximately two-thirds of banks’ assets are loans; financial instruments represent 18 percent of total assets. - Close to 60 percent of banks’ lending is property-related (residential and commercial real estate and construction). - Over 90 percent of all Norwegian residential mortgages are variable rate loans. ### Key solvency stress-test findings and comparisons - Overall resilience: - Stress tests suggest banks entered the COVID-19 crisis well-prepared, but current uncertainty calls for vigilance. - The solvency stress tests show resilience of the banking sector under all scenarios. - Severe scenario outcomes (Market Shock): - In the most severe scenario the CET1 capital ratio would decline by around 5 percentage points on average. - Average CET1 ratio for the 11 banks in scope drops to 12.4 percent by end of the risk horizon, from 17.4 at end-2019 — an average 5 percentage points fall. - Decline in average total capital ratio is close to 6.2 percentage points. - Increase in risk-weighted assets contributes to capital ratio decline by 2.3, 1.8, and 1.3 percent in the years 2020 to 2022, respectively. - Worst drops in CET1 ratios across banks range from 1.1 to 6.8 percent. - No material breaches of hurdle rates: buffers in excess of hurdle rates partially depleted for most banks and fully for three of them. - COVID scenarios outcomes: - Aggregate CET1 ratio for in-sample banks drops by about 4 percent under the central scenario and 4½ percent under the downside scenario by the end of the risk horizon. - Under the COVID downside scenario, one bank would exhaust its buffers in excess of the hurdle rate, though without breaching it. - COVID scenarios show less severe impact overall than the market shock scenario over the full horizon; COVID shock is frontloaded causing comparable or worse impact in the first year, but faster recovery and easier financial conditions mitigate subsequent impact. - Cross-exercise comparisons: - Norges Bank and Finanstilsynet ran top-down stress tests in parallel with IMF staff under the market shock scenario with broadly similar results; Norges Bank’s simulations produced larger declines in capital ratios. - FSA top-down: average CET1 reduced from 17.4 percent to 12.3 percent; loan losses sum to a total of NOK 185 bn; combined net profits across all banks accumulate to NOK -93 bn over 3 years. - Norges Bank top-down: estimates a larger average drop in CET1 ratio, in the order of -6.2 percentage points for the comparable sample of nine banks. - Three largest domestic banks’ bottom-up tests produced more moderate losses and declines in capital ratios than the top-down exercises; bottom-up tests show all measures of capital adequacy between 3 and 4 percentage points lower under the adverse scenario compared to baseline, with no hurdle breaches. ### Liquidity risk, funding vulnerabilities, and stressed LCR / cash-flow analysis - Wholesale funding and covered bonds: - Norwegian banks obtain nearly half their overall funding from wholesale markets and foreign exchange (FX) issuances. - About two-thirds of wholesale funding comes from covered bonds, split evenly between domestic and foreign currency issuances. - Covered bonds have partially substituted for senior-unsecured and short-term wholesale funding. - Short-term liquidity: - Stressed Liquidity Coverage Ratio (LCR) simulations point to a solid short-term liquidity situation: the average LCR across the 11 banks remained above the regulatory threshold even under extreme scenarios. - On average across the 11 banks, in all scenario combinations the average LCR would remain above 100 percent over a one-month horizon, though some banks would breach the threshold under the most severe combinations. - LCR in domestic currency (50 percent threshold imposed on some banks) remained above that threshold on average, with only one bank experiencing difficulties under some severe scenario combinations. - Norwegian banks are very liquid in EUR and USD; LCR in these currencies remain well above 100 percent in all cases. - Medium- to longer-term liquidity: - Cash-flow analysis beyond the 1-month LCR horizon reveals potential counterbalancing capacity (CBC) gaps for the whole system under severely adverse circumstances. - Under a severely adverse scenario, some banks would encounter difficulties beyond the 1-month horizon; the system would experience negative CBC gaps starting from the 3–4 months bucket. - Initial CBC composition (June 2019) for ten among the largest domestically incorporated banks: CBC comprised for more than half of central bank reserves; more than ⅓ of CBC is represented by covered bonds; government bonds make a modest share of overall CBC. - Funding- and derivative-related vulnerabilities: - Banks rely heavily on international capital markets for funding and are exposed to rollover risks in market turbulence. - Structural interest rate and currency transformation leads to systematic hedging in derivatives markets; this contains direct interest rate and exchange rate risks but increases exposure to margin calls and counterparty risk under extreme market shocks. - Capturing margining and derivative-related liquidity and counterparty risks requires granular transaction-level data not yet available. - Stressed LCR design and calibration: - LCR stress-testing used a 30-calendar-day stress period with augmented run-off rates and haircuts; simulation design combined three "haircut" scenarios with seven "outflows" scenarios (including retail funding shock, wholesale funding shock, or both). - Annex II reports stressed LCR outflow rates under four scenarios (Regulatory, Retail, Wholesale, Combined) with specific percent rates across many deposit and counterparty categories (e.g., deposits where the payout has been agreed within the following 30 days: 100 / 100 / 100 / 100). ### Interconnectedness, contagion, and market-based linkages - Market structure and cross-holdings: - Interbank deposit and repo markets are not particularly developed in Norway. - Banks are closely interlinked through derivative exposures and cross-holding of securities—especially covered bonds, which represent a larger pool of highly liquid assets than government bonds in Norway. - Substantial cross-ownership of covered bonds between banks adds to real estate exposure on banks’ asset sides. - Contagion analysis: - Norges Bank’s assessment of direct and indirect contagion effects found that contagion may lead to declines in banks’ capital, but such declines are small on average and large effects are unlikely. - Quantitative contagion outcomes: losses due to contagion are on average 1 pp of the banking sector’s CET1 ratio, but in some cases can reach 2.5 pp; losses result almost exclusively from indirect contagion. - Market-based interconnectedness: - Diebold and Yilmaz (2014) methodology on selected Scandinavian banks found two clusters: major Scandinavian banks (including DNB) and medium-sized Norwegian savings banks; index of total interconnectedness varied considerably over time, with a decline since 2017 and a rebound from a low in 2019. - Data limitations: - Availability of detailed derivative exposure data would materially improve analysis of interconnectedness and contagion using multi-layer network approaches. - Statistical recording issues complicate domestic vs. foreign holder identification of covered bonds (e.g., pension funds purchasing Norwegian covered bonds issued abroad recorded as “foreign” funding of mortgage companies). ### Climate-related transition risk analysis and scenarios - Scope and channels: - Mission explored propagation channels for transition risk from an abrupt increase in carbon pricing and accompanying measures that curb global fossil fuel supply and demand. - Two propagation channels explored: - Domestic cost of emissions: impact on firms’ earnings from a higher domestic carbon price and implications for their lenders. - External demand: impact on the Norwegian economy and banks of a fall in oil sector revenues caused by a higher global carbon price. - Carbon-price scenario calibrations and key figures: - Average price paid per ton of CO2-equivalents in 2018: NOK 266 (about US$30). - IMF-estimated range necessary to achieve Paris Agreement targets: US$50–100. - Two hypothetical uniform carbon tax increases analyzed: US$75 and US$150 per ton CO2-equivalent; also parallel shifts preserving sectoral differentiation to average US$75 or US$150. - Corporate cost of emissions channel — methodology and results: - Firm-level balance-sheet approach using Orbis firm data and Exiobase input-output sector emissions; "no-pass-through" assumption (firms fully absorb additional costs). - Debt at risk defined as share of firms for which the interest coverage ratio (ICR) drops below threshold (thresholds calibrated at 2 or 1). - Under a uniform carbon price hike to US$150, about 4 percent of all corporate bank exposures see their ICR drop below 2. - The most affected bank in the sample sees around 16 percent of corporate exposures drop below the threshold. - Sectors most materially impacted under uniform US$75: agriculture, forestry and fishing; water supply, sewerage, waste management and remediation activities; transportation and storage. - External demand channel — oil sector effects and loan-loss estimates: - Oil sector makes up roughly 16 percent of GDP, about one fifth of total investments and state revenues, and over one third of total exports. - A US$75 carbon tax is estimated to lead to a reduction in Norwegian oil revenues of about 22 percent. - A US$150 carbon tax is estimated to lead to a reduction in Norwegian oil revenues by more than 36 percent. - SVAR model estimates: - A carbon price of US$75 is estimated to increase loan loss rates by about 0.3 percentage points. - A carbon price of US$150 is estimated to increase loan loss rates by roughly 0.4 percentage points. - Historical benchmarks: average loss rate stood at 0.21 percent in 2018 for the whole banking system; average loss rate was 0.6 percent in 2016 during the 2014–16 oil price drop. - Portfolio/valuation effects from enforced reductions in oil production: - Carbon Tracker Initiative estimate: companies such as Equinor have to reduce their overall output by about 45 percent before 2040 to be consistent with Paris targets. - A 45 percent reduction of output could see an oil producer’s income fall by up to 80 percent, depending on global oil prices and firm characteristics. - Imposition of a carbon price of US$75–150 could further decrease output by approximately 8–12 percent. - Fall in share price for a company like Equinor estimated to reach up to -50 percent under some assumptions. - Spillover to household and institutional portfolios: a modeled 45 percent reduction in oil production lowering equity values would lower households’ financial