## FINLAND: IMF Financial Sector Assessment Program (FSAP) Technical Note — Executive Summary and Selected Appendices (1finea2023003)

## Source details

**Canonical URL:** [FINLAND: IMF Financial Sector Assessment Program (FSAP) Technical Note — Executive Summary and Selected Appendices (1finea2023003)](https://www.imf.org/-/media/files/publications/cr/2023/english/1finea2023003.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2023/english/1finea2023003.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2023/english/1finea2023003.pdf.json)

---

### Background and Scope
- Finland is a small open economy significantly exposed to global financial and economic conditions.
- The Technical Note assesses systemic risk in the Finnish banking sector using stress tests under a severe yet plausible adverse scenario spanning 2022–25 that includes global and regional inflationary pressures, monetary policy tightness, financial market turmoil (shocks to term and risk premiums), and a major slowdown of economic activity.
- Exercises covered four SIs and three LSIs representing more than 93 percent of total banking assets.

### Financial Sector Landscape (key statistics)
- Total banking sector assets: EUR 870.4 billion at end–2021.
- Three largest institutions (Nordea Bank, OP Financial Group, Municipality Finance) account for 93 percent of domestic banking assets.
- Subsidiaries and branches of foreign banking groups operating in Finland amount to 44 percent of GDP.
- Pension and insurance assets:
  - Statutory earnings-related pension (private sector): EUR 161 billion.
  - Public sector and specialized regimes: EUR 94 billion.
  - Total insurance assets: EUR 89.6 billion (EUR 73.0 billion life; EUR 16.6 billion nonlife).
  - Fund management sector assets: EUR 180.0 billion at end-2021.
- Finnish banks’ regulatory metrics:
  - Regulatory capital position: 21.2 percent.
  - Leverage ratio: 6.2.
  - Liquidity coverage ratio (LCR): 171 percent.
- Profitability and margins:
  - Gross interest margins: 47.6 percent.
  - Return on assets (ROA): 0.6 percent.
  - Return on equity (ROE): 8.2 percent.
- Funding structure:
  - Banks mainly funded through wholesale funding (43 percent of total liabilities).
  - Funding for SIs and larger LSIs: 32 percent from retail deposits (implying heavy reliance on wholesale funding).
  - Wholesale funding composition noted as 61 percent of stable funding (40 percent unsecured and 21 percent secured funding) in detailed liquidity analysis.

### Macrofinancial Challenges and Vulnerabilities
- Geopolitical and energy risks: ongoing monitoring; energy companies have received public liquidity guarantees and bridge financing to avoid spillovers.
- Structural constraints: rapid population ageing and low productivity trends constrain medium-term growth.
- Sectoral concentration:
  - More than half of bank lending is to households as mortgages to households or to housing companies, and unsecured consumer lending.
  - Loans to housing corporations: 40 percent of total non-financial corporate (NFC) debt (partly represent household liabilities).
- Household interest-rate sensitivity:
  - As of August 2021, at least 93 percent of total loans to euro area households by Finnish banks were variable rate.
  - Many households purchased rate collars.
  - Banks recommended to stress test DSTI of mortgage applications using an interest rate of 6 percent.
- Cross-border exposures and interconnections:
  - Finnish banks’ net foreign assets: 26.5 percent of GDP.
  - Largest cross-border exposures to Nordics and euro area; trade exposures largest to Germany, Sweden, U.S., and the Russian Federation.
  - Nordea’s redomiciliation in 2018 increased banking sector assets from 250 to 350 percent of GDP and deepened exposure to Sweden.

### Systemic Risks (summary)
- Key risk sources:
  - Concentrated banking sector dominated by a few institutions.
  - Household indebtedness at historically high levels and interest-rate sensitivity.
  - High interconnections with the Nordic region.
  - NBFI sector concentration: PICs hold large, highly correlated portfolios with potential pro-cyclical behavior.
- Cyber risks elevated due to ongoing threats from cyber criminals and state actors; the Finnish Government expanded crisis-management responsibilities (June 2022) for FIN-FSA and BoF to establish backup systems for payment continuity.
- Climate risks (transition and physical) assessed as low.

### Stress Test Framework and Scenarios
- Four types of stress-test exercises:
  - Top-down solvency stress test (IMF internal framework).
  - Liquidity stress test (LCR, cash-flow, NSFR).
  - Wholesale funding cost stress test (bond-yield panel model).
  - Contagion and interconnectedness stress tests (domestic interbank and cross-border).
- Scenario horizon: 2022 Q1–2025 Q4 (four years).
- Scenario features:
  - Baseline aligned with October 2022 WEO projections.
  - Adverse scenario reflects elevated inflation, aggressive policy tightening, financial market turmoil, supply disruptions, and recession with spiking risk premia.
- Balance-sheet projection approach:
  - Quasi-static: asset allocation and funding composition remain the same; balance sheets grow in line with nominal GDP path by bank-specific weighted exposures.
  - Floor on rate of change in the balance sheet: zero percent (to prevent deleveraging).
  - Assumptions: banks do not issue new shares; dividends paid at 30 percent of net income when positive and compliant with supervisory capital requirements.

### Solvency Stress-Test Results (aggregate and drivers)
- Baseline scenario:
  - Aggregate CET1 capital ratio increases from 19.4 percent to 27.7 percent between 2021–25.
  - Four-year cumulative credit impairments: 3.8 percent of starting capital by end-2025.
- Adverse scenario:
  - Aggregate CET1 ratio falls to 13.8 percent from 19.4 percent (aggregate CET1 capital ratio declines by 7.4 percentage points to 12 percent at end-2025 in a separate aggregated reporting line; source preserves both figures in different sections).
  - Four-year cumulative credit impairments: 59.1 percent of starting CET1 capital by end-2025.
  - Market risk losses (four-year cumulative): 8.6 percent of starting CET1 under stress (versus 0.1 percent in baseline).
  - NIM dynamics:
    - Average NIM in adverse scenario rises to 2.2 percent by end-2023 when policy rate increases to 6.8 percent.
    - Baseline NIM at end-2023: 1.6 percent with average policy rate of 3.5 percent.
    - Average annual risk-free rate increases from −0.3 to 5.7 percent over 2021–25 in the stress scenario.
    - Average NIM in adverse scenario drops to 1.5 percent at end-2025 due to base rate decreases while risk-free rate remains high.
  - Drivers of capital decline: net income losses and deterioration in credit exposure quality that increase RWAs (contributing to an 8.2 percentage point charge).
  - Most banks meet regulatory minima (CET1 4.5 percent; total capital 8 percent; Tier 1 6 percent) though some fail to meet CCoB/O-SIIB requirements.
- Sensitivity analysis (role of NIM):
  - Holding policy and interbank rates at baseline path (illustrative) produces deeper distress: average CET1 falls to 10.1 percent in 2025; total capital adequacy ratios 12.2 percent in 2025; some banks unable to meet minimum capital requirements.

### Liquidity Stress-Test Results
- LCR exercises (scenarios and outcomes):
  - Basel III scenario: all banks passed; average LCR 162 percent.
  - Scenarios with higher outflows (Outflows; Inflows-Outflows; Extreme LCR stress): aggregate LCRs of 79 percent, 76 percent, and 72 percent, respectively — below 100 percent threshold.
  - Conclusion: banks vulnerable when outflows rise (demand deposits and wholesale funding run-off).
- Cash-flow analysis:
  - Net-funding gaps across maturity buckets (1 day to 1 year) remain negative despite counterbalancing capacity from asset sales.
  - Banking system has insufficient buffers to sustain large outflows at all maturity buckets.
  - Vulnerability driven by importance and short-term nature of wholesale funding and non-operational deposits from financial and non-financial counterparties.
- NSFR analysis:
  - Available stable funding: 479 billion USD.
  - Required stable funding: 404 billion USD.
  - Aggregate NSFR ratio: 118 percent.
  - Most individual bank NSFR ratios exceed 120 percent.
  - Interpretation: aggregate one-year horizon stability coexists with short-term market-access vulnerabilities.
- Liquidity tail shocks and calibration:
  - Retail scenario run-off assumptions include 10 percent (stable retail) and 20 percent (less stable retail) in some calibrations; other scenarios use up to 60 percent run-offs for non-operational wholesale deposits and haircuts on Level 2 assets as specified in Appendix VI (detailed tables preserved in source).
  - LCR stressed asset haircuts: Level 1 Basel 100 percent → Stressed 95 percent; Level 2a Basel 85 percent → Stressed 50 percent; Level 2b various stressed reductions down to 0 percent for common equity.

### Access to Funding and Funding-Cost Tests
- Bond-yield panel model (dependent variable: bank-specific 5-year bond yields) uses ROA, provisions over exposures, liquid assets over assets, and German sovereign yields.
- Observed weighted average of 5-year bond yields:
  - End-2021: 0.5 percent (0.6 simple average).
  - Q2 2022: 2.8 percent (2.5 simple average).
- Stress scenario projections:
  - Cumulative increase in five-year bond yield on average: 11.3 percentage points.
  - Increase in weighted average relative to December 2021: 8 percentage points (3.4 and 0.6 percentage points in the baseline).
  - Contribution of credit spreads: 5.4 (average) percentage points of the increase; 2 (weighted average) percentage points in alternative metric.
  - Timing: largest total yield increase in first year (risk-free spike); largest spread increases in last two years (credit impairments and capital depreciation).
- Implication: banks may cease issuing new debt when yields reach high levels and instead seek alternative funding; higher yields reduce wholesale funding access and may force balance-sheet reduction if wholesale liquidity unavailable.

