## cr1706 — Finland FSAP Stress Testing Executive Summary

## Source details

**Canonical URL:** [cr1706 — Finland FSAP Stress Testing Executive Summary](https://www.imf.org/-/media/files/publications/cr/2017/cr1706.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2017/cr1706.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2017/cr1706.pdf.json)

---

### Major vulnerabilities and near-term risks
- Despite high capitalization levels, important vulnerabilities exist in the Finnish banking system.
- Near-term risks largely tilted to the downside from external and domestic sources.
- Profitability negatively affected by the low interest rate environment and low economic growth, despite being high compared to the rest of the euro area.
- Vulnerabilities highlighted:
  - Funding risks, contagion risks, and challenges related to long-term profitability.
  - Relatively low levels of total equity resulting in part from aggressive use by banks of their internal risk models.
  - Significant exposures by some banks to the derivatives business and consequent counterparty credit risks.

### Size, concentration, and funding structure
- Total banking sector assets amount to 230 percent of GDP.
- The top four banks control 93 percent of system assets.
- Subsidiaries of two foreign banks control 65 percent of total assets.
- A domestic private-owned bank makes up 21 percent of banking system assets.
- A domestic state-owned bank accounts for 6½ percent of assets.
- Wholesale funding and maturity transformation:
  - Loan-to-deposit ratio: 124 percent (and elsewhere reported 129 percent in another table).
  - Wholesale funding accounts for more than half of banks’ funding.
  - Main wholesale instruments: deposits from credit institutions, covered bonds, senior unsecured bonds, repo transactions, and commercial paper/time deposits.
  - A majority of wholesale funding is provided by foreign financial institutions, including parent banks.

### Stress test design, scope, and methodology
- Objective: assess capacity to withstand extreme but plausible macroeconomic shocks; explore weaknesses and transmission channels; identify priorities for policy actions and informational/methodological gaps.
- Coverage and sample:
  - TD stress test covered the four largest significant institutions, representing 93 percent of MFI assets and 84 percent of the banking sector retail deposits; included two subsidiaries of foreign banks.
- Components:
  - Top-Down (TD) macro tests (2016-2018 horizon, data through December 2015) using satellite models to map macro scenarios into PDs, LGDs, EADs, income and RWAs.
  - Sensitivity tests (static shocks).
  - Household Stress Test using HFCS micro data.
  - CVA Stress Test using banks’ data.
  - Liquidity tests: LCR-type, NSFR-type, cash flow maturity ladder, reverse liquidity sensitivity.
  - Contagion analyses: domestic interbank, banks–insurers, cross-border, and market-data-based network model (Diebold–Yilmaz style).
- Key modelling and projection rules:
  - Banks’ balance sheets projected to grow in line with nominal GDP; RWAs constrained not to decrease over the stress period.
  - Non-interest income and expenses projected to grow in line with nominal GDP with a floor of zero for operational expenses.
  - Funding-costs projected from a panel fixed-effects OLS model (annual changes 1999–2015): annual change in Euribor rate coefficient 0.6217*** (t-statistic 4.87); annual change in Finnish 10-year government bond rate 0.2262* (t-statistic 2.20); annual change in total funding -0.0107** (t-statistic -3.99); Constant 0.1919** (t-statistic -3.54); R-square 0.78; # of observations: 40.
  - Dividend distribution: undercapitalized banks not allowed to distribute; banks passing minima distribute according to official plans as of end-Q4 2015.

### Macroeconomic scenarios (3-year horizon: 2016–2018; first year of shock 2016)
- Severe stress scenario: large external and domestic shocks; includes a 3-notch sovereign downgrade; Finland-specific severe stress GDP growth path: -5.0, -3.1 and +0.2 percent in 2016, 2017 and 2018 respectively.
- Moderate stress scenario: milder euro-area and external shocks; one-notch sovereign downgrade.
- Key macro variable projections (Est. in % 2015; values for 2016, 2017, 2018 shown exactly as in source)
  - Real GDP growth
    - Baseline: 0.4 0.9 1.2 1.3
    - Severe stress: 0.4 -5.0 -3.1 0.2
    - Moderate stress: 0.4 -1.8 -1.1 1.3
  - Real investment growth
    - Baseline: -2.1 2.6 3.6 3.4
    - Severe stress: -2.1 -17.0 -15.8 -4.0
    - Moderate stress: -2.1 0.2 -0.9 -2.1
  - CPI Inflation rate
    - Baseline: 0.0 1.3 1.6 1.8
    - Severe stress: 0.0 -0.2 -1.4 -2.4
    - Moderate stress: 0.0 0.6 -0.9 -1.3
  - Unemployment rate
    - Baseline: 9.5 9.5 9.1 8.6
    - Severe stress: 9.5 15.4 16.3 17.0
    - Moderate stress: 9.5 10.5 10.7 10.0
  - Nominal annual interest rate (10-year government bond rate)
    - Baseline: 0.7 1.0 1.1 1.3
    - Severe stress: 0.7 3.4 2.8 2.6
    - Moderate stress: 0.7 1.4 1.5 1.6
  - House price annual change
    - Baseline: -0.5 0.0 -0.1 0.1
    - Severe stress: -0.5 -15.0 0.0 0.0
    - Moderate stress: -0.5 -0.2 -0.9 0.0