assets by about 12.5 percent (assuming equity holdings representative of the Oslo Børs All-share Index); insurers and pension funds would see portfolio drops in the range of 4 to 5 percent; non-money-market investment funds potential drops between 5 and 6 percent. - Caveats: - Analysis is static, single-factor, and partial-equilibrium; results are preliminary and conditional on numerous assumptions; further work required to capture general-equilibrium effects and side-effects. ### Key risks and vulnerabilities (structural and cyclical) - Real estate and household risk: - Residential real estate (RRE) prices increased by 70 percent over the last decade, rising more in larger cities (particularly Oslo, where they doubled). - Household debt levels now exceed 200 percent of disposable income on average. - Share of households with debt levels exceeding five times their gross income has been on an upward trajectory. - NPLs were below one percent overall, though NPLs on consumer loans (five percent of bank lending) were much higher at nine percent. - Banks’ provisions ensured about 85 percent coverage of NPLs. - Funding and concentration risk: - High reliance on wholesale funding and covered bonds increases rollover risk and concentration (reliance on international capital markets, EUR and USD issuances). - Mortgage companies issue covered bonds on true-sale basis and must maintain overcollateralization of at least two percent; many operate well above that threshold. - Other risks: - Climate transition risk given Norway’s reliance on oil production/export. - Cybersecurity risks to the mostly cashless payments system could cause severe dislocations and deposit contraction. - Financial integrity risks (e.g., alleged breach of customer due diligence rules at DNB) could lead to loss of confidence, financial losses, and sanctions. - Consumer lending: - Consumer lending is growing considerably from a modest base and is an increasing source of concern. - The FSA in 2018 found that 470,000 consumer loans became the subject of debt collection procedures, for a total of NOK 21bn (approx. 18 percent of total consumer loans for the surveyed banks), a figure 18 percent higher than in 2017. ### Sensitivity, market, IRRBB and concentration tests - IRRBB: - Analyzed using Basel Committee standard approach: Economic Value of Equity (EVE) and Net Interest Income (NII). - Six interest rate shock scenarios calibrated by FSA: parallel shock 200 bps; short-term rate shock 300 bps; long-term rate shock 150 bps. - Results: EVE reported by banks at end-2019 are all well below the 15 percent of Tier 1 capital Basel threshold for "outlier banks"; impact on NII always below 3 percent of Tier 1 capital. - Market risk: - Equity holdings represent between 0 and 1 percent of banks’ assets for banks in scope. - Historical VaR (1 percent, 1-month) on OSEAX: reduction in capital ratios from equity portfolio losses would not exceed 25 bps. - Government bond simulation (1 percent, 1-month) impact no more than 15 bps; corporate bond simulation no more than 5 bps. - FX risk and concentration: - Banks’ bottom-up FX risk historical simulations: FX risk exposures do not reach 1 percent even at the longest horizons. - Reverse stress test for concentration: simultaneous default of no less than the top 10 counterparts would be required to exhaust the banks’ surplus Tier 1 capital — indicates a low degree of concentration. ### Financial-sector evolution and interconnectedness (1998–2018) - Covered bonds law introduced in 2007 reshaped financial linkages and increased prominence of mortgage companies in the financial network. - Net financial exposures over time show mortgage companies became important in financing households by 2018. - Statistics recording issues: purchases of Norwegian covered bonds issued abroad by domestic pension funds and insurers complicate domestic vs. foreign holder identification. - Interbank network analyses and multi-layer network recommendations emphasize need for trade repository data for derivatives and multi-layer contagion models. ### Key policy recommendations and data / methodological priorities - Data collection and reporting: - Maintain accurate, updated, and easily retrievable maps of the major corporate groups — Finanstilsynet (FSA); I (Immediate). - Collect granular data for defaults on bank loans, including on residential and commercial mortgages — Norges Bank, FSA; ST (Short term). - Continue to collect and manage data from trade repositories and build models to analyze risks and contagion effects from the network of derivative exposures across banks — Norges Bank, FSA; I. - Collect from banks assets and liabilities data at a consolidated level, with breakdown by maturity bucket for time of repricing or duration and with the hedging positions — FSA; MT (Medium Term). - Collect data on the liquidity positions of foreign bank branches — FSA; MT. - Methodology and modelling: - Build more analytical and granular satellite models for credit risk, to link the macro landscape with relevant variables (PDs, LGDs, etc.) at bank and asset class level — Norges Bank, FSA; ST. - Deepen the analysis of margining arrangements adopted by banks and mortgage companies for derivative transactions and of related counterparty and liquidity risks — Norges Bank, FSA; I. - Implement stressed LCR and cash-flow analysis based on EU common reporting as useful complementary diagnostic tools for liquidity risk analysis — FSA; MT. - Develop an IFRS 9-compatible top-down credit loss model to better capture the loan loss impact under hypothetical downturn scenarios in top-down solvency stress tests. - Operational and supervisory steps: - Proceed swiftly with plans to utilize data collected from trade repositories and build models to analyze liquidity risks arising from derivative exposures (in particular liquidity drains caused by margin calls in turbulent markets). - Establish a regular collection of data on the liquidity position of branches of foreign banks. - Deepen analysis of margining arrangements (centrally cleared vs bilateral, symmetric vs asymmetric margining) and implications for liquidity, counterparty risk, CVA and capital requirements. - Adjust or replace the two indicators for the non-bank sector in Norges Bank’s heatmap to avoid persistent 'false alarms'. - Timing legend: I Immediate (within 1 year); ST Short term (within 1–2 years); MT Medium Term (within 3–5 years). *Source: IMF staff’s Technical Note (Norway FSAP), “EXECUTIVE SUMMARY”.* ### EXECUTIVE SUMMARY __________________________________________________________________________ 6 ### EXECUTIVE SUMMARY ### Context and system structure - Much of the FSAP work was conducted prior to the COVID-19 outbreak; the risk and vulnerability analysis integrates original work with quantification of COVID-19 impacts on bank solvency under two separate scenarios. The original “market shock” scenario explores additional risks that feature less prominently in the COVID scenarios. - Financial system size and composition: - Financial sector assets, excluding the Government Pension Fund Global (GPF-G), total 290 percent of GDP. - The GPF-G represents another 230 percent of GDP. - Banks represent 54 percent of the financial system (in terms of total assets). - Mortgage companies represent 23 percent of the financial system (in terms of total assets). - At the end of 2018 the banking sector comprised 141 domestic and foreign commercial banks. - Monetary policy context: - Norges Bank had raised its key policy rate four times between September 2018 and September 2019, to 1.5 percent, and during March–May 2020 it cut rates to zero percent in three quick steps. - Lending and balance-sheet features: - Approximately two-thirds of banks’ assets are loans; financial instruments represent 18 percent of total assets. - Close to 60 percent of banks’ lending is property-related (residential and commercial real estate and construction). - Over 90 percent of all Norwegian residential mortgages are variable rate loans. ### Key findings from solvency stress tests - Overall resilience: - Stress tests suggest banks entered the COVID-19 crisis well-prepared, but current uncertainty calls for vigilance. - The solvency stress tests show resilience of the banking sector under all scenarios. - Severe scenario outcomes: - In the most severe scenario the CET1 capital ratio would decline by around 5 percentage points on average. - Three banks would almost or completely exhaust their capital buffers above the conservative hurdle rates of about 10 percent CET1 ratio. - There would be no material breach of the remaining buffers and requirements, confirming the overall adequacy of current capital levels. - Comparisons of exercises: - Norges Bank and Finanstilsynet ran top-down stress tests in parallel with IMF staff under the market shock scenario with broadly similar results; Norges Bank’s simulations produced larger declines in capital ratios. - The three largest domestic banks ran bottom-up solvency stress tests under the same scenario; their results are more moderate in terms of losses and declines in capital ratios than the top-down exercises. ### Liquidity risk and funding vulnerabilities - Short-term liquidity: - Stressed Liquidity Coverage Ratio (LCR) simulations point to a solid short-term liquidity situation: the average LCR across the 11 banks remained above the regulatory threshold even under extreme scenarios. - Medium- to longer-term liquidity: - Cash-flow analysis beyond the 1-month LCR horizon reveals potential counterbalancing capacity (CBC) gaps for the whole system under severely adverse circumstances. - The potential gaps stem particularly from assumptions of severe limitations in roll-over of maturing debt issuances and derivative transactions. - Funding structure risk: - Banks rely heavily on international capital markets for funding and are exposed to rollover risks in market turbulence. - The structural interest rate and currency transformation in bank business models leads to systematic hedging in derivatives markets; this contains direct interest rate and exchange rate risks but increases exposure to margin calls and counterparty risk under extreme market shocks. - Data needs: - Capturing margining and derivative-related liquidity and counterparty risks requires granular data on derivative transactions that is not yet available, though authorities are investing in this direction. ### Interconnectedness and contagion - Market structure: - Interbank deposit and repo markets are not particularly developed in Norway. - Banks are closely interlinked through derivative exposures and cross-holding of securities—especially covered bonds, which represent a larger pool of highly liquid assets than government bonds in Norway. - Analysis and findings: - Availability of detailed