### Contagion and Interconnectedness
- Domestic interbank contagion:
  - Contagion risks from domestic interbank exposures are very limited; gross domestic exposures among large banks are smaller than regulatory capital.
  - No single domestic bank failure would trigger another failure in the modeled four-bank market; no cascade effect.
- Cross-border contagion (BIS-CBS spillover scenarios):
  - Cross-border exposures to Denmark, Norway, and Sweden represent 80 percent of total cross-border exposures.
  - Scenario A (exposure to foreign banks only):
    - Sweden: 38.6 percent impact on Finnish banking capital.
    - Denmark: 33.8 percent impact.
    - Norway: 23.2 percent impact.
    - France: 17.8 percent impact.
    - Overall index of vulnerability: 6.6 percent.
  - Scenario B (credit shock to total exposure including nonbank claims):
    - A systemic default in any Scandinavian country yields a 100 percent impact on Finnish banks’ capital.
    - U.S.: 38.9 percent impact.
    - France: 21.4 percent impact.
    - Overall index of vulnerability: 39.8 percent.
  - Interpretation: strong Nordic linkages create major cross-border vulnerability despite limited domestic contagion.

### Key Policy Recommendations (preserved wording and addressees/timing where specified)
- 1 — Enhance liquidity buffers to cover a predetermined threshold of wholesale funding outflows over a five-day horizon. Addressee: FIN-FSA. Timing: NT.
- 2 — Closely monitor banks’ banking book quality and introduce independent stress test exercises of smaller scale on specific areas of risk (e.g., credit risk only). Addressees: FIN-FSA, and BoF. Timing: MT.
- 3 — Lead an effort to conduct a top-down Nordic-wide stress test coordinated exercise, considering interlinkages and spillovers, liquidity-solvency interactions, and expanding the coverage to both banks and NBFIs. Addressees: FIN-FSA, and BoF. Timing: MT.
- Additional supervisory steps and recommendations:
  - Direct banks to adjust funding structures toward longer-term wholesale funding sources and demand deposits where feasible.
  - Consider asking banks to hold liquidity buffers to cover a predetermined threshold of wholesale funding outflows over a five-day horizon.
  - Run more frequent liquidity stress tests and require banks to hold a more sufficient stock of HQLA to withstand stress-test results.
  - Closely monitor banking and trading books and intervene early to mitigate potential increases in banks’ risk premia due to asset-quality deterioration.
  - Collaborate with Nordic authorities on coordinated Nordic-wide stress testing covering banks and NBFIs, including liquidity-solvency interactions.

### Methodological and Modeling Notes (selected technical details preserved)
- Credit risk modeling:
  - IFRS9-based provisioning and transition-matrix methodology with Beta linking to adapt aggregate PD to stage 1 and stage 2 exposures; NPL ratios projected via cointegrated VEC model using unemployment, investment to GDP, and property prices.
  - LGD: collateralized lending calibrated via structural modelling using reported LTVs; unsecured via Frye-Jacobs; MuniFin LGD ceiling 25 percent.
  - PD adaptation formula preserved: PD_it = N( G( aPD_t ) + G( aPD_i ) − G( aPD_0 ) ) with G inverse CDF and N CDF of standard normal (formula preserved from source).
- Market risk valuation:
  - Modified duration approach with formulae preserved (MD_i = MacD_i / (1 + y_i) and MD_i update MD_{i-1} / (1 + Δy_i + Δrisk_free_i)); 50 percent hedging assumed for interest-rate and FX risk in debt holdings.
  - Equity revaluation floor: ΔEQUITY = −0.3 × prefmktpos (EQUITY_LLC + EQUITY_Shlcic) (formula preserved).
- Net Interest Income (NII) and pass-through:
  - Pass-through rates to policy rate for lending/liabilities: retail lending 0.5; corporate lending 0.7; interbank lending 0.5; term deposits 0.4; overnight deposits 0.3; other deposits 0.3.
  - Pass-through rates to risk-free rate for debt securities: 0.8; unsecured wholesale funding: 0.6; secured wholesale funding: 0.5.
- Bond-yield model coefficients (panel estimation):
  - ROA t: Estimate −3.4671; SE 1.3745; tStat −2.5224; pValue 0.0126.
  - Log POE t: Estimate 0.9483; SE 0.3189; tStat 2.9739; pValue 0.0034.
  - LOTA t: Estimate −0.0038; SE 0.0020; tStat −1.8738; pValue 0.0627.
  - Risk-Free t: Estimate 0.9817; SE 0.1428; tStat 6.8733; pValue 0.0000.

*Source: EXECUTIVE SUMMARY, 1finea2023003 - FINLAND: IMF Financial Sector Assessment Program (FSAP) Technical Note — Executive Summary.*

### EXECUTIVE SUMMARY __________________________________________________________________________ 6

### EXECUTIVE SUMMARY

### Background
- Finland is a small open economy significantly exposed to global financial and economic conditions.
- Post-GFC, Finland experienced a long recession led by the decline of its information and communications technology (ICT) sector; structural reforms improved competitiveness, growth, and employment but at a lower rate of growth.
- The economy was less affected by COVID-19 relative to other economies, but is navigating a weaker outlook given the war in Ukraine, despite limited direct exposures to Russia.
- The Technical Note assesses systemic risk in the Finnish banking sector using stress tests under a severe yet plausible adverse scenario that includes global and regional inflationary pressures, monetary policy tightness, financial market turmoil (shocks to term and risk premiums), and a major slowdown of economic activity.

### Financial Sector Landscape
- Total banking sector assets were EUR 870.4 billion at end–2021.
- The banking system is highly concentrated and dominated by a few institutions; the three largest—Nordea Bank, OP Financial Group, and Municipality Finance—account for 93 percent of domestic banking assets and are designated as SIs supervised by the SSM within the ECB.
- Subsidiaries and branches of foreign banking groups operating in Finland amount to 44 percent of GDP.
- Pension insurance companies (PICs) and fund management are the largest parts of the NBFI sector; statutory earnings-related pension for private sector workers had EUR 161 billion, public sector and specialized regimes had EUR 94 billion, total insurance assets were EUR 89.6 billion (EUR 73.0 billion life; EUR 16.6 billion nonlife), and the total fund management sector had EUR 180.0 billion in assets at end-2021.
- Finnish banks’ regulatory capital position: 21.2 percent; leverage ratio: 6.2; liquidity coverage ratio (LCR): 171 percent.
- Profitability and margins: gross interest margins of 47.6 percent; return on assets (ROA) 0.6 percent; return on equity (ROE) 8.2 percent.
- Funding vulnerabilities: banks are mainly funded through wholesale funding (43 percent of total liabilities), and hold significant derivatives exposure.

### Macrofinancial Challenges
- Authorities are monitoring geopolitical risks and energy market challenges—energy companies have faced liquidity needs and have been provided public liquidity guarantees and bridge financing to avoid spillovers into the financial sector.
- Rapid population ageing and low productivity trends constrain medium-term growth prospects.
- Potential for declining bank profitability and risks from residential real estate market bifurcation (Helsinki region versus other parts), which could affect smaller cooperative banks concentrated in declining regions.
- The impact of COVID-19 on commercial real estate (CRE) remains uncertain as work and lifestyle patterns evolve.
- Structural change: Nordea’s redomiciliation in 2018 increased banking sector assets from 250 to 350 percent of GDP, deepening exposure to other Nordic countries, particularly Sweden; FIN-FSA and BoF have increased staffing resources.

### Systemic Risks and Vulnerabilities
- Key sources of financial stability risk:
  - Concentrated banking sector dominated by a few institutions.
  - Household indebtedness at historically high levels, exacerbated by the pandemic.
  - High interconnections within the Nordic region.
  - NBFI sector concentration: PICs account for a large share of non-bank assets, have highly correlated portfolios, and exhibit potential pro-cyclical behavior.
- The exercises covered four SIs and three LSIs representing more than 93 percent of total banking assets.

### Stress Test Approach and Scenarios
- Four types of stress test exercises conducted:
  - Top-down solvency stress test.
  - Liquidity stress test.
  - Wholesale funding cost stress test.
  - Contagion and interconnectedness stress test (domestic and cross-border).
- Scenario features: global and regional inflationary pressures, monetary policy tightness, financial market turmoil (shocks to term and risk premiums), and a major economic slowdown.

### Solvency and Market Findings
- Aggregate CET1 ratio falls to 13.8 from 19.4 percent under the severe adverse scenario, remaining above regulatory requirements.
- Second-round effects: decreased profitability and deterioration of asset quality combined with an increase in the risk-free rate can substantially affect banks’ bond yields and access to wholesale funding.
- Market-risk channels (increase in risk premia and bond yields) can amplify funding pressures even when solvency metrics remain above regulatory minima.

### Liquidity Findings
- Banks are vulnerable to liquidity shocks due to reliance on short-term wholesale funding, especially sight deposits susceptible to large withdrawals and outflows.
- Under a stressed liquidity scenario, aggregate LCR falls to 79 percent (below the 100 percent threshold) in the case of large outflows.
- Domestic interconnectedness analysis: contagion risks from domestic interbank exposures are very limited.
- Cross-border analysis: the Finnish banking sector is vulnerable to a potential systemic event in Nordic countries due to strong linkages and high exposures.

### Access to Funding and Contagion
- Second-round effects through funding markets are important: deterioration in asset quality can raise banks’ bond yields and reduce access to wholesale funding.
- Cross-border interconnectedness with Nordic counterparties is a significant channel of vulnerability despite limited domestic contagion.

### Key Policy Recommendations (Table 1)
- 1 — Enhance liquidity buffers to cover a predetermined threshold of wholesale funding outflows over a five-day horizon. Addressee: FIN-FSA. Timing: NT.
- 2 — Closely monitor banks’ banking book quality and introduce independent stress test exercises of smaller scale on specific areas of risk (e.g., credit risk only). Addressees: FIN-FSA, and BoF. Timing: MT.
- 3 — Lead an effort to conduct a top-down Nordic-wide stress test coordinated exercise, considering interlinkages and spillovers, liquidity-solvency interactions, and expanding the coverage to both banks and NBFIs. Addressees: FIN-FSA, and BoF. Timing: MT.

### Authorities’ Considerations
- Encourage banks to adjust funding structures toward longer-term wholesale funding sources that are less susceptible to outflow risk.
- Consider asking banks to hold liquidity buffers to cover a predetermined threshold of wholesale funding outflows over a five-day horizon.
- Closely monitor banking and trading books and intervene early to mitigate potential increases in banks’ risk premia due to asset-quality deterioration.
- Collaborate with Nordic authorities to conduct a coordinated Nordic-wide stress test covering banks and the NBFI sector, including a top-down analysis under a Nordic-specific scenario that includes both EU and non-EU Nordic countries and focuses on interlinkages, spillovers, and liquidity-solvency interactions.