### Solvency stress-test results and capital implications
- Under a fully-loaded Basel III basis, severe scenario effects:
  - Aggregate CET1 ratio would drop by 14 percentage points compared with the starting point.
  - CET1 would be 15 percentage points below the baseline in 2018.
  - Main drivers of CET1 decline (contributions over stress horizon):
    - Funding costs: -6.4 percentage points of RWAs.
    - Loss provisions: -6.1 percentage points of RWAs.
    - Changes in RWAs: -2.8 percentage points of RWAs due to shocked credit parameters.
  - Under severe stress:
    - Capital of two of the four largest banks would fall slightly below minimum regulatory Capital Adequacy Requirements.
    - Recapitalization needs: around EUR550 mn or 0.3 percent of GDP (CAR-based).
    - When measured by leverage ratio, all four banks would fail to meet the 3 percent hurdle rate in 2018; recapitalization needs based on leverage ratio equivalent to 2 percent of GDP.
    - Aggregate leverage ratio (four largest banks) would drop from 4.2 percent in 2015 to 2.3 percent at end-2018.
    - Number of undercapitalized banks (CET1<4.5%): 2; Number of undercapitalized banks (leverage ratio<3%): 4.
- Interpretation: discrepancy between RW-based and leverage-based outcomes indicates banks’ internal models may underestimate risks.

### Household sector stress and retail credit amplification
- Household stress-test methodology: HFCS micro data (representative sample ~10900 households, updated to 2015) with income/expenditure shock mechanics and default rule PD = 1 if IM<0 and NA<0.
- Key household stress outcomes:
  - Pre-shock mortgage portfolio implied PD: 3.9 percent.
  - Severe stress scenario implied mortgage PD: 13.6 percent.
  - Counterfactual without decline in wages and social transfers: average PD 7.6 percent.
  - Implied household PDs rise from 3.9 percent to almost 13.6 percent under a scenario of a significant drop in incomes combined with a 15 percent shock to real estate prices and without any decline in average expenditure (multiplier 3.5).
- Credit losses and provisions (three-year horizon):
  - Severe stress scenario credit losses in the loan book: EUR6.9 bn, equivalent to 1.5 percent of total banking system assets.
  - Baseline scenario credit losses: EUR0.8 bn, equivalent to 0.2 percent of total banking system assets.
  - Capital needs in severe stress driven by losses and an almost 10 percent increase in RWAs driven by higher credit risks.

### Credit risk satellite-model key inputs and sensitivities
- Asset composition:
  - Total loans: 42 percent of total banking sector assets.
  - Debt securities: 22 percent.
  - Derivatives: close to 19 percent.
- IRB/STA distribution:
  - 47 percent of loan portfolio under STA; 76 percent of credit risk RWAs under IRB.
- PD construction:
  - PiT and TTC PDs used; hybrid PDs approximated as 30 percent PIT + 70 percent TTC.
  - System average PDs:
    - 2015 system average PD: 1.9 percent.
    - Severe stress: system average PD increases to 6 percent in 2018.
    - Moderate stress: peak PD of 3.9 percent.
- Quantified PD sensitivities (preserved exactly):
  - Mortgage: unemployment +1 percentage point → PD +0.6 percentage points; GDP growth -1 percentage point → PD +0.2 percentage points.
  - Non-financial corporates: GDP growth -1 percentage point → PD +0.1 percentage points; Euribor +100 basis points → PD +0.2 percentage points.
  - SMEs: unemployment +1 percentage point → PD +0.7 percentage points.
- LGD pass-through assumptions:
  - Baseline: No changes.
  - Moderate stress: 50% pass-through from real estate price shock to LGD.
  - Severe stress: 100% pass-through.