derivative exposure data would materially improve analysis of interconnectedness and contagion using multi-layer network approaches. - Norges Bank’s assessment of direct and indirect contagion effects found that contagion may lead to declines in banks’ capital, but such declines are small on average and large effects are unlikely. ### Climate-related transition risk analysis - Scope of analysis: - The mission explored propagation channels for transition risk from an abrupt increase in carbon pricing, and accompanying measures that curb global fossil fuel supply and demand. - Two propagation channels were explored: - Domestic cost of emissions: impact on firms’ earnings from a higher domestic carbon price and implications for their lenders. - External demand: impact on the Norwegian economy and banks of a fall in oil sector revenues caused by a higher global carbon price. - Results: - Within the limits of a static, single-factor, partial equilibrium analysis, estimated impacts on banks and the financial system are relatively contained but non-negligible. - A tentative portfolio channel analysis emphasized valuation effects of necessary reductions in oil production on oil-related companies and the financial wealth of households and other sectors. - Policy implication: - The results underline the importance for financial institutions and micro- and macroprudential supervisors to adequately identify, measure, and manage climate transition risks. ### Key recommendations on vulnerability analysis and stress testing (excerpt from Table 1) - Data (Agency; Time) - Maintain accurate, updated, and easily retrievable maps of the major corporate groups — Finanstilsynet (FSA); I - Collect granular data for defaults on bank loans, including on residential and commercial mortgages — Norges Bank, FSA; ST - Continue to collect and manage data from trade repositories and build models to analyze risks and contagion effects from the network of derivative exposures across banks — Norges Bank, FSA; I - Collect from banks assets and liabilities data at a consolidated level, with breakdown by maturity bucket for time of repricing or duration and with the hedging positions — FSA; MT - Collect data on the liquidity positions of foreign bank branches — FSA; MT - Methodology (Agency; Time) - Build more analytical and granular satellite models for credit risk, to link the macro landscape with relevant variables (PDs, LGDs, etc.) at bank and asset class level — Norges Bank, FSA; ST - Deepen the analysis of margining arrangements adopted by banks and mortgage companies for derivative transactions and of related counterparty and liquidity risks — Norges Bank, FSA; I - Implement stressed LCR and cash-flow analysis based on EU common reporting as useful complementary diagnostic tools for liquidity risk analysis — FSA; MT - Timing legend: I Immediate (within 1 year); ST Short term (within 1-2 years); MT Medium Term (within 3-5 years) *Source: IMF staff’s Technical Note (Norway FSAP), “EXECUTIVE SUMMARY”.* ### 8. Wholesale markets are an important source of funding. As credit demand structurally ### 1norea2020008 - 8. Wholesale markets are an important source of funding. As credit demand structurally ### Wholesale funding and covered bonds - Norwegian banks obtain nearly half their overall funding from wholesale markets and foreign exchange (FX) issuances. - About two-thirds of wholesale funding comes from covered bonds, split evenly between domestic and foreign currency issuances. - Covered bond issuances have partially substituted for other riskier sources of wholesale funding, such as senior-unsecured and short-term wholesale funding. - There is substantial cross-ownership of covered bonds between banks, as they hold these as HQLA, which further adds to real estate exposure on banks’ asset sides. - Domestic market insufficiency leads banking groups regularly to tap global capital markets with foreign currency (mainly EUR and USD) and generally fixed-rate issuances. ### Capitalization, liquidity, profitability, and asset quality (pre-COVID-19 and entering COVID-19) - Banks’ total regulatory capital ratio was 24.2 percent as of end-2019. - CET1 capital ratio was 18.0 percent as of end-2019; local regulatory requirements are consistent with Basel III standards. - Two domestic systemically important institutions face an additional two percent requirement. - Banks’ liquidity levels were in full compliance with the LCR requirements (which follow the EU framework). - Liquidity coverage in foreign currencies generally exceeded that in Norwegian krone owing to a shortage of domestic HQLA. - Bank profitability was strong in peer comparison, owing to low operating expenses (partly due to high digitalization) and low credit losses. - Asset quality: NPLs were below one percent overall, though NPLs on consumer loans (five percent of bank lending) were much higher at nine percent. - Banks’ provisions ensured about 85 percent coverage of NPLs. ### Mortgage companies and covered bond issuance structure - By law, covered bonds can only be issued by separate and dedicated entities within a banking group (covered bond mortgage companies). - These entities acquire mortgage loans from their parent banks on a 'true sale' basis. - Covered bond mortgage companies are subject to the same capitalization requirements as banks and must maintain overcollateralization of at least two percent, with many operating well above that threshold. - Some mortgage companies are not authorized to issue covered bonds and instead issue ordinary unsecured senior bonds; these are mostly small, with notable exceptions Kommunalbanken and Eksportfinans. - Kommunalbanken is wholly owned by the government, issues almost all its long-term debt abroad, finances lending to local government, and has been declared a Domestic Systemically Important Bank (D-SIB). ### Key risks and vulnerabilities - Underlying vulnerabilities pertain to banks’ high exposure to domestic real estate (residential and commercial) and reliance on wholesale funding. - A deterioration in the ability of highly-leveraged households or corporates to service loans, a sharp real estate price correction, or both, would affect banks’ asset quality; banks’ high reliance on wholesale funding, including covered bonds backed by residential mortgages, compounds the risks. - Residential real estate (RRE) prices increased by 70 percent over the last decade, rising more in larger cities (particularly Oslo, where they doubled). - Household debt levels now exceed 200 percent of disposable income on average. - The share of households with debt levels exceeding five times their gross income has been on an upward trajectory. - Commercial real estate (CRE) has also boomed, especially Oslo’s prime office market. - Corporate debt levels are about average in international peer comparison. - Prior to COVID-19, macroprudential and monetary tightening helped cool the residential market while CRE prices continued to rise. - Countercyclical capital buffer (CCyB) planned upper level was 2½ percent from end-2019. - Introduction of temporary household sector tools included a stressed-interest rate debt servicing test, LTV and DTI caps, and amortization requirements. - Survey example: interest-only loans with LTV ratio above 60 percent fell from almost 9 to below 3 percent between 2016 and 2018. - Consumer lending is growing considerably from a modest base and is an increasing source of concern. - The FSA in 2018 found that 470,000 consumer loans became the subject of debt collection procedures, for a total of NOK 21bn (approx. 18 percent of total consumer loans for the surveyed banks), a figure 18 percent higher than in 2017. - Banks’ typical balance-sheet shows assets prevalently floating rate and NOK-denominated, financed by a mix of fixed/floating rate and NOK/FX funding, exposing banks to: - losses from decreasing interest rates and depreciating NOK; - hedging positions that are fix-receiving on interest rates and long on FX, leading to exposure to margin calls when interest rates increase and/or NOK appreciates, and to increased counterparty risk when interest rates decrease and/or NOK depreciates. - Additional structural risks: - climate transition risk given Norway’s reliance on oil production/export; - cybersecurity risks to the mostly cashless payments system, where a cyber-attack on a critical payment infrastructure could cause severe dislocations and deposit contraction; - financial integrity risks (e.g., alleged breach of customer due diligence rules at DNB) that could lead to loss of confidence, financial losses, and sanctions. ### Risks linked to reliance on international capital markets (from hedging and funding strategy) - Liquidity risk: - rollover risk (inability to execute planned issuances or high cost of alternatives); - potential liquidity drains from margin calls (variation margin) and/or requests for additional initial margin during increased market volatility. - Counterparty risk: - favorable market price changes can lead to credit losses if counterpart defaults, or to increased credit valuation adjustments (CVA) and related capital requirements. - Concentration risk: - reliance on a single or few—possibly connected—counterparts for bilateral transactions and/or direct or indirect clearing members for cleared transactions. ### Solvency stress tests — scenarios, assumptions, and methodology - Three scenarios were analyzed over a three-year horizon: ‘COVID Central’, ‘COVID Downside’, and ‘Market Shock’. - The FSA and Norges Bank ran parallel top-down solvency stress tests based on the Market Shock scenario. - Risk horizon for the exercise: from end-2019 to end-2022. - COVID scenarios incorporate effects of measures already taken by the authorities to ease financial conditions. - The exercise used a quasi-static allocation balance sheet assumption (asset allocation and funding composition remain the same throughout the risk horizon). - Interest income derived from evolution of interest-bearing assets and liabilities and applied interest rates; non-interest income evolves in line with asset growth. - Dividends assumed paid out only in case of positive income, at a payout ratio consistent with Norway’s historical experience. - Resilience measured in terms of capital ratios against current and future requirements and some capital buffers; hurdle rates include all capital requirements and buffers except the Capital Conservation Buffer (CCB) and Countercyclical Buffer (CCyB). - The D-SIB buffer is treated as structural and assumed unusable in the scenario. - Adjustments made to RWAs and hurdle rates due to the EU capital framework (CRR/CRD IV) entering into full force from end-2019, including: - introduction of the SME ‘supporting factor’ and removal of the Basel I floor for IRB banks, reducing