*Source: EXECUTIVE SUMMARY, 1finea2023003 - FINLAND: IMF Financial Sector Assessment Program (FSAP) Technical Note — Executive Summary.*

### 9.      Cyber risks to the financial sector are elevated, while climate risks are limited. Ongoing

### 9.      Cyber risks to the financial sector are elevated, while climate risks are limited. Ongoing

### Elevated cyber risk and limited climate risk
- Ongoing threats from cyber criminals and state actors pose risks to the Finnish financial system, particularly in the context of the war in Ukraine.
- The Finnish Government passed legislation in June 2022 further expanding the crisis management responsibilities of the FFSA and BoF to establish a backup system to maintain continuity of customers' daily banking payments.
- Both transition and physical climate risks are low; Finland is among the lowest risk countries in climate change vulnerability indices.

### Systemic risks and vulnerabilities
- Household indebtedness and interest rate sensitivity:
  - As of August 2021, at least 93 percent of total loans to euro area households by Finnish banks were variable rate, making households vulnerable to increases in interest rates.
  - Many households purchased rate collars which generally mitigate near-term impact of higher interest rates.
  - Banks are recommended to stress test the DSTI of mortgage applications using an interest rate of 6 percent and banks seem to follow this recommendation.
  - An increasing share of household debt is in the form of housing company loans; loans to housing corporations are 40 percent of total non-financial corporate (NFC) debt but to some extent represent household liabilities.
- Sectoral concentration and interconnections:
  - More than half of bank lending is to households as mortgages to households or to housing companies, and unsecured consumer lending.
  - The Finnish banking sector is highly interconnected with the wider Nordic region; Nordea Bank has large cross-border exposures.
  - Finnish banks’ net foreign assets are 26.5 percent of GDP.
  - BIS data show largest cross-border exposures are to the Nordics and euro area; largest trade exposures to Germany, Sweden, the United States (U.S.), and the Russian Federation.
- Funding risks:
  - 32 percent of bank funding comes from retail deposits for SIs and larger LSIs in Finland, implying heavy reliance on wholesale funding.
  - Banks are exposed to the risk of tightening global financial conditions; increases in cost of funding and limited ability to raise additional deposit funding may lead to reductions in balance sheets and credit if wholesale liquidity is not available.

### Key macrofinancial risks (as identified by the FSAP)
- Intensifying spillovers from Russia’s war in Ukraine (further sanctions, trade and financial disruptions, commodity price volatility).
- Commodity price shocks from continuing supply disruptions and negative demand shocks.
- De-anchoring of inflation expectations and stagflation: supply shocks to food and energy increase headline and core inflation, triggering more aggressive central bank tightening and a potential recession.
- Local Covid-19 outbreaks in slow-to-vaccinate countries or emergence of vaccine-resistant variants causing extended supply chain disruptions, slower growth, capital outflows, and debt distress in some EMDEs.

### Scenarios used in stress testing
- Two macroeconomic scenarios are used: a baseline (expected) scenario and an adverse scenario.
  - The scenario spans four years: 2022–25.
  - The baseline scenario is aligned with the October 2022 World Economic Outlook projections.
  - The adverse scenario reflects main risks in the RAM: higher inflation in the U.S. and advanced European economies, persistent geopolitical tensions, continued pandemic-related shortages, sustained demand and increases in food and energy prices, higher policy rates in euro area and U.S., and a resulting recession with tightened financial conditions and spiking risk premiums.

### Solvency stress test overview and coverage
- The FSAP solvency stress test is a top-down exercise covering four SIs and three LSIs that cover 93 percent of the banking sector assets.
- The stress test uses the IMF’s internal solvency stress-testing framework and includes market risk (equity, foreign exchange (FX), commodities, and interest rate risk) and income projections.
- The derivatives book is not considered due to lack of granular data; variation margins and derivatives portfolio cannot be stressed meaningfully.
- The stress test was conducted using supervisory data for Q2 2022 provided by the SSM; satellite models estimated using aggregate data from the Bank of Finland (BoF) and the FIN-FSA.

### Balance sheet projection assumptions
- A quasi-static approach: asset allocation and funding composition remain the same; balance sheets grow in line with the nominal GDP path specified in the scenario.
- To prevent banks from deleveraging, a floor on the rate of change in the balance sheet is set at zero percent.
- Balance sheet growth is estimated at a bank-specific level using weighted average GDP growth of all countries where the bank has significant exposure.
- Other factors: revaluation of assets from FX movements and conversion of a proportion of off-balance sheet items to the balance sheet.

### Credit risk methodology
- Credit risk covers domestic and cross-border household lending, corporate lending, corporate bonds in the banking book (amortized cost (AC)), and corporate bonds in the trading book (FVOCI).
- Wholesale debt instruments differentiated between AC, FVPNL, and FVOCI holdings:
  - FVPNL credit risk embedded in the market risk methodology (price changes reflect risk-free rate movement or credit risk premia).
  - AC securities: credit impairments estimated as a banking book asset.
  - FVOCI securities: estimated through both market risk and banking book credit impairment estimation.
- Probabilities of Default (PDs) estimation:
  - All financial institutions use IFRS9; credit impairments calibrated accordingly.
  - Scenario transition matrices projected via Beta linking where aggregate PD is projected and adapted to stage 1 and stage 2 exposures.
  - NPL ratios projected using an econometric model; aggregate household and domestic corporate NPL ratios are cointegrated and projected through a VEC model with exogenous macrofinancial variables (best-fit variables: unemployment, investment to GDP, and property prices). The interest rate environment is incorporated indirectly through investment to GDP.
  - Domestic consumer lending PDs derived through respective NPL ratio projection; cross-border household NPL ratio assumed to follow same growth path as domestic NPL ratio. SME PDs derived via corporate NPL ratio.
  - Large domestic corporate lending PDs use two PD paths: one from NPL ratio projection and one from a single-equation time-series model using Moody’s average corporate EDFs and macrofinancial variables; combined via linear programming.
  - Cross-border corporate lending PDs estimated through single-equation time-series models using country-specific average corporate Moody’s EDFs.
  - All NPL ratios and PD paths estimated at aggregate level and adapted to bank-specific PDs via formula (1):
    - PD_it = N( G( aPD_t ) + G( aPD_i ) - G( aPD_0 ) )  [formula preserved as in source; G is inverse CDF of standard normal, N is CDF of standard normal].
- Special-case institution:
  - Municipality Finance PLC (Kuntarahoitus Oyj–MuniFin) is government guaranteed; its banking book PD corresponds to Finland’s sovereign PD implied by sovereign spreads from stress and baseline scenarios.
- Loss Given Default (LGD) estimation:
  - LGD for collateralized lending calibrated through structural modelling using reported LTVs, starting-point reported LGDs, and property price paths.
  - LGD for unsecured lending calibrated through the Frye-Jacobs method.
  - For MuniFin, a ceiling to the LGD is set at 25 percent (the LGD level used in calibration of sovereign CDS spreads for developed countries).
- Risk Weighted Assets (RWAs):
  - STA and IRB portfolios differentiated.
  - RWAs change due to balance sheet growth, new provisions for credit losses, exchange rate movements, and triggered off-balance sheet items.
  - ASRF model implemented for unexpected losses for IRB portfolios per Basel III.
  - Regulatory through-the-cycle (TTC) PDs calibrated through scenario point-in-time (PiT) projections using a smoothing parameter recommended by Finnish authorities; regulatory downturn (DT) LGD is the maximum between reported DT LGD at period 0 and estimated PiT LGD.

### Market risk methodology
- Market risk sources assessed: interest rates, exchange rates, FX, and equity prices; impacts both capital resources (via P&L or OCI) and capital requirements.
- Market valuation losses for debt securities (sovereigns, financials, large corporates) estimated using a modified duration approach:
  - Modified duration MD_i = MacD_i / (1 + y_i)  [formula preserved].
  - Modified duration updated each year via MD_i = MD_{i-1} / (1 + Δy_i + Δrisk_free_i)  [formula preserved].
  - Change in value of a security computed via percentΔFVcredit spread and percentΔFVrisk free formulas (formulas (4), (5), (6) preserved).
- Sovereign debt holdings: yield curves constructed by linear interpolation of short- and long-term interest rates as specified in scenarios; losses = portfolio size × average duration × changes in yields and FX changes for foreign-currency debt.
- Non-sovereign yields move in line with sovereign yield plus a credit spread at the three-year horizon.
- Debt holding valuations assume 50 percent hedging for interest rate risk and FX.
- Market valuation losses for commodity, FX, and equity securities estimated as starting positions multiplied by scenario price/FX/equity changes.
  - Equity revaluation subject to a floor constraint: ΔEQUITY = −0.3 × prefmktpos (EQUITY_LLC + EQUITY_Shlcic)  [formula (7) preserved].