### CVA, derivatives, market and concentration risks
- CVA stress test:
  - Exposures grouped by country segments and STA/AM categories; average shock to RWAs across tested banks: 30 percent.
  - Result would be 4 to 5 times higher if all CRR-exempt exposures were included.
- Derivatives and cross-border concentration:
  - Derivatives constitute a large share of cross-border exposures: 42 percent of cross-border claims and 36 percent of liabilities.
  - Nordea’s derivatives concentration in Nordea Bank Finland increased cross-border exposure to UK clearing houses; planned conversion to branch (2017) will change location of exposure.
- Market risk sensitivity results:
  - Interest-rate repricing gap (below one year) aggregated: EUR -1.9 bn.
  - Severe scenario direct interest-rate loss: EUR134 mn (no material effect on CAR).
  - Potential valuation losses on own-sovereign debt in severe stress: EUR293 mn, contributing 0.3 percentage points to CAR decline.
  - FX exposure: net open FX position EUR130 mn (0.03 percent of assets, 0.6 percent of Tier 1 capital); severe scenario FX loss EUR4 mn.
- Concentration risk (largest exposures, extreme zero recovery):
  - Default of the largest exposure of each bank → capital shortfall EUR268 mn (0.1 percent of GDP) relative to 8.5 percent Tier 1 hurdle.
  - Default of five largest exposures per bank → total Tier 1 shortfall EUR1.5 bn (0.7 percent of GDP).
  - Default of ten largest exposures per bank → total Tier 1 shortfall EUR2.3 bn (1.1 percent of GDP).

### Liquidity and funding stress-test results
- Funding structure (Dec 2015, excluding capital and including contingent credit lines):
  - Retail sight and time deposits: 45 percent of funding.
  - Secured wholesale funding: 28 percent.
  - Unsecured wholesale funding: 27 percent (credit institutions 20 percent plus operational deposits).
  - Undrawn committed credit/liquidity facilities: 8 percent.
  - Foreign-currency liabilities: 17 percent of total liabilities.
- LCR-based stress-test outcomes (aggregate/system-wide):
  - Aggregate LCR at Dec 2015: 124 percent (baseline).
  - Adverse Scenario 1 (retail shock): Aggregate LCR → 84 percent; liquidity shortfall EUR17 bn; shortfall expressed in GDP: 8 9 percent of GDP.
  - Adverse Scenario 2 (dry-up unsecured wholesale funding): Aggregate LCR → 66 percent; shortfall EUR45 bn; shortfall in GDP: 22 percent of GDP.
  - Combined shock: System-wide LCR → 61 percent; liquidity shortfall EUR57 bn; shortfall in GDP: 27.5 percent of GDP.
  - Foreign-currency LCR shortfalls: EUR1.1 bn to EUR3 bn (0.5 percent to 1.4 percent of GDP).
  - Reverse liquidity test (retail withdrawals only): to bring aggregate liquidity ratio below 100 percent, withdrawal rates would need to reach 30 percent for guaranteed retail deposits in a month and 40 percent for uncovered retail deposits in a month.
- NSFR-based stress test:
  - Available stable funding: EUR192.2 bn; Required stable funding: EUR168.7 bn; Aggregate NSFR: 113.9 percent.
  - Liquidity mismatches at the one-year horizon: EUR25.4 bn (12.2 percent of GDP); some individual banks show shortfalls.
- Outflow analysis (maturity buckets): 16–30 days bucket gap: EUR1.7 bn (0.8 percent of GDP).
- Overall: funding and liquidity risks are the main vulnerability; material exposure to funding risks in foreign currencies for some banks.

### Contagion and cross-border interconnectedness
- Domestic contagion:
  - Direct domestic contagion via bilateral exposures limited; no single domestic bank failure would trigger another bank’s failure via direct domestic interbank exposures among the four largest banks.
  - Bank–insurer interlinkages: contagion risks very limited; no cascade effects identified.
- Cross-border linkages and vulnerabilities:
  - Cross-border exposures between credit institutions account for 63 percent of total cross-border exposures, mainly intra-group within Nordic banking groups.
  - Significant cross-border derivatives concentration and reliance on UK CCPs.
  - Market-data implicit-linkages (Diebold–Yilmaz): Sweden and Finland share the strongest implicit linkages; Finland most strongly connected to developed economies in continental Europe among Nordics.
  - Cross-border funding shocks: a 34 percent contraction of funding from one country would deplete one bank’s liquid assets; a 39 percent unmet payment share from one country would cause Tier 1 minima breach for the most exposed banks.
- Policy implication: strong Nordic linkages warrant regional coordination and monitoring.