RWAs; - authorities increased the systemic risk buffer (SRB) from 3 percent to 4.5 percent, with the whole SRB of 4.5 percent applied only to domestic exposures (to be implemented end-2020 for IRB banks and end-2022 for non-IRB banks). - Top-down exercises abstract from the three-stage IFRS 9 classification and adopt a traditional two-stage approach (performing and non-performing loans), relying on banks’ PDs and LGDs for IRB exposures or NPLs for standardized approach exposures. ### Financial sector policy response to COVID-19 (summarized) - Monetary policy actions: - three rate cuts brought the main policy rate to zero percent (from 1.5 percent). - Norges Bank provided extraordinary NOK loans with maturities ranging from 1-week to 1-year with full allotment; collateral requirements eased by removing limits on the use of non-government securities. - The FSA permitted use by banks of HQLA held to satisfy the LCR requirement, provided it is properly reported. - Norges Bank agreed a U.S. dollar swap line with the Federal Reserve for up to US$30 billion and provided U.S. dollar liquidity to Norwegian banks. - Macroprudential measures: - relaxation of the countercyclical capital buffer (CCyB) from 2.5 percent to 1 percent to ease constraints on bank lending. - authorities indicated no increase in the CCyB anticipated until at least the first quarter of 2022. - temporary relaxation allowing banks to deviate from LTV, DTI, and other requirements for up to 20 percent of new loans during 2020:Q2 and Q3 (compared to previous “speed limit” of 10 percent; 8 percent in Oslo). - Microprudential actions: - appeals by the FSA and MoF for banks and insurers to restrict dividend payouts until economic uncertainty is reduced. - enhanced regulatory reporting of short sales of domestic equity shares. *https://www.imf.org/-/media/files/publications/cr/2020/english/1norea2020008.pdf* ### 27. The following scenarios were simulated over a 3-year horizon: (Table 3 and Figure 6): ### 27. The following scenarios were simulated over a 3-year horizon ### Scenario descriptions - COVID central scenario - Reflects the projected baseline outlook—for Norway and the global economy—as of June 2020. - Norwegian mainland GDP contracts by almost 5½ percent in 2020. - The economy starts to rebound from the second half of 2020. - COVID downside scenario - Corresponds to persistent uncertainty and a further deepening of the downturn. - Approximated by a downward divergence of a 1 standard deviation of the core variables (GDP, employment, oil price) from the central path. - Results in a GDP decline of about 7 percent in 2020 and a more gradual recovery. - Market shock scenario - Envisages a steep decline of GDP in the first two years of the risk horizon, with a modest recovery in the third year. - Drop in GDP under the market shock scenario is more severe than historical episodes (late 1980s banking crisis and global financial crisis). - Property prices assumed to drop approximately 35 percent for residential and more than 50 percent for commercial real estate over the risk horizon. - Equity prices decline by 40 percent in the first two years. - Oil price drops to US$27 per barrel over the same time span (from around US$65 per barrel for Brent oil at the cut-off date, end-June 2019). ### Key variables in the macroeconomic scenarios (Table 2) - Real GDP - Mainland (y/y percent change) - COVID Central: 2020 -5.5, 2021 3.8, 2022 3.3 - COVID Downside: 2020 -7.1, 2021 2.6, 2022 0.6 - Market Shock: 2020 -1.94, 2021 -3.27, 2022 0.51 - Unemployment rate - COVID Central: 2020 8.2, 2021 6.6, 2022 4.5 - COVID Downside: 2020 8.8, 2021 7.2, 2022 5.1 - Market Shock: 2020 5.89, 2021 8.87, 2022 9.29 - Consumer price index (y/y percent change) - COVID Central: 2020 2.5, 2021 2.8, 2022 2.0 - COVID Downside: 2020 2.5, 2021 2.8, 2022 2.0 - Market Shock: 2020 2.1, 2021 2.2, 2022 2.2 - Crude oil price - COVID Central: 2020 36.2, 2021 37.5, 2022 40.8 - COVID Downside: 2020 24.1, 2021 19.5, 2022 19.0 - Market Shock: 2020 34.0, 2021 27.0, 2022 26.0 - Equity index (OSEAX) (y/y percent change) - COVID Central: 2020 -24.2, 2021 35.3, 2022 30.7 - COVID Downside: 2020 -39.1, 2021 20.2, 2022 14.1 - Market Shock: 2020 -33.0, 2021 -10.0, 2022 14.0 - House price index (y/y percent change) - COVID Central: 2020 0.8, 2021 7.5, 2022 7.1 - COVID Downside: 2020 -1.8, 2021 7.2, 2022 6.0 - Market Shock: 2020 -16.6, 2021 -11.9, 2022 -11.2 - 3-month money market rate (average) - COVID Central: 2020 1.0, 2021 0.3, 2022 0.3 - COVID Downside: 2020 1.0, 2021 0.3, 2022 0.3 - Market Shock: 2020 2.6, 2021 3.2, 2022 3.3 ### Macroeconomic and market modeling notes - Remaining global variables simulated via Flexible System of Global Models (FSGM). - Domestic variables path estimated by Norges Bank with their DSGE model (NEMO). - Market variables (equity and house prices) estimated with models based on core macro variable paths. ### Staff’s top-down solvency stress test — Methodology - Credit risk evolution linked statistically to macro scenario variables. - For standardized-approach exposures (and exposures of IRB banks treated as standardized), satellite models link NPL ratios to macro variables. - Corporate lending: quarterly NPL time series from 2014, breakdown into 19 industries (17 individually modeled). - Retail lending: single quarterly series tracing back to 1990. - Models estimated via Bayesian Model Averaging (BMA). - Dependent variables subject to logistic transformation. - LGD and write-off assumptions for satellite-model based estimates: - Starting realized LGD: retail 30 percent, corporate 35 percent. - Starting write-off ratio assumed 30 percent. - Both parameters assumed to double under adverse conditions. - For IRB banks: - PDs follow the same evolution as default rates with a bank-specific sensitivity coefficient calibrated by regressing each bank’s IRB corporate PDs on average EDF of Norwegian corporates. - Same sensitivity applied to non-corporate IRB portfolios. - LGDs assumed to increase as a result of a drop of up to 15 percent in recovery rates (i.e., 1-LGD) over the horizon. - PD and LGD projections used to estimate RWAs and expected losses. - Interest income assumed to change according to volume changes and path for lending/borrowing rates in macro scenario. - Wholesale funding costs linked to CET1 ratio in a satellite model; statistically significant but economically negligible effect. - Non-interest income follows asset growth. - Losses in trading portfolios arise in the market shock scenario from sharp equity price drops and increases in interest rates and credit spreads. - Foreign currency long positions lead to modest gains under market shock due to NOK appreciation. ### Results — COVID scenarios - Aggregate CET1 ratio for in-sample banks drops by about 4 percent under the central scenario and 4½ percent under the downside scenario by the end of the risk horizon. - Under the COVID downside scenario, one bank would exhaust its buffers in excess of the hurdle rate, though without breaching it. - COVID scenarios show less severe impact overall than the market shock scenario over the full horizon; COVID shock is frontloaded causing comparable or worse impact in the first year, but faster recovery and easier financial conditions mitigate subsequent impact. ### Results — Market shock scenario - NPL ratios increase substantially, up to levels not seen since the mid-1990s. - Loan loss rates rise around 8 times the current level for retail lending and more than tenfold for corporate loans. - All market segments impacted; mining (including oil extraction and related services) and transport and storage particularly affected. - Assuming about three quarters of retail portfolio losses relate to mortgages, about 30 percent of overall losses would be linked to real estate-related lending. - Loan loss provisions are the main driver of sharp decline in capital ratios. - Average CET1 ratio for the 11 banks in scope drops to 12.4 percent by end of the risk horizon, from 17.4 at end-2019 — an average 5 percentage points fall. - Worst drops in CET1 ratios across banks range from 1.1 to 6.8 percent. - Decline in average total capital ratio is close to 6.2 percentage points. - Increase in risk-weighted assets contributes to capital ratio decline by 2.3, 1.8, and 1.3 percent in the years 2020 to 2022, respectively. - No material breaches of hurdle rates: buffers in excess of hurdle rates partially depleted for most banks and fully for three of them; some banks see worst drop in second year and recover by end of horizon. - Under the previous prudential regime (pre-CRR/CRDIV alignment and lower SRB), the average drop in CET1 would have been some 80 bps lower for the 11 banks in scope. ### Authorities’ top-down solvency stress tests — FSA and Norges Bank - FSA top-down (20 banks + macro bank) using Norwegian Aggregate Model (NAM) with FSA overlay, PD-generating satellite model (SEBRA), and bank profitability/solvency model; adverse scenario follows FSAP/Norges Bank market shock. - Only two banks breach assumed CET1 requirements under the adverse scenario. - Average CET1 reduced from 17.4 percent to 12.3 percent. - Reduction for individual banks varies between 1.6 and 6.7 percentage points. - 9 of 20 banks reach lowest CET1 in 2021; remaining in 2022. - Excluding CCB and CCyB, only two banks (not within the 11 of the FSAP exercise) are in breach of lower CET1 requirement. - Combined net profits across all banks accumulate to NOK -93 bn over 3 years. - Loan losses sum to a total of NOK 185 bn. - Norges Bank top-down (macro bank of nine large banks) using NEMO with policy rate unchanged and satellite bank model for capital adequacy. - Loan losses calculated using IMF “rule of thumb,” adjusted for IFRS9. - Macro bank sees fall of CET1 to level of the crisis hurdle rate, but without breaching requirements. - Total loan losses for the macro bank amount to around 2 to 3 percent per year. - Removal of the CCB and CCyB during stress period results in macro bank not breaching requirements but falling to the hurdle rate level. - Cross-exercise comparison - FSAP and FSA results are very similar on average. - Norges Bank estimates a larger average drop in CET1 ratio, in the order of -6.2 percentage points for the comparable sample of nine banks. - The excess CET1 at the end of the risk horizon is larger under the current prudential regime than it would have been under the previous regime. ### Banks’ bottom-up solvency stress tests - Three Norwegian banks performed bottom-up solvency stress tests using the same market shock assumptions. - Two banks experience lowest capital ratios in year three; one bank lowest in year two with recovery in year three. - Across the three banks, all measures of capital adequacy are between 3 and 4 percentage points lower under the adverse scenario compared to the baseline. - Measures include total regulatory capital, common equity, and tier 1 capital. - Despite significant declines, banks do not breach hurdle rates in bottom-up tests. *Source: IMF staff (Norway Financial Sector Assessment content unit 27).