### Net Interest Income (NII) methodology
- NII stress tests measure banking income vulnerability to interest rate changes; interest rate risk estimated on interest-bearing assets and liabilities.
- BoF provided historical time-series of banking sector-wide average effective interest rates (stocks) for categories of interest-bearing assets and interest-paying liabilities.
- Interest rates are projected for stress and baseline scenarios via time-series regressions where dependent variable is the effective interest rate of total stock of interest-bearing assets/liabilities and independent variables include policy rate, interbank rate, and risk-free rate.
- Interest-bearing assets classified as:
  - i) consumer lending (secured and unsecured)
  - ii) corporate lending
  - iii) debt securities
  - iv) interbank lending
- Interest-bearing liabilities classified as:
  - i) term deposits
  - ii) overnight deposits
  - iii) other deposits
  - iv) interbank borrowing
  - v) collateralized wholesale funding
  - vi) uncollateralized wholesale funding

*Source: IMF staff calculations and FSAP material in the provided chapter.*

### 34.      The change in banking sector-wide interest rate projections has been implemented for

### 34.      The change in banking sector-wide interest rate projections has been implemented for

### Interest-rate projection methodology
- Projected bank-specific lending and borrowing rates are reported in the Financial Reporting (FINREP) regulatory data submissions.
- Projected bank-specific interest rates are multiplied with the projected stock of respective assets or liabilities after judgmental adjustments (e.g., excluding NPLs from interest-bearing assets) to estimate bank-specific interest-income and interest-expense levels.
- Projection formula (as presented):
  - 푦푦
    푖푖푖푖
    =푌푌
    푖푖0
    (
    푦푦
    푖푖0
    +∆푟푟
    푖푖푖푖
    )
    (8)
  - Where, y_it is the projected interest income/expense of the interest-bearing asset/liability Y_i at time t, and Δr_it is the aggregate projected change of the average interest rate of the specific type of interest-bearing assets/liabilities from time 0 to time t.
- Results for average asset and liability rates are shown in Figures 9 and 10 (figures referenced in source).
- Approach notes:
  - A similar approach has been used in the EBA 2018 stress test.
  - Interest rates on new business (flows) were not available; the FSAP implemented an alternative to the usual “repricing ladder” methodology.
- Pass-through rates for interest-bearing assets/liabilities linked with the policy rate:
  - i) retail lending (both secured and unsecured): 0.5
  - ii) corporate lending: 0.7
  - iii) interbank lending: 0.5
  - iv) term deposits: 0.4
  - v) overnight deposits: 0.3
  - vi) other deposits: 0.3
- Pass-through rates for interest-bearing assets/liabilities linked with the risk-free rate:
  - i) debt securities: 0.8
  - ii) unsecured wholesale funding: 0.6
  - iii) secured wholesale funding: 0.5
- More details referenced in Appendix VI.

### Non-Interest Income (non-II), other expenses, and tax/dividend assumptions
- Non-interest income projection:
  - Projected based on a Monte Carlo simulation due to many components not dependent on macro-financial variables.
  - Parameters based on observed historical trend and volatility of bank-specific non-II data.
  - 10,000 alternative non-II paths are projected.
  - The average of the paths is used as the baseline projection.
  - The path in the 10th percentile is used for the adverse scenario.
  - Both baseline and stressed non-II projections are adapted to nominal GDP growth, in line with the balance-sheet growth rule.
- Other expenses and rest Other Comprehensive Income (rOCI):
  - Projected according to nominal GDP growth under assumption that ratios remain constant.
  - Other expenses are 0.8 percent of total assets.
  - rOCI is 0.02 of total assets.
- Income tax and dividends:
  - Income tax calculated as a fixed rate when profit before tax (PBT) is positive (without counting OCI).
  - The tax rate is set at 20 percent.
  - The rate of dividends is the average observed dividend rate of every bank over the last five years.

### Solvency stress-test results (aggregate and drivers)
- Overall system resilience:
  - Banks meet hurdle rates over the stress testing horizon (Figure 11 referenced).
- Baseline scenario:
  - Aggregate CET1 capital ratio increases from 19.4 to 27.7 percent between 2021–25.
- Adverse scenario:
  - Aggregate CET1 capital ratio declines by 7.4 percentage points to 12 percent at end-2025.
  - Banks record weakened profits in the two first years on average; they record losses in the last year of the scenario.
  - Decline in capital ratio mainly due to net income losses and deterioration of credit exposure quality, which drives an increase in risk-weighted assets (RWA), contributing to a charge of 8.2 percentage points.
  - All banks meet the minimum capital requirements, but not all meet their CCoB/O-SIIB.
- Credit impairments:
  - Four-year cumulative credit impairments are 59.1 percent of starting CET1 capital by end-2025 under the adverse scenario.
  - Under the baseline scenario, four-year cumulative impairments are 3.8 percent of starting capital.
  - Most credit risk impairments are recorded during the two last years of the scenario.
- Net interest margin (NIM) dynamics:
  - Average NIM in the adverse scenario rises to 2.2 percent by end-2023 when the policy rate increases to 6.8 percent.
  - Baseline NIM is 1.6 percent at end-2023, with the average policy rate during the year at 3.5 percent.
  - Average annual risk-free rate increases from -0.3 to 5.7 percent over 2021–25 in the stress scenario.
  - Average NIM in the adverse scenario drops to 1.5 percent at end-2025 due to the decrease in the base rate while the risk-free rate remains high.
- Market risk:
  - Four-year cumulative market risk losses are 8.6 percent of starting CET1 under stress, versus 0.1 percent in the baseline.
  - Market risk losses are high during the first year of the adverse scenario but fall in subsequent years.
- Non-interest income under stress:
  - Stressed non-interest income is lower than baseline but remains positive on average.
- Heterogeneity across banks:
  - SIs record greater impairments in the trading book and higher volatility of returns in the market book.
  - Larger banks have higher risk appetite; smaller banks more risk averse with smaller trading books.
  - Some heterogeneity arises from distinct business models.

### Sensitivity analysis (role of NIM)
- Purpose:
  - Tests the effect of high NIM on banks’ ability to absorb macrofinancial shocks by keeping the policy rate and interbank rate in the adverse scenario at the baseline path.
  - Illustrative scenario; not necessarily plausible because increase in interest rates is one of the stress factors of the adverse scenario.
- Results:
  - Banks record lower capital ratios compared with the main stress test; some cannot meet minimum capital requirements.
  - Profitability is affected by lower NIM, but most banks continue to record profits.
  - Average CET1 falls to 10.1 percent in 2025.
  - Total capital adequacy ratios are 12.2 percent in 2025.
  - Some banks are unable to meet minimum capital requirements in 2025.
- Conclusion:
  - Banks would have faced significant distress if interest rates had remained low in the adverse macroeconomic scenario.

### Liquidity stress tests — overview
- Three liquidity exercises conducted: LCR stress test, cash-flow-based analysis, and a qualitative NSFR analysis.
- Scope and data:
  - Three major banks and three LSIs included; exercises implemented in EUR.
  - LCR measures ability to cover 30-day weighted net outflow with high-quality liquid assets (HQLA).
  - Cash-flow analysis simulates cash outflows over maturity buckets from 1 day to 360 days.
  - NSFR considers longer-term available funding relative to funding needs.
  - Tests use consolidated regulatory data (COREP and FINREP).
  - Stress tests use end-June 2022 data for the three SI banks; for LSIs end-March or end-May 2022 data used depending on exercise.
- Funding structure vulnerability:
  - Wholesale funding (secured and unsecured) accounts for 61 percent of stable funding (40 percent unsecured and 21 percent secured funding).
  - Retail financing at 39 percent.
  - Dependence on wholesale funding increases liquidity outflows under stress and heightens vulnerability to tightening in global financial conditions.

### LCR analysis (scenarios and outcomes)
- Primary stress scenarios:
  1. Liquid assets stress: Basel III scenario plus higher stress on eligible liquid assets
  2. Inflows stress: Basel III scenario plus greater drop in inflows
  3. Outflows stress: Basel III scenario plus a greater rise in outflows
  4. Inflows-outflows stress: Basel III scenario plus more significant drop in inflows and a greater rise in outflows (scenarios 2 and 3 combined)
  5. Extreme LCR stress: Higher stress on eligible liquid assets, drop in inflows and rise in outflows (scenarios 1, 2, and 3 combined)
- Passage thresholds:
  - LCR stress tests passed if ratio of liquid assets to net outflows > 100 percent.
- Results:
  - Basel III scenario: all banks passed; average LCR level of 162 percent.
  - Banks can sustain higher haircut in liquid assets and a greater drop in inflows (scenarios 1 and 2).
  - Scenarios involving higher outflows (scenarios 3, 4, 5): almost all banks fail; aggregate LCRs of 79 percent, 76 percent, and 72 percent, respectively.
- Outflows vulnerability:
  - Higher run-off rates applied to demand deposits of retail and unsecured wholesale funding drive failures in outflow scenarios.

### Cash-flow analysis (method and results)
- Method:
  - Uses contractual cash flows across maturity buckets to estimate net-funding gaps and counterbalancing capacity.
  - Net-funding gap = inflows minus outflows in each time bucket; cumulative net-funding gap is sum across buckets.
  - Counterbalancing capacity = sum of cash inflows banks can generate under stress at reasonable prices per bucket; cumulative counterbalancing capacity sums across buckets.
  - Scenario draws on solvency stress-test assumptions for consistency (stressed market values, market reaction, tightening of monetary conditions).
- Results:
  - Banking system has insufficient buffers to sustain outflows at all maturity buckets.
  - Negative cumulative net funding gaps persist despite counterbalancing capacity from asset fire sales across all maturity buckets.
  - Vulnerability driven by importance of wholesale funding and liabilities from non-operational deposits from financial and non-financial counterparties.
  - Large scale of wholesale funding and short-term maturity make system vulnerable to large withdrawals.

### Net Stable Funding Ratio (NSFR) analysis
- Aggregate results:
  - Available stable funding: 479 billion USD.
  - Required stable funding: 404 billion USD.
  - Aggregate NSFR ratio: 118 percent.
  - At individual bank level, most ratios exceed 120 percent.
- Interpretation:
  - Large scale of wholesale funding allows NSFR to exceed 100 percent despite creating vulnerabilities related to market access and short-term nature.

### Liquidity results and policy implications (recommendations and findings)
- Key vulnerabilities:
  - Heavy reliance on wholesale funding, especially unsecured wholesale funding (~40 percent of total available funding as of mid-2022).
  - About 70 percent of unsecured wholesale funding consists of sight deposits from corporate and financial institutions, very vulnerable to large outflows.
  - Short-term nature of wholesale funding amplifies cash-outflows at early maturity buckets, producing large negative cumulative net funding gaps despite counterbalancing capacity.
  - LCR drops from 162 percent in the Basel III scenario to 79 percent when scenario includes an increase in outflow run-off rates (e.g., retail deposits and wholesale funding factors).
- Overall conclusion:
  - While NSFR indicates sufficient resources at an aggregate one-year horizon, LCR and cash-flow analyses reveal significant vulnerabilities in short-term liquidity under stress due to wholesale funding structure and potential large outflows.