### Key policy implications and recommended actions
- Regulatory and supervisory actions:
  - Review banks’ Internal Ratings-Based (IRB) models and finalize plans to introduce regulatory floors for risk-weights calculated by IRB banks.
  - Increase monitoring of liquidity, funding, and derivatives-related risks; examine banks’ liquidity positions in foreign currencies.
  - Use regular micro-data based household stress tests to benchmark PDs and LGDs for retail portfolios and to challenge aggressive modeling techniques.
  - Perform liquidity stress tests based on cash flows at various maturities and apply customized LCR along the maturity ladder.
  - Discourage cross-ownership of covered bonds.
  - Collect additional data on cross-border derivatives activities with breakdown between centrally-cleared and over-the-counter transactions.
  - Lead a Nordic-wide coordinated stress test exercise to account for interlinkages and spillovers, include liquidity–solvency interactions, and expand coverage beyond banks to cross-border regional insurance companies and financial market infrastructure.
- Other recommendations:
  - High vigilance given structural strong interconnectedness within the Nordic region; consider expanding protocols for formal region-wide sharing of supervisory data.

### Key numerical design and parameter highlights (preserved exactly)
- Market share covered: 93 percent of the MFI assets.
- Stress test horizon: 3 years (2016/2018).
- Multiplier a for updating TTC PDs with shocked PIT PDs: 10%.
- Pre-shock system average PD: 1.9 percent; Severe stress system average PD in 2018: 6 percent.
- Household mortgage PD: pre-shock 3.9 percent → severe stress 13.6 percent.
- Credit losses (three-year): Severe stress EUR6.9 bn (1.5 percent of assets); Baseline EUR0.8 bn (0.2 percent of assets).
- Aggregate LCR (Dec 2015) baseline: 124 percent; Adverse Scenario 1 LCR: 84 percent (EUR17 bn shortfall); Adverse Scenario 2 LCR: 66 percent (EUR45 bn shortfall); Combined shock shortfall EUR57 bn (27.5 percent of GDP).
- Aggregate NSFR: 113.9 percent; Liquidity mismatches one-year horizon: EUR25.4 bn (12.2 percent of GDP).
- Leverage ratio: 2015 start 4.2 percent → 2.3 percent end-2018 (severe); leverage hurdle 3 percent binding from 2018.
- Recapitalization needs:
  - CAR-based: EUR550 mn (0.3 percent of GDP).
  - Leverage-based: equivalent to 2 percent of GDP.

*Italic: IMF staff report as extracted from cr1706.*

### EXECUTIVE SUMMARY __________________________________________________________________________ 6

### EXECUTIVE SUMMARY

### Major vulnerabilities and near-term risks
- Despite high capitalization levels, important vulnerabilities exist in the Finnish banking system.
- Near-term risks are largely tilted to the downside and stem from both external and domestic sources.
- Profitability is negatively affected by the low interest rate environment and low economic growth, despite being high compared to the rest of the euro area.
- Vulnerabilities highlighted:
  - Funding risks, contagion risks, and challenges related to long-term profitability.
  - Relatively low levels of total equity resulting in part from an aggressive use by banks of their internal risk models.
  - Significant exposures by some banks to the derivatives business and consequent counterparty credit risks.

### Size, concentration, and funding structure
- Total banking sector assets amount to 230 percent of GDP.
- The top four banks control 93 percent of system assets.
- Subsidiaries of two foreign banks control 65 percent of total assets.
- A domestic private-owned bank makes up 21 percent of banking system assets.
- A domestic state-owned bank accounts for 6½ percent of assets.
- Wholesale funding plays a major role:
  - Loan-to-deposit ratio is 124 percent.
  - Wholesale funding accounts for more than half of banks’ funding.
  - Main wholesale instruments: deposits from credit institutions, covered bonds, senior unsecured bonds, repo transactions, and commercial paper/time deposits.
  - A majority of wholesale funding is provided by foreign financial institutions, including parent banks.

### Stress test design and scenarios
- Stress tests examined resilience to solvency, liquidity, and contagion risks via top-down (TD) macroeconomic scenarios and sensitivity analyses.
- Macroeconomic scenario tests assessed impacts over a three-year horizon (2016-2018), based on data available through December 2015.
- Tests used satellite models and methodologies developed by Fund staff for projecting effects on individual banks’ profitability and capitalization.
- TD liquidity tests used a maturity ladder analysis and supervisory information, on an aggregate basis and by currencies.
- Contagion tests covered domestic interbank exposures, cross-exposures between domestic banks and insurers, and cross-border interlinkages with supervisory and market data.
- External risks incorporated include: protracted weak growth in advanced economies (particularly the euro area), substantial declines in house and other asset prices in Nordic neighbors, foreign banks’ withdrawal of liquidity, and a surge in global financial market volatility driving higher interest rates and funding costs.
- Domestic risks incorporated include: domestic confidence shock causing consumption and investment collapse, decline in wages, reduction in public safety net, house price declines, and a sovereign stress scenario involving a 3-notch downgrade of the sovereign rating.