* ### 44. The authorities should consider: ### 1norea2020008 - 44. The authorities should consider: ### Sensitivity tests — Interest Rate Risk in the Banking Book (IRRBB) and FX Risk - Banks actively hedge risks from changes in interest rates and exchange rates; direct analysis limited by data gaps (hedging positions not available; maturity-bucket breakdown only at solo level). - IRRBB is analyzed using the Basel Committee standard approach from two perspectives: - Economic Value of Equity (EVE): change in net present value of assets and liabilities under stressed interest rate scenarios ('stock' perspective). - Net Interest Income (NII): difference between total interest income and total interest expense within a 1 year horizon ('flow' perspective). - Six interest rate shock scenarios derived: (i) parallel shock up; (ii) parallel shock down; (iii) steepener shock (short rates down and long rates up); (iv) flattener shock (short rates up and long rates down); (v) short rates shock up; and (vi) short rates shock down. - FSA shock calibration parameters: 1) parallel shock: 200 bps; 2) short-term rate shock: 300 bps; 3) long-term rate shock: 150 bps. - Results: - EVE reported by banks at end-2019 are all well below the 15 percent of Tier 1 capital Basel threshold for "outlier banks". - Parallel shock 'up' is not always the largest loss driver; non-parallel shocks (steepening or short-end shocks) can be more damaging for some banks. - Impact on NII is generally contained, always below 3 percent of Tier 1 capital. ### Market risk - Equity holdings represent between 0 and 1 percent of banks’ assets for banks in scope of the top-down exercise. - Equity risk estimated via historical Value-at-Risk (1 percent confidence level, 1-month horizon) using OSEAX daily quotes since 2013: - Reduction in capital ratios from equity portfolio losses would not exceed 25 bps at that confidence level and horizon. - Government bond simulation (co-movement since 1990 of 10-year benchmark yields across several sovereigns) with assumed durations of 5 or 10: - At 1 percent confidence level and 1-month horizon, losses would impact capital ratios by no more than 15 bps. - Corporate bond simulation (S&P Norway 10+ Year Investment Grade Corporate Bond Index since 2010): - At 1 percent confidence level and 1-month horizon, losses would affect capital ratios by no more than 5 bps. - Overall: contained market risk consistent with limited trading activity. ### Banks’ sensitivity tests (bottom-up) - Banks provided IRRBB and FX risk estimates following the Basel approach for IRRBB. - FX risk assessed via historical simulation of 12 years of daily changes at a 99 percent confidence level over horizons: 1 week, 1 month, 3 months, 1 year; assessed overall net open position and separately for EUR, USD, GBP at minimum. - Results: - IRRBB EVE and NII exposures remain below 2 percent of Tier 1 capital. - FX risk exposures do not reach 1 percent even at the longest horizons. - Conclusion: IRRBB and FX risk are small across the banks in scope due to systematic hedging. ### Concentration risk - Reverse stress test ordering large exposures by connected-client groups; assumed direct impact on capital and 50 percent LGD. - Finding: simultaneous default of no less than the top 10 counterparts would be required to exhaust the banks’ surplus Tier 1 capital — indicates a low degree of concentration. ### Recommendations on IRRBB data collection - The FSA is encouraged to collect consolidated data from banks on assets and liabilities with a breakdown by maturity bucket for time of repricing (or duration) and include hedging positions. ### Liquidity stress tests — overview and context - Mortgage-centered business model: property-related lending ~ close to 60 percent of loan portfolios on average; mortgages used as collateral for covered bonds. - FSA regularly collects data on collateral availability and applies a standard stress test assumption of a 30 percent fall in real estate prices; under that assumption, there would be no lack of collateral across the system for covered bonds backed by residential real estate. - High reliance on wholesale funding and relatively low incidence of deposits mean continued issuance of covered bonds and unsecured senior bonds; liability mix includes large share denominated in foreign currency and fixed rate while assets are predominantly NOK-denominated and floating-rate → continuous need to hedge interest rate and FX risk via derivatives. - Data limitations: granular, forward-looking data on derivative transactions (needed to model potential future margin calls) are not available; supervisory reporting includes only backward-looking account of margin calls in past two years (part of LCR reporting). ### Stressed LCR (1-month horizon) - LCR overview: - 30-calendar-day stress period. - For stress-testing, standard LCR assumptions augmented via increased run-off rates and haircuts. - Simulation design: combination of three "haircut" scenarios with seven "outflows" scenarios. - Outflow scenarios include one regulatory, three FSAP stress scenarios (retail funding shock, wholesale funding shock, or both), and three FSA/Norges Bank scenarios. - Results: - On average across the 11 banks, in all scenario combinations the average LCR would remain above 100 percent over a one-month horizon, though some banks would breach the threshold under the most severe combinations. - LCR in domestic currency (50 percent threshold imposed on some banks) remained above that threshold on average, with only one bank experiencing difficulties under some severe scenario combinations. - Norwegian banks are very liquid in EUR and USD; LCR in these currencies remain well above 100 percent in all cases. ### Cash flow analysis (up to one year) - Cash flow analysis complements stressed LCR by exploring balance of outflows, inflows, and counterbalancing capacity (CBC) over 18 maturity buckets (overnight to one year). - Initial CBC composition (June 2019) for ten among the largest domestically incorporated banks: - CBC comprised for more than half of central bank reserves. - More than ⅓ of CBC is represented by covered bonds. - Government bonds make a modest share of overall CBC. - Note: high share of central bank reserves driven by one single bank; CBC varies substantially by currency (central bank reserves mainly in EUR and USD). - Scenarios: mildly adverse and severely adverse (run-off and haircut rates calibrated accordingly). - Results: - Under mildly adverse scenario, all banks would handle net outflows up to one year with initial CBC. - Under severely adverse scenario, some banks would encounter difficulties beyond the 1-month horizon; the system would experience negative CBC gaps starting from the 3-4 months bucket. - Outflow analysis highlights rollover risk from sizeable dependence on debt issuance and potential liquidity drains from margin calls on derivative contracts; detailed transaction-level data would be required to quantify margin-call related liquidity risk (data currently not available). ### Recommendations on liquidity risk and data - Authorities should consider: - implementing stressed LCR and cash-flow analysis as complementary diagnostic tools; - deepening analysis of margining arrangements adopted by banks and mortgage companies for derivative and security lending transactions (e.g., centrally cleared vs bilateral transactions, symmetric vs asymmetric margining, etc.) and their implications for liquidity risk; - proceeding swiftly with plans to utilize data collected from trade repositories and build models to analyze liquidity risks arising from derivative exposures (in particular liquidity drains caused by margin calls in turbulent markets); - establishing a regular collection of data on the liquidity position of branches of foreign banks. ### Other operational recommendations (from earlier paragraph 44) - Authorities should consider: - collecting data useful to build historical series of default rates, analyze residential mortgages (based on stock information, i.e., existing loans), regularly monitor commercial real estate evolution and performance; - building more analytical and granular satellite models for credit risk, able to link the macro landscape with relevant variables (PDs, LGDs, etc.) at bank and asset class level; - deepening analysis of margining arrangements adopted by banks and mortgage companies for derivative and security lending transactions (e.g., centrally cleared vs bilateral transactions, symmetric vs asymmetric margining, etc.) and of the related counterparty risk, for its impact on profits and losses (via CVA) and risk-weighted assets (via CVA capital charges); - proceeding swiftly with plan to collect data from trade repositories and build models to analyze the counterparty risk arising from the network of derivative exposures across banks; - developing an IFRS 9-compatible top-down credit loss model to better capture the loan loss impact (more frontloaded than under an incurred loss accounting regime) under hypothetical downturn scenarios in top-down solvency stress tests. *Source: 1norea2020008 - 44. The authorities should consider (PDF chapter/section).