*Source: IMF staff calculations (excerpts from the referenced chapter).*

### 57.      The analyses suggest the need for tighter liquidity regulation. Ideally, banks should

### 1finea2023003 - 57.      The analyses suggest the need for tighter liquidity regulation. Ideally, banks should

### Liquidity regulation and supervisory recommendations
- Authorities should direct Finnish banks to adjust their wholesale funding over time, aiming to increase the proportion of longer-term and demand deposits, to the extent that is feasible.
- Authorities are recommended to run more frequent liquidity stress test exercises and should require banks to hold a more sufficient stock of HQLA to withstand the stress test results.
- The analyses suggest the need to: 
  - closely monitor banks’ banking book quality, in particular for banks with a higher credit risk appetite;
  - introduce independent stress test exercises on a smaller scale throughout the year focused on specific areas of risk (e.g., credit risk or interest rate risk only), allowing for multiple scenarios and sensitivity analyses.

### Access to funding analysis — overview
- A second-round effect stress test measured the impact of solvency stress test results on bank access to wholesale funding.
- The Finnish financial system is large and most of its funding is from wholesale sources.
- The analysis estimated how banking financial position, in combination with the macro-financial environment, affect banking bond yields.
- Banks that do not issue bonds were excluded. The analysis focuses on marginal funding cost only (i.e., the cost of a bank issuing new debt).

### Model specification
- A panel model was calibrated where the dependent variable is bank-specific 5-year bond yields and independent variables are indices for banks’ financial positions and the risk-free rate.
- Banking position fundamentals consist of four dimensions: i) solvency, ii) profitability, iii) asset quality, and iv) liquidity.
- Implemented variables:
  - Profitability: Return on assets (total profits over total assets)
  - Asset quality: Provisions over credit exposures
  - Liquidity: Liquid assets to total assets (the liquid assets ratio has been assumed fixed in projections)
  - Risk-free rate: German sovereign bond yields
- Banking fundamental projections from the solvency stress test exercise were implemented in the model; bond yield projections were generated for both baseline and stress scenarios.

### Results on bond yields and key statistics
- Observed weighted average of 5-year bond yields at end-2021: 0.5 percent (0.6 simple average).
- Observed in Q2 2022: 2.8 percent (2.5 present simple average).
- Over the course of the stress scenario:
  - cumulative increase in the five-year bond yield on average: 11.3 percentage points;
  - increase in the weighted average according to the total five-year yield value in December 2021: 8 percentage points (3.4 and 0.6 percentage points in the baseline scenario).
- Contribution of credit spreads to the increase:
  - 5.4 (average) percentage points of the increase is due to credit spreads;
  - 2 (weighted average) percentage points of the increase is due to credit spreads.
- Timing of increases:
  - Greatest increase in total yield is in the first year of the stress scenario (when the risk-free rate increases the most).
  - Greatest increase in spreads is observed in the last two years of the stress scenario (largest increase in credit impairments and capital depreciation).
- Interpretation:
  - Projected yields are indicative of ease of access to market funding, not projected rates for interest expenses.
  - When yields increase to a high level, banks are expected to stop issuing new debt securities and look for alternative funding sources until yields normalize.

### Implications for Finnish banks
- Despite strong capitalization, Finnish banks may face constraints to wholesale funding access during a stress period when banking solvency decreases and risk-free rates increase.
- Potential liquidity distress due to wholesale funding outflows (as shown in liquidity stress test results) may amplify funding-access problems.
- The decade-long low interest rate environment supported easy access to wholesale funding; higher market rates may render that business model less viable.
- Potential losses and portfolio quality deterioration, together with an increase in interest rates, may reduce access to funding and cause liquidity issues.

### Domestic interbank contagion and interconnectedness
- Analysis based on matrix of bilateral domestic interbank gross credit exposures of the six large banks using COREP large exposures data; reporting date end-March 2022.
- Stress test assumes hypothetical default of each bank, one at a time, with subsequent rounds if defaults propagate.
- Banking default criterion: failure to meet minimum capital requirements (either 4.5 percent CET1 ratio or 8 percent total capital ratio).
- Model outputs:
  - index of vulnerability — probability of default of a counterparty due to contagion;
  - index of contagion — probability of a contagious systemic event if the counterparty defaults.
- Two spillover scenarios tested:
  - i) simple credit shock scenario;
  - ii) credit and funding shock scenario.
- Parameters used:
  - Lambda (credit shock): set to 65 percent.
  - Delta (funding shock): set to 50 percent.
  - Rho (funding shock): set to 35 percent.
- Findings:
  - Contagion risks from domestic interbank exposures are very limited; domestic interbank positions are small compared to banks’ capitalization.
  - For the six banks, the sum of their gross domestic exposures to the other three banks is smaller than their regulatory capital.
  - No single failure of a domestic bank would trigger the failure of another bank in either scenario; no cascade effect in this four-bank market.
  - As at end-March 2022, none of the banks is found undercapitalized at the regulatory minimum after shocks on domestic interbank exposures.
  - All banks have low vulnerability to spillovers in the model, although some banks show an index of vulnerability significantly higher than others.
  - The index of contagion (average percentage loss of other banks due to failure of a given bank) is low overall.

### Cross-border contagion and interconnectedness
- Cross-border exposures to Denmark, Norway, and Sweden represent 80 percent of total cross-border exposures.
- Two BIS-CBS based spillover scenarios (March 2022) assessed:
  - Scenario A: reporting banks’ exposure to foreign banks only, considering both credit and funding shocks.
  - Scenario B: credit shock to total exposure of the banking sector, including claims to banks, governments, and the nonfinancial sector.
- Same parameters as interbank analysis: lambda = 65 percent, delta = 50 percent, rho = 35 percent.
- Default threshold for Finnish banking sector: cumulative CET1 ratio below 4.5 percent.
- Results:
  - Scenario A:
    - Finland most vulnerable to Sweden: a potential distress in Swedish banking sector will have a 38.6 percent impact on banking capital in Finland.
    - Denmark: 33.8 percent impact.
    - Norway: 23.2 percent impact.
    - France: 17.8 percent impact.
    - Overall index of vulnerability in Scenario A: 6.6 percent.
  - Scenario B:
    - A systemic default in any Scandinavian country will cause a 100 percent impact on Finnish banks’ capital.
    - The U.S. impact: 38.9 percent.
    - France: 21.4 percent.
    - Overall index of vulnerability in Scenario B: 39.8 percent.
  - Sweden, Denmark, and Norway are the most contagious countries in both scenarios, with Scenario B magnitudes significantly greater than Scenario A.

### Summary of actionable supervisory steps
- Direct banks to adjust funding structure toward longer-term and demand deposits where feasible.
- Increase frequency of liquidity stress testing and require higher HQLA buffers to meet stress outcomes.
- Closely monitor banking book quality, especially for banks with higher credit risk appetite.
- Implement smaller-scale, focused stress tests throughout the year (e.g., credit risk only, interest rate risk only) with multiple scenarios and sensitivity analyses to detect potential balance sheet deteriorations that may be masked by high capitalization.

*Source: 1finea2023003 - 57.      The analyses suggest the need for tighter liquidity regulation. Ideally, banks should*

### 2. Channels of

### 2. Channels of Risk Propagation

### Methodology (Solvency Stress Test)
- FSAP team satellite models and methodologies.
- Balance-sheet regulatory approach.
- Market risk:
  - Treated as an add-on component, with a separate calibration.
  - Market risk stress scenario impacts capital resources (either via profit and loss or via Other Comprehensive Income (OCI)) and capital requirements (RWA).
  - Impact on capital resources comprises positions in the trading book as well as other fair valued items in the banking book.
  - Impact on RWA for market risk evolves with balance sheet assumptions.
- Traded risk impact:
  - Revaluation of trading assets (FVPL) and securities classified as fair value through other comprehensive income (FVOCI) securities by counterparty: central government (by country issuers), credit institutions, other financial institutions, and nonfinancial corporates.
  - Credit spreads on sovereign, credit institutions and corporate securities interpolated using bank-specific residual maturity at the book and issuer level (sovereign issuers by country and individual corporate issuers by ISIN codes).
  - Credit spreads on other securities estimated on a hypothetical portfolio using a duration proxy.
  - Valuation effects assessed using a modified duration approach.
  - Hedges are considered as ineffective under stress.
- Losses on securities portfolios based on duration approach.
- Losses on equities (both long and short positions) based on stock market price movement specified by the scenario.
- Credit risk:
  - For internally modelled exposures (IRB): projection of PiT and TTC PDs, LGD, EAD and RWA.
  - For STA exposures: projection of new flows of defaulted exposures, coverage ratio for defaulted loans, and risk weight downgrade for performing exposures.
  - Credit risk projections for IRB and STA exposures cover credit institutions, nonbank financial corporates, and households.
  - Corporate PDs for largest exposures proxied by Moody’s EDFs.
  - Resulting impact translated into credit loss impairment charges and shifts to RWAs due to capital charges for defaulted assets.
- Provisioning for IRB and STA modeled using IFRS9 transition matrix approach:
  - Transition matrices, PiT PDs, PiT LGDs for loan and securities classified under financial asset measured through amortized cost (AC), and other comprehensive income (FVOCI) modeled using COREP data.
- Funding costs projected at the portfolio level using funding structure by product (retail and wholesale deposits, secured and unsecured debt securities, repo, etc.) and maturity bucket (overnight vs. term).
  - Funding projections capture systematic risk (linked to the scenario) and idiosyncratic risk (for spreads on debt instruments issued over benchmark).
  - Funding cost projections utilized bank level data on 12 Irish banks from COREP templates.
  - Lending rates projected at the system level and attached to bank-specific interest rates and outstanding amount at cut-off date (interest rate on corporate and household loans and debt securities).

### Stress Test Horizon
- 2022 Q1–2025 Q4 (4 years)

### Tail Shocks Scenario
- Two Scenarios:
  - A baseline scenario based on the April 2022 WEO macroeconomic projections.
  - An adverse scenario capturing the key risks in the RAM, relying on GFM, a structural macroeconometric model disaggregated into forty national economies (documented in Vitek (2018)). Scenarios for foreign countries where Finland has significant exposure are extracted from GFM and are internally consistent with country scenarios of other ongoing FSAPs.