* ### 70. The evolution of the financial sector in the past twenty years has been continuous and ### 1norea2020008 - 70. The evolution of the financial sector in the past twenty years has been continuous and ### Financial-sector evolution (1998–2018) - The evolution has been continuous and gradual, with the introduction of the covered bond law in 2007 being the main factor behind the most notable change. - Covered bonds legislation introduced in 2007 deeply reshaped financial linkages, increasing the prominence of mortgage companies in the financial network. - Analysis uses Statistics Norway financial accounts by sector to map linkages among Norwegian sectors and with the rest of the world. ### Net financial exposures and sectoral roles - 1998: - Largest net financial exposures: central government and households towards non-financial corporations (especially equity holdings). - Households towards insurance companies and pension funds (insurance policies and pension entitlements). - Banks’ position was relatively balanced: credit to households and non-financial corporations somewhat larger than borrowing. - Mortgage companies’ role overall negligible. - 2008: - Similar picture to 1998, but banks and mortgage companies had a more relevant role in financing the private sector. - 2018: - Most relevant change: mortgage companies became important in the financing of households. - Statistics recording issue: - When Norwegian pension funds and insurance companies purchase Norwegian covered bonds issued abroad, statistics record this as “foreign” funding of mortgage companies, complicating domestic vs. foreign holder identification. - Norwegian insurance companies and pension funds increasingly hold assets through mutual funds, posing additional challenges for interconnectedness analysis. - Households: - Norwegian households hold a large amount of assets in insurance companies and pension funds, reflecting pension entitlements beyond the public pillar I pension system. - Households’ assets towards insurers are on average larger than their bank deposits. - The relatively low share of liquid assets may be partly explained by Norway’s strong welfare system. - Dataset / visualization notes: - Graphs in Box 3 represent net financial assets (financial assets—financial liabilities) among sectors between 1998 and 2018. - Each vertex size is proportional to total assets; color shows net position; edges represent net bilateral exposures. ### Interbank network and market-based interconnectedness - Methodology: - Interconnectedness analyzed using market-based approach following Diebold and Yilmaz (2014), using equity returns for selected Scandinavian banks to measure spillovers. - Principal component analysis on equity returns reveals two clusters: major Scandinavian banks (including DNB) and medium-sized Norwegian savings banks. - For interconnectedness measures, smaller savings banks and Santander (foreign subsidiary) were excluded in some parts to focus on large Scandinavian banks. - Sample and period: - Analysis includes DNB, Danske, Nordea, Handelsbanken, SEB, and Swedbank, starting January 2007. - Rolling window of 2 years on daily stock returns used to decompose 10-day ahead forecast errors from a VAR. - Key results: - Index of total interconnectedness varied considerably over time, with a decline since 2017. - DNB and Danske estimated to have been at the receiving end of spillovers in the region (conditional on the sample); DNB can turn into a net contributor when compared to smaller Norwegian banks not in this sample. - Interconnectedness rebounded from a low in 2019. - Limitations: - Data limitations prevented a complementary analysis based on bilateral exposure data; available cross-holdings of securities capture only one layer. - Other layers (e.g., derivatives exposures) could be concentrated and act as shock amplifiers under extreme market volatility. ### Contagion assessment (direct and indirect) - Norges Bank framework: - Evaluates how fire sales of common asset holdings lead to valuation losses for banks (indirect contagion) and how recapitalization can cause direct contagion. - Model calibrated on securities holdings and large exposures data. - Quantitative outcomes: - Losses due to contagion are on average 1 pp of the banking sector’s CET1 ratio, but in some cases can reach 2.5 pp. - Losses result almost exclusively from indirect contagion. ### Recommendations (authorities should consider) - Maintain an accurate, updated, and easily retrievable map of the components and mutual relationships (in terms of control or other relevant influence) of major corporate groups: especially banking groups and also groups of borrowers. This supports checks of reliability of group maps used by banks for large exposures and concentration risk reporting. - Proceed swiftly to expand contagion analysis to further layers of interconnectedness among banks, particularly using data from trade repositories and building models to analyze contagion effects from the network of derivative exposures across banks. - Adjust or replace the two indicators for the non-bank sector in Norges Bank’s heatmap, as they have been ‘flashing red’ for some time without raising significant concerns upon closer inspection; maintaining them in current form risks repeated explanations and possible perception as a ‘false alarm’. ### Climate-related risks: Transition risk analysis — overview - Definitions: - “Transition risks”: risks originating from the transition to an economy that emits fewer greenhouse gases (distinct from physical risks). - Transition risks can be driven by changes in policy, advances in technology, or both. - Norway’s exposure: - Norway ranks high in adaptability and has relatively low exposure by the University of Notre Dame Global Adaptation Index; transition risks are highly relevant due to Norway being a major producer of oil and gas. - Analysis focus: - Two primary questions: 1. How does an increase in domestic carbon prices impact Norwegian banks’ credit exposures? 2. How does a fall in oil sector revenues affect Norwegian banks’ loan losses? - Identified two propagation channels and considered potential portfolio effects; sensitivity tests are partial equilibrium and do not account for use of revenues from higher carbon prices. ### Carbon pricing in Norway (key figures) - Average price paid per ton of CO2-equivalents in 2018: NOK 266 (about US$30). - Global average carbon price referenced: US$2. - IMF-estimated range necessary to achieve Paris Agreement targets: US$50–100. - No sectoral breakdown of actual carbon tax rates currently available for Norway; tax rates approximated based on actual carbon taxes paid by sectors and total emissions. ### Corporate cost of emissions channel: methodology and scenarios - Approach: - Firm-level balance sheet approach to calculate whether additional cost from higher carbon taxes makes firms unable to service debt. - Two hypothetical uniform carbon tax increase scenarios: US$75 and US$150 per ton CO2-equivalent. - Also tested scenarios preserving current sectoral differentiation by shifting the whole tax curve to average US$75 or US$150. - Output: change in “debt at risk” for each bank operating in Norway. - Key assumptions: - ‘No-pass-through’ assumption: firms fully absorb additional costs (no price or quantity adjustments), likely leading to some overestimation of firm-level impact relative to a full general equilibrium solution. - Individual firms assumed representative of their sectors to transition between sector- and firm-level data. - Emissions and firm data: - Scope 1 greenhouse gas emissions estimated for sectors using input-output tables from Exiobase. - Sector emissions matched to firm-level data from Orbis. - Orbis dataset includes more than 300.000 firms and is assumed representative of the economy as a whole. - Debt at risk definition: - “Debt at risk” = share of firms (in each sector) for which the interest coverage ratio (ICR) drops below a threshold due to the policy shock. - Depending on calibration, threshold set at either 2 or 1. - Analysis focuses on firms with ICR dropping from between one and two to below one (area of increased risk), though when linking to bank lending all firms are taken into account. - Sectoral impacts (under uniform increase to US$75): - Sectors most materially impacted: agriculture, forestry and fishing; water supply, sewerage, waste management and remediation activities; transportation and storage. - Impact strongest in sectors that currently have very low carbon taxes while being carbon-intensive (e.g., waste management enjoys significant exemptions). *Source: 1norea2020008 (IMF chapter).* ### 88. Under a carbon price of US$150, several other sectors are also at risk, notably ### 1norea2020008 - 88. Under a carbon price of US$150, several other sectors are also at risk, notably ### Carbon price scenarios and sectoral impacts - Under a uniform carbon price of US$150: - Wholesale and retail trade and accommodation and food services are notably at risk. - Emissions-intensive sectors that already face a high carbon tax are affected at US$150 (while imposing a US$75 carbon tax in these cases did not imply a significant change from current carbon taxes). - Imposing a parallel shift up of current carbon prices, such that the average corresponds to US$75 or US$150, while preserving differentiation across sectors: - Produces a roughly similar picture to uniform taxes. - Sectors unaffected under the uniform US$75 tax rate due to already high carbon prices (e.g., G=wholesale and retail trade and I=accommodation and food service activities) would be impacted by a parallel shift. ### Banks’ corporate debt at risk from higher carbon prices - Methodology and thresholds: - Banks’ debt at risk is calculated as the share of total sectoral exposure for which ICRs drop below a threshold value. - Configurations shown assume carbon price hikes to uniform levels of US$75 and US$150, or parallel shifts up in current prices to an average of US$75 and US$150, and ICR thresholds of 2 and 1. - Key statistics: - Following a uniform carbon price hike to US$150, about 4 percent of all corporate bank exposures see their ICR drop below 2. - The most affected bank in the sample sees around 16 percent of corporate exposures drop below the threshold. ### External demand channel: impact of higher global carbon prices on Norway’s oil sector and bank loan losses - Oil sector importance: - Oil sector makes up roughly 16 percent of GDP, about one fifth of total investments and state revenues, and over one third of total exports. - Estimated impacts on Norwegian oil revenues from global carbon price increases: - A US$75 carbon tax is estimated to lead to a reduction in Norwegian oil revenues of about 22 percent. - A US$150 carbon tax is estimated to lead to a reduction in Norwegian oil revenues by more than 36 percent. - Comparison: fall in revenues by about 43 percent during the 2014–16 drop in oil prices. - SVAR model results on loan losses: - Quarterly data from 2010: Q1 to 2019: Q2 for oil sector revenues, loan loss rates, GDP and the NOK/USD exchange rate are used. - Loan losses of banks and mortgage corporations show a significant reaction to shocks in oil revenues. - Maximum reaction occurs after 6 quarters and has an average size of roughly 1 standard deviation. - Estimated increases in loan loss rates: - A carbon price of US$75 is estimated to increase loan loss rates by about 0.3 percentage points. - A carbon price of US$150 is estimated to increase loan loss rates by roughly 0.4 percentage points. - Historical benchmarks: - The average loss rate stood at 0.21 percent in 2018 for the whole banking system. - The average loss rate was 0.6 percent in 2016 during the 2014–16 oil price drop. - Note on methodology: - Supply curves are based on break-even prices for global oil production sites as provided by Rystad Energy. - Demand curves are based on median price elasticity of oil demand from estimates in the literature. - The tax wedge is calculated on the basis of a standard estimate about embedded emissions in each barrel of oil (0.43 ton CO2/barrel) and a global carbon tax equal to either $75 or $150 per ton CO2-equivalent. - The analysis assumes the average firm in each sector is representative of the firms each bank is exposed to in that sector. ### Potential portfolio effects from enforced reductions in oil production - Scenario setup: - Three scenarios modeled to assess the impact of a forced reduction in oil production on the Norwegian economy. - Scenario 1: Norwegian oil majors’ output drops by 45 percent as a result of policy measures that do not entail a change in oil prices. - Scenarios 2 and 3: Output reduction is partly achieved by increasing the global carbon price to US$75 or US$150. - In each scenario, a given reduction of output is translated into lower earnings using an approximate breakdown of oil majors’ fixed and variable costs; the effect of carbon price on earnings is calculated in line with the external demand channel methodology. - Key quantitative findings: - To achieve emissions consistent with Paris targets, the Carbon Tracker Initiative estimates companies such as Equinor have to reduce their overall output by about 45 percent before 2040. - Depending on global oil prices and a firm’s ability to innovate, a reduction of output by 45 percent could see an oil producer’s income fall by up to 80 percent, due to decreased margin over generally high fixed costs. - The imposition of a carbon price of US$75–150 could further decrease output by approximately 8-12 percent. - Applying a dividend discount model, the fall in the share price for a company like Equinor is estimated to reach up to -50 percent, depending on the increase in carbon prices and assumptions about cost of capital and growth rates. - All scenarios assume a linear reduction in output such that the desired level is reached by 2040. - Spillovers to other equities and balance sheets: - Other Norwegian equities correlate on average one-to-one with the oil sector over time horizons up to one year; thus a fall in oil shares could lead to an equally large drop in other Norwegian shares. - Household and asset manager exposures (2018 data): - Norwegian households on average held about 22 percent of their financial assets directly in domestic equity, and another 3 percent indirectly via pension claims, life insurance policies, and holding of investment fund shares. - A 45 percent reduction in oil production that lowers equity values as modeled would lower the value of households’ financial assets by about 12.5 percent (assuming equity holdings representative of the Oslo Børs All-share Index). - Insurers and pension funds (collectively holding 10.6 percent of their financial assets as domestic equity) would see portfolio drops in the range of 4 to 5 percent. - Non-money-market investment funds (holding 11.7 percent of portfolios in domestic equity) would see potential drops between 5 and 6 percent. - Caveats: - Analysis is preliminary and conditional on a number of assumptions. - Further work is necessary to integrate side-effects that could weaken or strengthen the overall results. - Policies that result in significant changes in output could potentially have large effects on the global oil price; relaxing assumptions about oil price behavior could generate additional insight. ### Risk Assessment Matrix (selected climate- and oil-related threats) - Oversupply in the oil market: - Overall level of concern: High. - Likelihood of severe realization in the next 1–3 years: Medium. - Expected impact on financial stability if realized: Medium. - Key channels: - Oil prices remain depressed for an extended period. - Supply exceeds expectations due to failures of production coordination. - The global transition to a low-carbon economy accelerates. - Expected impacts: - Sharp decline in demand for oil-related mainland goods and services. - Liquidity conditions tighten and cost of capital rises. - Falling profit margins of energy-related companies weaken debt-servicing ability and increase banks’ corporate NPLs. *Source: IMF staff estimates and analysis in the Norway Financial Sector Assessment (extracted content).* ### Annex II. Table 2. Stressed LCR: Outflows Scenarios 1–4 ### Annex II. Table 2. Stressed LCR: Outflows Scenarios 1–4 ### Overview - Table reports stressed LCR outflow rates (In percent) under four scenarios: Scenario 1 (Regulatory), Scenario 2 (Retail), Scenario 3 (Wholesale), Scenario 4 (Combined). ### Outflows from Unsecured Transactions/Deposits — Retail deposits - deposits where the payout has been agreed within the following 30 days: 100 / 100 / 100 / 100 - deposits subject to higher outflows — category 1: 10 / 20 / / 20 - deposits subject to higher outflows — category 2: 15 / 30 / / 30 - stable deposits: 5 / 10 / 5 / 10 - derogated stable deposits: 3 / 5 / 3 / 5 - other retail deposits: 10 / 20 / 20 / 20 ### Operational deposits (maintained for clearing, custody, cash management or other comparable services in the context of an established operational relationship) - covered by DGS: 5 / 10 / 15 / 15 - not covered by DGS: 25 / 25 / 35 / 35 - maintained in the context of IPS or a cooperative network — not treated as liquid assets for the depositing institution: 25 / 35 / 25 / 35 - maintained in the context of IPS or a cooperative network — treated as liquid assets for the depositing credit institution: 100 / 100 / 100 / 100 - maintained in the context of an established operational relationship (other) with non-financial customers: 25 / 25 / 35 / 35 - maintained to obtain cash clearing and central credit institution services within a network: 25 / 25 / 25 / 25 ### Non-operational deposits - correspondent banking and provisions of prime brokerage deposits: 100 / 100 / 100 / 100 - deposits by financial customers: 100 / 100 / 100 / 100 - deposits by other customers — covered by DGS: 20 / 20 / 40 / 40 - deposits by other customers — not covered by DGS: 40 / 40 / 60 / 60 ### Additional outflows - collateral other than Level 1 assets collateral posted for derivatives: 20 / 30 / 30 / 30 - Level 1 EHQ Covered Bonds assets collateral posted for derivatives: 10 / 25 / 25 / 25 - material outflows due to deterioration of own credit quality: 100 / 20 / 20 / 20 - impact of an adverse market scenario on derivatives, financing transactions and other contracts — hlba approach: 100 / 100 / 100 / 100 - impact of an adverse market scenario on derivatives, financing transactions and other contracts — amao approach: 100 / 100 / 100 / 100 - outflows from derivatives: 100 / 100 / 100 / 100 - short positions — covered by collateralized SFT: 0 / 0 / 0 / 0 - short positions — other: 100 / 100 / 100 / 100 - callable excess collateral: 100 / 100 / 100 / 100 - due collateral: 100 / 100 / 100 / 100 - liquid asset collateral exchangeable for non-liquid asset collateral: 100 / 100 / 100 / 100 ### Loss of funding on structured financing activities - structured financing instruments: 100 / 100 / 100 / 100 - financing facilities: 100 / 100 / 100 / 100 - assets borrowed on an unsecured basis: 100 / 100 / 100 / 100 - internal netting of client’s positions: 50 / 50 / 50 / 50 ### Committed facilities — Credit facilities - to retail customers: 5 / 10 / 5 / 10 - to non-financial customers other than retail customers: 10 / 10 / 20 / 20 - for funding promotional loans of retail customers: 5 / 10 / 10 / 10 - for funding promotional loans of non-financial customers: 10 / 20 / 20 / 20 - other: 40 / 60 / 60 / 60 - to regulated institutions other than credit institutions: 40 / 75 / 75 / 75 - within IPS or cooperative network if treated as liquid asset by the depositing institution: 75 / 100 / 100 / 100 - to other financial customers: 100 / 100 / 100 / 100 ### Committed facilities — Liquidity facilities - to retail customers: 5 / 15 / 10 / 15 - to non-financial customers other than retail customers: 30 / 40 / 50 / 50 - to personal investment companies: 40 / 50 / 50 / 50 - to purchase assets other than securities from non-financial customers: 10 / 10 / 10 / 10 - other: 100 / 100 / 100 / 100 - for funding promotional loans of retail customers (to credit institutions): 5 / 15 / 10 / 15 - for funding promotional loans of non-financial customers (to credit institutions): 30 / 40 / 50 / 50 - other (to credit institutions): 40 / 50 / 50 / 50 - within IPS or cooperative network if treated as liquid asset by the depositing institution: 75 / 100 / 100 / 100 - to other financial customers (to credit institutions): 100 / 100 / 100 / 100 ### Other liability outflows - liabilities resulting from operating expenses: 0 / 0 / 0 / 0 - in the form of debt securities if not treated as retail deposits: 100 / 100 / 100 / 100 - others: 100 / 100 / 100 / 100 ### Outflows From Secured Lending And Capital Market-Driven Transactions — Counterparty is central bank - level 1 excl. EHQ Covered Bonds collateral: 0 / 0 / 0 / 0 - level 1 EHQ Covered Bonds collateral: 0 / 0 / 0 / 0 - level 2A collateral: 0 / 0 / 0 / 0 - level 2B asset-backed securities (residential or automobile, CQS1) collateral: 0 / 0 / 0 / 0 - level 2B covered bonds: 0 / 0 / 0 / 0 - level 2B asset-backed securities (commercial or individuals, Member State, CQS1) collateral: 0 / 0 / 0 / 0 - other Level 2B assets collateral: 0 / 0 / 0 / 0 - non-liquid assets collateral: 0 / 0 / 0 / 0 ### Outflows From Secured Lending And Capital Market-Driven Transactions — Counterparty is non-central bank - level 1 excl. EHQ Covered Bonds collateral: 0 / 0 / 0 / 0 - level 1 EHQ Covered Bonds collateral: 7 / 7 / 7 / 7 - level 2A collateral: 15 / 15 / 15 / 15 - level 2B ABS (residential or automobile, CQS1) collateral: 25 / 25 / 25 / 25 - level 2B covered bonds: 30 / 30 / 30 / 30 - level 2B asset-backed securities (commercial or individuals, Member State, CQS1) collateral: 35 / 35 / 35 / 35 - other Level 2B assets collateral: 50 / 50 / 50 / 50 - non-liquid assets collateral — counterparty is central govt, PSE<=RW20, MDB: 25 / 25 / 25 / 25 - non-liquid assets collateral — other counterparty: 100 / 100 / 100 / 100 *Annex II. Table 2. Stressed LCR: Outflows Scenarios 1–4 — Norway (International Monetary Fund).