### Risks Covered and Behavioral Adjustments
- Risks covered:
  - Credit (on loans and debt securities).
  - Market (valuation impact of debt instruments through repricing and credit spread risk as well as the P&L impact of net open positions in market risk factors such as foreign exchange risks).
  - Interest rate risk (IRRBB) on the banking book.
  - Concentration risk by sensitivity analysis.
  - Solvency and liquidity risk interactions, mainly through asset haircuts.
- Behavioral adjustments:
  - Quasi-static approach for balance sheet growth: asset allocation and composition of funding remain the same; balance sheet grows in line with nominal GDP paths of major geographical exposures and subject to reduced credit demand in material jurisdictions and FX shock from revaluation effects on foreign currency loans specified in the stress test scenario.
  - Rate of change of balance sheets set at a floor of zero percent to prevent deleveraging; this constraint is binding in the adverse scenario.
  - RWAs projection:
    - Standardized portfolios: RWAs change due to balance sheet growth, new inflows of non-performing loans, new provisions for credit losses, exchange rate movements, and conversion of a portion of off-balance sheet items to on-balance sheet items.
    - IRB portfolios: through-the-cycle PDs, downturn LGDs and EAD for each asset class/industry used to project risk weights.
  - Interest income from non-performing loans is not accrued.
  - Assumption: banks do not issue new shares or make repurchases during the stress test horizon.
  - Dividends assumed to be paid out at 30 percent of current period net income after taxes (i.e., only if net income is positive) by banks that were in compliance with supervisory capital requirements.

### Regulatory and Market-Based Standards and Parameters (Solvency)
- National regulatory framework: Basel III regulatory minima on CET1 (4.5 percent) and include any requirements due to systemic buffers for three other systemically important institution (O-SII).
- Evaluated metrics:
  - CET1 against 4.5 percent.
  - Total banking capital adequacy ratio against the 8 percent level.
  - Tier 1 capital ratio against the 6 percent benchmark.
  - Leverage ratio against the 3 percent Basel III minimum requirement.
- Same hurdle rate used for baseline and adverse scenario.
- Hurdle rates for CET1, T1 and total capital adequacy do not include capital conservation and capital countercyclical buffers as well as pillar 2 requirement.
- Banks ending the stress test horizon with capital level or leverage ratio below relevant hurdle rates are considered to have failed the test.

### Reporting Form for Results (Solvency)
- Outputs reported using charts and tables, potentially including:
  - Evolution of capital ratios for the system and as groups (retail banks and large international banks).
  - Impact of different result drivers, including profit components, losses due to realization of different risk factors.
  - Capital shortfall as sum of individual shortfalls; in euros and in percent of nominal annual GDP.
  - Number of banks and corresponding percentage of assets below the regulatory minimum (or below the minimum leverage ratio).

---

### Banking Sector: Liquidity Stress Test (Top-Down by IMF)

### Institutional Perimeter and Data
- Top-Down by FSAP team.
- Institutions included: Six banks subcategorized as SIs (three banks) and LSIs (three banks). One SI is not included due to lack of data.
- Market share coverage: about 80 percent of the banking sector, with 73 percent for SIs and 7 percent for LSIs.
- Latest data: April 2022.
- Source: supervisory data (LCR, NSFR, and ALMM Maturity Ladder template).
- Scope of consolidation: banking activities of the consolidated banking group for banks headquartered in Finland. Foreign subsidiaries assessed on an unconsolidated level covering domestic activities only.

### Channels of Risk Propagation (Liquidity)
- Basel III LCR and cash-flow based liquidity stress test using maturity buckets by banks, incorporating both contractual and behavioral (where available) assumptions about combined interaction of funding and market liquidity and different level of central bank support.
- Liquidity test in EUR, USD, and Sterling.

### Risks and Buffers (Liquidity)
- Risks: funding liquidity and market liquidity.
- Buffers: counterbalancing capacity, including liquidity obtained from markets and/or the central bank’s facilities. Expected cash inflows included in the cash-flow based and LCR-based analysis.

### Tail Shocks (Liquidity) — Size of the Shock
- Run-off rates calibrated to reflect scenarios of system-wide deposit runs and dry-up of unsecured wholesale and retail funding, with additional run-off for non-resident deposits calibrated following historical events, recent international experience, and IMF expert judgment.
- Retail scenario key assumptions:
  - (i) 10 percent run-off rates for stable retail deposits and 20 percent for less stable retail deposits;
  - (ii) 10-35 percent for operational deposits and 20-40 percent for non-operational deposits;
  - (iii) no changes in liquid asset weights.
- Wholesale scenario key assumptions:
  - (i) 5 percent run-off rates for stable retail deposits and 10 percent for less stable retail deposits;
  - (ii) 15-35 percent for operational deposits and 40-60 percent for non-operational deposits;
  - (iii) no changes in liquid asset weights.
- Combined run-off and price shock scenario key assumptions:
  - (i) 10 percent run-off rates for stable retail deposits and 20 percent for less stable retail;
  - (ii) 15-35 percent for operational deposits and 40-60 percent for non-operational deposits;
  - (iii) liquid assets weight reduction of 0-5 percent for level 1 assets, 3-20 for level 1 covered bonds, 5-15 percent for level 2A assets and 5-25 for level 2B assets.
- Liquidity shocks simulated for:
  - 1–month for LCR.
  - 5-days, 1-month, 3-months, and 1-year for the cash-flow based approach.
- Haircuts of high-quality liquid assets (HQLA) calibrated against ECB haircuts, past Euro Area FSAPs, and market shocks for investment securities and money market instruments in the solvency stress test.

### Regulatory Standards (Liquidity)
- Consistent with Basel III regulatory framework (LCR).
- Liquidity shortfall by bank.

### Reporting Format (Liquidity)
- Liquidity ratio or shortfall by groups of banks and aggregated (system wide).
- Number of banks that can meet or fail their obligations.

---

### Banking Sector: Interconnectedness Analysis (Top-Down by IMF)

### Institutional Perimeter and Data
- Top-Down by FSAP team.
- Institutions included for cross-border contagion: country-pair bilateral exposure across Nordic/Baltic region, rest of Euro Area, US, and Russia.
- Data and baseline date: BIS consolidated banking statistics.

### Channels of Risk Propagation (Interconnectedness)
- Balance-sheet model: Network model by Espinosa-Vega and Solé (2010).

### Tail Shocks (Interconnectedness)
- Size of the shock: pure contagion from financial distress in foreign countries.
- Default threshold: banks default if their CET1 capital ratios fall below 4.5 percent (regulatory minimum).

### Reporting Format (Interconnectedness)
- Capital shortfall systemwide, by bank and by group: contagion and vulnerability scores.
- Amplification and cascade effects, direction, and size of spillovers within the network.

---

### Banking Sector: Funding Cost (Top-Down by IMF)

### Institutional Perimeter and Data
- Top-Down by FSAP team.
- Institutions included: Two banks that do not issue bonds.
- Market share coverage: about 85 percent of the banking sector.
- Data and baseline date:
  - Publicly available market data on banking bond yields (July 2022).
  - Historical bank-specific balance sheet and PnL data from Bloomberg.
  - Solvency stress-testing projections.

### Channels of Risk Propagation (Funding Cost)
- Methodology: Panel regression between cost of funding and bank specific performance indicators.

### Risks, Behavioral Response, and Tail Shock
- Risks: credit spreads and interest rate.
- Firm behavioral response: firms are not allowed to raise capital.
- Size of the shock: drop in banking profitability and asset quality due to solvency stress test.

### Reporting Format (Funding Cost)
- Relationship between banking performance and access to funding.
- Projection of marginal wholesale funding cost under the alternative scenarios.
- Market-based analysis; no capital thresholds are applied.

---

### Appendix III — Probabilities of Default: Econometric Estimation

### General Approach
- All PDs (and NPL ratios) estimated through time series modelling.
- PDs and NPL ratios range between [0, 1]; econometric models implemented on the logistic transformation:
  - LogitY = lp (Y / (1−Y))
- Econometric model used to simulate 10,000 alternative paths under both baseline and adverse scenarios; respective average paths used as projections.
- For EDF projections under the stress scenario, the path in the 90th percentile selected to better capture the impact of the adverse macroeconomic environment.

### NPL Ratios
- Historical household and corporate NPL ratios cointegrated; modeled together through a vector error correction model (VECM) with exogenous macrofinancial variables.
- VECM estimated on quarterly observations from Q4 2005 to Q4 2001 (65 observations). Optimal order zero (BIC).
- Exogenous variables: quarterly change in unemployment, quarterly change of real investment over real GDP ratio, quarterly log-difference of property prices, and dummy for structural break in Q4 2013.
- Selected VECM estimates (Table 1) include:
  - Constant (Consumer): -1.0455 (SE 0.2905; tStat -3.5993; pValue 0.0003)
  - Constant (Corporate): 0.0135 (SE 0.2707; tStat 0.0498; pValue 0.9603)
  - Adjustment (Consumer, Consumer): 0.0562 (SE 0.0154; tStat 3.6587; pValue 0.0003)
  - Impact (Consumer, Consumer): -0.4066 (SE 0.1111; tStat -3.6587; pValue 0.0003)
  - d Unemployment t (Corporate): 0.1645 (SE 0.0660; tStat 2.4921; pValue 0.0127)
  - d Investment to GDP t-1 (Consumer): -21.3959 (SE 9.8672; tStat -2.1684; pValue 0.0301)
  - d LN Property Prices t-1 (Consumer): -3.0129 (SE 1.1182; tStat -2.6944; pValue 0.0071)
  - dummy (Consumer): 0.5676 (SE 0.1083; tStat 5.2436; pValue 0.0000)
  - dummy (Corporate): 0.8951 (SE 0.1009; tStat 8.8722; pValue 0.0000)

### Corporate PDs (Country Models)
- Finland Corporate PD Model:
  - Estimated on quarterly observations from Q3 2001 to Q4 2021 (78 observations).
  - Optimal specification: ARIMA(1,0,0).
  - Dependent variable: logit transformation of Moody’s average Finnish corporate EDF.
  - Exogenous variables: annual real GDP growth, annual real investment growth, and EURIBOR.
  - Selected estimates (Table 2):
    - Constant: -0.8277 (SE 0.2376; tStat -3.4838; pValue 0.0009)
    - ARt-1: 0.8179 (SE 0.0548; tStat 14.9148; pValue 0.0000)
    - GDP growth t-3: -7.0611 (SE 3.3351; tStat -2.1172; pValue 0.0378)
    - Investment growth t-2: -0.6494 (SE 0.3146; tStat -2.0642; pValue 0.0427)
    - EURIBOR t-1: 3.1773 (SE 1.2758; tStat 2.4904; pValue 0.0151)
- Denmark Corporate PD Model:
  - Estimated from Q3 2006 to Q4 2021 (57 observations). Optimal specification ARIMA(0,1,0).
  - Exogenous variables: quarterly change of Denmark’s interbank rates, quarterly log-difference of Denmark’s real GDP, and quarterly change of Denmark’s output gap.
  - Selected estimates (Table 3):
    - d Interbank t-1: 0.1957 (SE 0.0878; tStat 2.2298; pValue 0.0301)
    - d LN GDP t-3: -8.2813 (SE 4.3703; tStat -1.8949; pValue 0.0637)
    - d Output gap t: -0.2053 (SE 0.0414; tStat -4.9536; pValue 0.0000)
- Norway Corporate PD Model:
  - Estimated from Q3 2001 to Q4 2021 (78 observations). Optimal specification ARIMA(0,1,0).
  - Exogenous variables: quarterly change of Norway’s interbank rates, and quarterly change of Norway’s output gap.
  - Selected estimates (Table 4):
    - d Interbank t-1: 0.2762 (SE 0.0543; tStat 5.0858; pValue 0.0000)
    - d Output gap t-1: -0.2617 (SE 0.0615; tStat -4.2578; pValue 0.0001)
- Sweden Corporate PD Model:
  - Estimated from Q3 2001 to Q4 2021 (78 observations). Optimal specification ARIMA(0,1,0).
  - Exogenous variables: quarterly change of Sweden’s interbank rates, quarterly log-difference of Sweden’s real GDP and quarterly change of Sweden’s output gap.
  - Selected estimates (Table 5):
    - d Interbank t-1: 0.1799 (SE 0.0468; tStat 3.8433; pValue 0.0003)
    - d LN GDP t-3: -3.9048 (SE 2.1997; tStat -1.7752; pValue 0.0800)
    - d Output gap t: -0.0867 (SE 0.0222; tStat -3.9064; pValue 0.0002)

### Finland Sovereign PD
- MuniFin only does government guaranteed lending; Finland’s sovereign PD used as PD for MuniFin’s lending.
- Historical Moody’s Finland sovereign EDF estimated through single time series model with exogenous variables.
- Estimated on quarterly observations from Q3 2001 to Q4 2021 (78 observations).
- Optimal specification: ARIMA(1,0,0).
- Dependent variable: logit transformation of the EDF.
- Exogenous variables: annual real GDP growth and quarterly change in output gap.
- Selected estimates (Table 6):
  - Constant: -0.4475 (SE 0.2366; tStat -1.8116; pValue 0.0634)
  - ARt-1: 0.9076 (SE 0.4577; tStat 19.027; pValue 0.0000)
  - GDP growth t-2: -3.68 (SE 2.4105; tStat -1.5267; pValue 0.1322)
  - d Output gap t: -0.0462 (SE 0.0309; tStat -1.4965; pValue 0.1399)

*From: 1finea2023003 - 2. Channels of*

### Appendix IV. Finland Corporate Probabilities of Default:

### Appendix IV. Finland Corporate Probabilities of Default: Econometric Estimation

### Aggregate PD combination methodology
- Combined aggregate corporate PD (PD_t_c) is a weighted average:
  - PD_t_c = w_a PD_t_a + w_b PD_t_b
  - Subject to w_a + w_b = 1
  - PD_t_a: PD derived through the NPL
  - PD_t_b: corporate EDF
  - Combination weights w_a and w_b estimated by linear programming

### Normalization and weight calibration
- Normalization of observed and predicted individual PDs:
  - A_i,t_ad = A_i,t / (1/T ∑_{t=1}^{T} A_i,t) 
  - A_t_mean = (1/I) ∑_{i=1}^{I} A_i,t_ad
  - F_i,t_ad = F_i,t / (1/T ∑_{t=1}^{T} A_i,t)
  - Where A_i,t is observed historical PD, F_i,t is predicted (fitted) historical PD for i∈{a,b}
- Weights estimated via linear programming minimizing sum of absolute deviations:
  - min_w ∑_{i=1}^{T} (ε_i1 + ε_i2)
  - s.t. ∑_i w_i F_i,t_ad − A_t_mean + ε_i1 − ε_i2 = 0 ∀ t∈[1,T]
  - ∑_i w_i = 1
  - w, ε1, ε2 ≥ 0

### Estimated combination weights
- NPL base projection weight: 0.56
- EDF based projection weight: 0.44

*Italic final source attribution: Appendix IV. Finland Corporate Probabilities of Default: Econometric Estimation*

### Appendix V. Interest Rates: Econometric Estimation

### Modeling approach
- Average interest rates of interest-bearing assets and liabilities projected via time-series regression models.
- Dependent variables: respective average interest rates (EURIBOR for lending and deposits, risk-free rate for debt securities).
- Models estimated on quarterly frequency, generally from Q2 2009 to Q4 2021 (46 observations), except debt securities (Q1 2015 to Q4 2021, 24 observations).

### Retail lending rates (ARIMA(1,1,0) with exogenous variables)
- Exogenous variables: quarterly change of EURIBOR and unemployment.
- Sample: Q2 2009 to Q4 2021 (46 observations).
- Table 1: Finland: Retail Lending Rate Model
  - Art-1: Estimate 0.2179; SE 0.0865; tStat 2.5188; pValue 0.0158
  - d EURIBOR t: Estimate 0.5073; SE 0.0608; tStat 8.3463; pValue 0.0000
  - d Unemployment t-2: Estimate 0.0010; SE 0.0054; tStat 1.7970; pValue 0.0797

### Corporate lending rates (VECM for large corporates and SMEs)
- Series cointegrated; modeled jointly via VECM. Optimal order: zero (BIC).
- Exogenous variables: quarterly change in EURIBOR and quarterly log-difference of equity prices.
- Sample: Q2 2009 to Q4 2021 (46 observations).
- Table 2: Finland: Corporate Lending Rate Model (selected coefficients)
  - Constant (Large): Estimate -0.4920; SE 0.1209; tStatistic -4.0684; pValue 0.0000
  - Constant (SME): Estimate 0.0626; SE 0.0382; tStatistic 1.6402; pValue 0.1010
  - Adjustment (Large, Large): Estimate -0.1810; SE 0.0393; tStatistic -4.6065; pValue 0.0000
  - Impact (Large, Large): Estimate -0.7301; SE 0.1585; tStatistic -4.6065; pValue 0.0000
  - Impact (Large, SME): Estimate 0.7595; SE 0.1649; tStatistic 4.6065; pValue 0.0000
  - d EURIBOR t (Large): Estimate 1.3901; SE 0.3985; tStatistic 3.4887; pValue 0.0005
  - d EURIBOR t (SME): Estimate 0.6588; SE 0.1258; tStatistic 5.2388; pValue 0.0000
  - d LN Equity Price t-3 (Large): Estimate -4.1771; SE 1.5602; tStatistic -2.6772; pValue 0.0074
  - d LN Equity Price t-3 (SME): Estimate -0.7054; SE 0.4924; tStatistic -1.4325; pValue 0.1520

### Interbank lending rate (ECM)
- Cointegrated with EURIBOR; estimated via single-equation ECM.
- Sample: Q2 2009 to Q4 2021 (46 observations).
- Table 3: Finland: Interbank Lending Rate Model
  - d EURIBOR t: Estimate 0.4675; SE 0.1056; pValue 0.0000
  - Lambda: Estimate -0.0340; SE 0.0179; pValue 0.0650
  - Error correction: Constant -0.1030; EURIBOR t-1 4.9769

### Deposit rates (term, overnight, other)
- Three single-equation time series models with d EURIBOR t as exogenous variable; autoregression included when required.
- Sample: Q2 2009 to Q4 2021 (46 observations).
- Table 4: Finland: Deposit Rates Model (selected coefficients)
  - ARt-1 (Overnight): Estimate -0.2783; SE 0.1301; pValue 0.0382
  - d EURIBOR t:
    - Term: Estimate 0.3508; SE 0.0724; pValue 0.0000
    - Overnight: Estimate 0.2940; SE 0.0376; pValue 0.0000
    - Other: Estimate 0.2658; SE 0.0820; pValue 0.0023

### Debt securities rates
- Three series: debt securities in assets, unsecured debt securities in liabilities, secured debt securities in liabilities.
- Exogenous variable: quarterly change of risk-free rate; secured debt includes d LN Property Prices t.
- Estimated via single-equation models; unsecured and secured liabilities cointegrated with risk-free rate and estimated as ECMs.
- Sample: Q1 2015 to Q4 2021 (24 observations).
- Table 5: Finland: Debt Security Rates Models (selected coefficients)
  - d Risk Free t:
    - Assets: Estimate 0.7819; SE 0.1497; pValue 0.0000
    - Liabilities Unsecured: Estimate 0.6059; SE 0.2780; pValue 0.0402
    - Liabilities Secured: Estimate 0.4941; SE 0.1125; pValue 0.0003
  - d LN Property Prices t (Secured): Estimate -2.2130; SE 1.4451; pValue 0.1406
  - Lambda (Assets): Estimate -0.1747; SE 0.0804; pValue 0.0409
  - Lambda (Secured Liabilities): Estimate -0.1200; SE 0.0861; pValue 0.0302
  - Error correction constants:
    - Liabilities Unsecured: Constant 0.1431
    - Liabilities Secured: Constant -0.1113
  - Risk Free t-1:
    - Liabilities Unsecured: 0.2777
    - Liabilities Secured: 0.1083

### Scenario treatment for debt securities in liabilities
- Baseline scenario: projections use the actual forecast of the model.
- Stress scenario: 10,000 alternative paths simulated; the path in the 90th percentile selected to capture expected increase in risk premia.