* ### 2018. Additional analysis on sample identical to FSAP (11 ### 1norea2020008 - 2018. Additional analysis on sample identical to FSAP (11 banks) ### Sample and Institutional Perimeter - One “macro bank” comprising nine large banks (about 60 percent of NO banking market by assets): DNB Bank, SpareBank 1 SR-Bank, Sparebanken Vest, SpareBank 1 SMN, Sparebanken Sor, SpareBank 1 Ostland, SpareBank 1 Nord-Norge, Sbanken and Sparebanken More. - ST comprises 11 largest domestic banks which hold approximately 60.5 percent of domestic banking sector assets. The ST does not include branches of foreign banks operating in Norway. - 20 institutions for which there is data coverage (including 11 largest). - Largest 11 banks which hold approximately 60.5 percent of the domestic banking sector assets. ### Data Sources and Baseline Dates - Internal audited data (where available) and other internal data. - Data from CRD IV reporting, reporting of banks’ corporate client exposures, and other supervisory and public data sources. - Supervisory and publicly available data. - Source: Supervisory data. - Baseline date: June 30, 2019. ### Overall Approach and Horizon - Balance sheet-based approach. - FSA: Balance-sheet approach based on consolidated data (source: FINREP), covering 20 banking groups. - FSA: For distribution of loan losses: unconsolidated data (source: ORBOF) covering smaller banks. - NB: Balance sheet-based approach based on consolidated public accounts, delivered by SNL/S&P MI. - Balance sheet-based approach. - Horizon: 3 years (2020–2022). ### Satellite Models for Macro-Financial Linkages and Risk Distribution - FSA: Total loan losses generated by proprietary macro model and assigned to individual banks according to risk in loan books. - NB: Satellite models for loan losses. - Satellite models for PDs, LGDs, and NPL ratios for credit losses. - FSA: Satellite-proxy PD model for distribution of loan losses on loans to NFC. - FSA: Satellite model for market risk. - NB: Banks’ loan losses in the stress scenario follow “rule of thumb” for total losses on corporate and household loans as a function of GDP developments. - Market losses from holdings of debt instruments (sovereign and other) based on modified duration and shocks to rates as assumed under scenarios. - FSA: Non-interest income projections based on growth in total assets. - FSA: NII developments based on output from macro model and expert judgement. - Non-interest income projections based on nominal GDP growth and expert judgment. ### Scenario and Tail-Shock Analysis (Solvency) - Scenario-based tests assess the impacts on the entire portfolio including the loans and, if applicable, the trading book. The COVID scenarios have been used exclusively in the FSAP team exercise. - The COVID central scenario is based on a preliminary version of the June 2020 WEO projections. The Market Shock and COVID downside scenarios are based on a given deviation of GDP from its long-term trend and COVID central, respectively. They all involve a series of domestic and global macroeconomic and financial variables. - The Market Shock Scenario is simulated using the IMF’s Flexible System of Global Models for the global variables and Norges Bank’s NEMO model for Norwegian variables. For the COVID scenarios the paths of most variables are obtained as conditional forecasts in a Vector AutoRegression conditional on the June 2020 WEO projections for GDP, unemployment and inflation, plus the WEO forecast for the oil price. A few remaining variables (e.g., interest rates) are calibrated judgmentally. - All scenarios are driven by a combination of external shocks and amplified by domestic characteristics. They include existing vulnerabilities and policy constraints. - Under the Market Shock Scenario, the Norwegian economy goes through an L-shaped growth path, with annual GDP growth of -1.94 percent, -3.27 percent, and +0.51 percent during 2020, 2021, and 2022, respectively. This corresponds to a cumulative deviation of real GDP growth of close to -9 percentage points over the first two years compared to the long-term trend (almost 3 standard deviations). - The COVID scenarios are based on a growth path with a different profile: in both scenarios GDP bottoms out in the second quarter of 2020, but at different depths and speed of subsequent rebound. In particular, under the COVID central GDP growth is -5.5, 3.8 , and 3.3 percent in 2020, 2021, and 2022, respectively; under the COVID downside it is -7.1, 2.6, and 0.6 percent in the three years, respectively. The drop of GDP in the first year corresponds to a divergence from the long-term growth path of more than 4 and 5 standard deviations under the COVID central and downward scenarios, respectively. - This economic slowdown is accompanied by an increase in the unemployment rate of close to 6 percentage points over the 3-year horizon under the Market Shock, and of approximately 1 and 1.6 percent under the COVID central and COVID downside scenarios, respectively (after peaking at 4.7 and 5.3 percentage points above 2019 in 2020, respectively). - The cumulative house price decline reaches -35 percent over the risk horizon under the Market Shock scenario, while under the two COVID scenarios house prices record a cumulative growth of 16.2 (central) and 11.6 percent (downside). - In all scenarios NOK depreciates by 8.7 percent in the first year, leading to a cumulative appreciation / depreciation of 10.7 percent at the end of the third year. ### Behavioral Adjustments and Constraints - Passive balance sheet assumption: - Balance sheets are assumed to be static, apart from credit growth, which is linked to nominal GDP growth. - Balance sheet composition remains constant throughout the stress test horizon. - The rate of increase of lending and funding is applied as of the end of the previous period, without taking into account the impact of defaulted exposures and the stock of outstanding loans during the current period. - Asset disposals and acquisitions are not permitted, except where in line with aggregate credit growth. - Banks’ credit portfolio composition is assumed to remain unchanged. - Capital increases are not permitted, unless these were approved prior to the cut-off date. - Defaulted exposures do not generate interest income after they become impaired. - Dividend payouts: - Bottom-up and FSAP Team: Dividend payouts made according to the most recent payout experience in case of positive net income and no payouts in case of negative net income. - FSA: Banks are assumed to make dividend payouts of 50 percent for periods with positive net income and no payouts in case of negative net income (no dividend payouts in the whole stress period in the NB exercise). ### Sensitivity Analysis - Sensitivity analyses evaluate impacts of single risk factors (one at a time) on the existing capital buffers: - FX shock - Interest Rate Risk in the Banking Book (IRRBB) - Credit concentration risk - Climate change transition risks. ### Calibration of Risk Parameters and Regulatory Standards - Projected losses distributed across different asset classes. - Point in time credit risk proxies/parameters calibrated by FSAP team. - FSA: National framework. - Hurdle rates: CET1, Tier 1, and total capital ratios, including SRB and D-SIB buffers and Pillar 2 requirements; leverage ratio, including 2 percent buffer and additional (1 percent) buffer for D-SIBs. ### Output Presentation and Reporting (Solvency) - FSA: System-wide capital shortfall. - Number of banks and percentage of banking system assets in the system that fall below the capital hurdle. Norges Bank: CET1 ratios for the ‘macro’ bank and for each of the 9 banks comprising the ‘macro’ bank are reported. - Impact of different result drivers, including profit components, losses due to realization of different risk factors. ### Liquidity Stress Testing: Data, Methodology, Risks and Buffers - Stress test can be run on all Norwegian banks (unconsolidated). The model also includes links between banks and covered bond companies. - FSA and Norges Bank are collaborating on developing a liquidity stress testing framework for Norwegian banks and mortgage companies. - Based on data availability, foreign branches could partially be involved into the exercise. - Data sources: CRD IV— LCR and NSFR; Non-CRD IV reporting—Balance sheet data, "Refinancing under stress."; Supervisory and publicly available data. Baseline date: June 30, 2019. - Methodology: Cash flow analysis of inflows and outflows from assets, liabilities and off-balance sheet items; Cash-flow based liquidity stress test using maturity buckets; Basel III LCR and NSFRs; Separate analysis for NOK, EUR and USD. - Risks modelled: bank-specific stress (e.g., bank ratings downgrade); domestic/global market stress triggering house price decline and NOK depreciation; shock to funding (stressed outflow and inflow factors) and available liquidity (haircuts). - Buffers: Banks’ liquidity reserves with haircuts; LCR buffer; Extended liquidity reserve (includes non-LCR available securities and bank deposits); Possible new issuances of covered bonds (where loans are readily available for transfer to the CB company and/or there are free cover pool assets within the CB company); Available and unencumbered liquid assets. ### Liquidity Tail Shocks, Regulatory Standards and Reporting - Projections are based on the expected behavior of banks, customers and depositors as well as other banks and market actors. - Run-off rates calculated following historical events and based on IMF methodology (for cash flow analysis). - The stress factors are generally applied for a 30-day period. The model assumes declining stress. The same stress factors are applied for the next period (day 30 to 90) and reduced to zero after three months. - Bank run and dry up of wholesale funding markets, taking into account haircuts to liquid assets. - Regulatory standards: National regulatory framework. LCR: 100 percent, consistent with Basel III LCR framework. NOK LCR: 50 percent (only applies to seven largest banks). - Output presentation: Survival horizon – time from initial event to net liquidity < 0. Net liquidity equals the difference between financing gap and the bank’s liquidity reserves. System-wide liquidity gaps. ### Interconnectedness: Methodology, Risks and Reporting - Methodologies used: combined direct and indirect contagion model based on Cont and Schaaning (2017) and Hueser et al. (2017); balance sheet-based interbank model by Espinosa-Vega and Solé (2010); market price-based spillover model by Diebold and Yilmaz (2014); cross-border network model by Espinosa-Vega and Solé (2010). - Risks: risk of indirect contagion due to price impact from fire sales of cross holdings; risk of direct contagion due to potential bail-in of MREL cross-holdings; credit and funding losses related to interbank cross-exposures (and cross-border banking exposures). - Buffers: Banks’ own capital buffers; Banks’ own capital and liquidity buffers. - Tail shock: Initial shock results from top-down stress test and may trigger funding difficulties. Pure contagion: Assumed failure of institutions. - Output presentation: Contagion analysis: additional amplification (pp of CET1 ratio) of shock from solvency stress test; network analyses with supervisory data; system-wide capital shortfall; number of undercapitalized and failed institutions, and their shares of assets in the system; evolution and direction of spillovers. *Annex IV. Banking Sector Stress Testing Matrix (STeM) — material from the source document.* --- _Source: https://www.imf.org/-/media/files/publications/cr/2020/english/1norea2020008.pdf_