*Italic final source attribution: Appendix V. Interest Rates: Econometric Estimation*

### Appendix VI. Liquidity Stress Test Scenario Specification

### LCR Scenario weights and liquid asset haircuts (Table 1)
- Level 1 Assets: Basel 100 percent; Stressed 95 percent
  - Level 1 includes:
    1. Cash
    2. Qualifying marketable securities (sovereigns, central banks, PSEs, and MDBs)
    3. Qualifying central bank reserves
    4. Domestic sovereign or central bank debt for nonzero risk-weighted entities
- Level 2a Assets: Basel 85 percent; Stressed 50 percent
  - Level 2a includes:
    1. Qualifying marketable securities from sovereigns, central banks, PSEs, and MDBs (with 20 percent risk weighting)
    2. Qualifying corporate debt securities rated AA- or higher
    3. Qualifying covered bonds rated AA- or better
- Level 2b Assets:
  - Qualifying Mortgage-Backed Securities: Basel 75 percent; Stressed 50 percent
  - Qualifying corporate debt securities rated between A+ and BBB-: Basel 50 percent; Stressed 25 percent
  - Qualifying common equity shares: Basel 50 percent; Stressed 0 percent

### Inflows roll-off rates (Table 2) — Basel vs Stress
- Level 1 assets: Basel 0 percent; Stress 0 percent
- Level 1 assets (extremely liquid): Basel 7 percent; Stress 0 percent
- Level 2a assets: Basel 15 percent; Stress 5 percent
- Level 2b assets:
  - Eligible RMBS: Basel 25 percent; Stress 10 percent
  - Other: Basel 50 percent; Stress 30 percent
- Margin lending backed by all other collateral: Basel 50 percent; Stress 30 percent
- All other assets: Basel 100 percent; Stress 75 percent
- Credit or liquidity facilities: Basel 0 percent; Stress 0 percent
- Operational deposits held at other financial institutions: Basel 0 percent; Stress 0 percent
- Other inflows, by counterparty:
  - Retail counterparties: Basel 50 percent; Stress 25 percent
  - Nonfinancial wholesale counterparties, transactions not listed above: Basel 100 percent; Stress 20 percent
  - Other inflows from non-financial counterparties, non-principal repayment: Basel 50 percent; Stress 50 percent
  - Financial institutions and central banks, transactions not listed above: Basel 100 percent; Stress 50 percent
- Net derivative cash inflows: Basel 100 percent; Stress 50 percent
- Other (contractual) cash inflows: Basel 100 percent; Stress 50 percent
- Loans with an undefined contractual end date: Basel 20 percent; Stress 10 percent

### Outflows run-off rates (Table 3) — Basel vs Stress (selected items)
- Retail Deposits:
  - Demand deposits — Stable deposits: Basel 5 percent; Stress 20 percent
  - Demand deposits — Less stable retail deposits: Basel 10 percent; Stress 30 percent
  - Term deposits, residual maturity > 30d: Basel 0 percent; Stress 100 percent
  - Other forgone retail deposits: Basel 0 percent; Stress 100 percent
- Unsecured Wholesale Funding:
  - Demand and term deposits, residual maturity < 30d, small business — Stable deposits: Basel 5 percent; Stress 20 percent
  - ... — Less stable deposits: Basel 10 percent; Stress 30 percent
  - Operational deposits generated by clearing, custody, and cash management activities: Basel 25 percent; Stress 50 percent
  - Portion covered by deposit insurance: Basel 5 percent; Stress 5 percent
  - Cooperative banks in an institutional network: Basel 25 percent; Stress 25 percent
  - Nonfinancial corporates, sovereigns, central banks, MDBs, PSEs — Fully covered by deposit insurance: Basel 20 percent; Stress 20 percent
  - ... — Not fully covered by deposit insurance: Basel 40 percent; Stress 60 percent
  - Other legal entity customers: Basel 100 percent; Stress 100 percent
- Secured Funding:
  - Secured funding with a central bank, or backed by Level 1 assets: Basel 0 percent; Stress 0 percent
  - Secured funding backed by Level 2A assets: Basel 15 percent; Stress 20 percent
  - Secured funding backed by non-Level 1 or non-Level 2a asset (domestic sovereign, MDBs, or domestic PSEs as a counterparty): Basel 25 percent; Stress 50 percent
  - Funding backed by RMBS eligible for Level 2B: Basel 25 percent; Stress 50 percent
  - Funding backed by other Level 2B assets: Basel 50 percent; Stress 50 percent
  - Other secured funding transactions: Basel 100 percent; Stress 100 percent
- Additional Requirements (selected):
  - Valuation changes on non-Level 1 posted collateral securing derivatives: Basel 20 percent; Stress 20 percent
  - Excess collateral held by bank related to derivative transactions that could be called anytime: Basel 100 percent; Stress 100 percent
  - Liquidity needs related to collateral contractually due on derivatives transactions: Basel 100 percent; Stress 100 percent
  - Increased liquidity needs related to derivative transactions allowing collateral substitution: Basel 100 percent; Stress 100 percent
- ABCP, SIVs, conduits, SPVs, or similar:
  - Liabilities from maturing: Basel 100 percent; Stress 100 percent
  - Asset backed securities: Basel 100 percent; Stress 100 percent
- Undrawn but committed credit and liquidity facilities (Table 3 concluded):
  - Retail and small business: Basel 5 percent; Stress 50 percent
  - Nonfinancial corporates, sovereigns, central banks, MDBs, PSEs:
    - Credit facilities: Basel 10 percent; Stress 50 percent
    - Liquidity facilities: Basel 30 percent; Stress 50 percent
  - Supervised banks:
    - Credit facilities: Basel 40 percent; Stress 50 percent
    - Liquidity facilities: Basel 100 percent; Stress 100 percent
  - Other financial institutions:
    - Credit facilities: Basel 40 percent; Stress 50 percent
    - Liquidity facilities: Basel 100 percent; Stress 100 percent
  - Other legal entity customers, credit and liquidity facilities: Basel 100 percent; Stress 100 percent
- Other contingent funding liabilities:
  - Trade finance: Basel 5 percent; Stress 50 percent
  - Customer short positions covered by customers' collateral: Basel 50 percent; Stress 10 percent
  - Other product and services: Basel 75 percent; Stress 25 percent
  - Additional contractual outflows: Basel 100 percent; Stress 100 percent
  - Net derivative cash outflows: Basel 100 percent; Stress 100 percent
  - Any other contractual cash outflows (not listed above): Basel 100 percent; Stress 100 percent

### Cashflow analysis weights (Table 4)
- Time buckets and rates:
  - Up to 1 d:
    - Cash-Outflow Rate 60 percent
    - Cash-Inflows Rate 50 percent
    - Cumulative Assets Sales 30 percent
  - 1d to 1 wk:
    - Cash-Outflow Rate 60 percent
    - Cash-Inflows Rate 50 percent
    - Cumulative Assets Sales 70 percent
  - 1wk to 1m:
    - Cash-Outflow Rate 40 percent
    - Cash-Inflows Rate 50 percent
    - Cumulative Assets Sales 90 percent
  - 1m to 2ms:
    - Cash-Outflow Rate 40 percent
    - Cash-Inflows Rate 30 percent
    - Cumulative Assets Sales 100 percent
  - 2ms to 3ms:
    - Cash-Outflow Rate 40 percent
    - Cash-Inflows Rate 30 percent
    - Cumulative Assets Sales 100 percent
  - 3ms to 6ms:
    - Cash-Outflow Rate 40 percent
    - Cash-Inflows Rate 10 percent
    - Cumulative Assets Sales 100 percent
  - 6ms to a Y:
    - Cash-Outflow Rate 40 percent
    - Cash-Inflows Rate 10 percent
    - Cumulative Assets Sales 100 percent
  - 1y to 2ys:
    - Cash-Outflow Rate 20 percent
    - Cash-Inflows Rate 10 percent
    - Cumulative Assets Sales 100 percent

### Fire-sales haircuts (Table 5)
- Unencumbered Assets and Collateral:
  - Fire-Sales Market Haircut 20 percent
  - Counterbalancing Haircut 10 percent
- Other Eligible Securities:
  - Fire-Sales Market Haircut 50 percent
  - Counterbalancing Haircut 30 percent

*Italic final source attribution: Appendix VI. Liquidity Stress Test Scenario Specification*

### Appendix VII. Bond Yield Model Calibration

### Model specification
- Panel model where dependent variable: bank-specific 5-year bond yields.
- Independent variables:
  - Profitability: Return on assets (ROA)
  - Asset quality: Provisions over total credit exposures (POE)
  - Liquidity: Liquid assets over total assets (LOTA)
  - Risk-free rate: German sovereign bond yields
- Solvency measures tested but not significant.
- Data: unbalanced quarterly panel, cross-section N = 5 banks, time dimension ranges from 30 to 47; total observations 175.
- Longest series range: Q3 2010 to Q1 2022.
- Model implemented in first differences; all variables seasonally adjusted.

### Estimated coefficients (Table 1: Finland: Panel Model Estimation)
- ROA t: Estimate -3.4671; SE 1.3745; tStat -2.5224; pValue 0.0126
- Log POE t: Estimate 0.9483; SE 0.3189; tStat 2.9739; pValue 0.0034
- LOTA t: Estimate -0.0038; SE 0.0020; tStat -1.8738; pValue 0.0627
- Risk-Free t: Estimate 0.9817; SE 0.1428; tStat 6.8733; pValue 0.0000

*Italic final source attribution: Appendix VII. Bond Yield Model Calibration*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2023/english/1finea2023003.pdf_