### Solvency stress test results and capital implications
- The severe stress scenario combines extreme but plausible macrofinancial shocks and significantly affects the banking system.
- Key drivers of solvency stress: funding risks from high reliance on unsecured wholesale funding; credit risk in domestic retail portfolios; high share of volatile trading income.
- Under the severe stress scenario:
  - Capital of two of the four largest banks would fall slightly below minimum regulatory Capital Adequacy Requirements.
  - Recapitalization needs would be around EUR550 mn or just 0.3 percent of GDP.
  - When considering total equity in relation to assets (leverage ratio), all four banks would fail to meet the 3 percent hurdle rate in 2018 when it becomes binding. Recapitalization needs based on the leverage ratio would be equivalent to 2 percent of GDP.
  - The discrepancy between RW-based capital outcomes and leverage-based outcomes indicates that banks’ internal models used to calculate risk weighted assets may underestimate risks.

### Household sector stress and credit risk amplification
- Solvency results are largely driven by a significant deterioration in households’ situation.
- Current low credit risks reflect the macroeconomic environment and social set-up in which private consumption and households’ disposable income are actively sheltered from economic shocks.
- In an economic downturn, credit risks can remain low if government/social safety nets continue to protect households; however, a substantial reduction in wages and social transfers may lead to a large increase in retail portfolio losses.
- Capital needs in the severe stress scenario arise from losses and an almost 10 percent increase in risk-weighted assets driven by higher credit risks.

### Liquidity and funding stress test results
- Funding and liquidity risks are the main vulnerability.
- Global liquidity stress tests show most banks would be exposed to liquidity risks under:
  - Large deposit withdrawals,
  - A scenario more severe than Basel III LCR metrics,
  - A dry-up of unsecured wholesale funding.
- Some banks display material exposure to funding risks in foreign currencies.
- NSFR results: maturity mismatches are moderate at the aggregate one-year horizon, but at the individual level some banks would experience liquidity shortfalls reflecting excessive maturity transformation.

### Contagion and cross-border interconnectedness
- Direct domestic contagion risks through bilateral exposures are limited.
- Contagion analysis finds very limited risks from domestic interbank exposures and from cross-exposures between domestic banks and insurance companies.
- Cross-border bank linkages are a concern and require monitoring:
  - The Finnish banking system is vulnerable to cross-border spillovers due to Nordic region linkages.
  - Subsidiaries of foreign banks would be vulnerable to failure of their parents on their asset side.
  - Market data-based analysis confirms strong implicit financial linkages among the Nordics: Sweden and Finland share the strongest implicit linkages, and Finland is the Nordic country most strongly connected to developed economies in continental Europe.

### Policy implications and recommended actions
- Findings may warrant a regulatory review of banks’ Internal Ratings-Based (IRB) models and increased monitoring of liquidity, funding, and derivatives-related risks.
- Recent FSA steps to introduce regulatory floors on risk-weights for mortgage exposures for IRB banks will increase capital buffers to cover unexpected losses due to potential system-wide vulnerabilities.
- Additional recommended actions include:
  - Use regular micro-data based household stress tests to benchmark PDs and LGDs for retail portfolios and to challenge banks if aggressive modeling techniques lower risk parameters.
  - Finalize plans to introduce regulatory floors for risk-weights calculated by IRB banks.
  - Examine banks’ liquidity positions in foreign currencies in stressed environments and perform liquidity stress tests based on cash flows at various maturities.
  - Discourage cross-ownership of covered bonds.
  - Lead a Nordic-wide coordinated stress test exercise to account for interlinkages and spillovers, include liquidity-solvency interactions, and expand coverage beyond banks to cross-border regional insurance companies and financial market infrastructure.
  - Collect additional data on cross-border derivatives activities with a breakdown between centrally-cleared and over-the-counter transactions.

*FINLAND — INTERNATIONAL MONETARY FUND: EXECUTIVE SUMMARY*

### 4.      In general, the objective of the FSAP stress testing exercise is to assess the capacity of

### 4.      In general, the objective of the FSAP stress testing exercise is to assess the capacity of 

### Overview
- Objective: assess the capacity of the banking system to withstand extreme but plausible macroeconomic shocks; explore weaknesses and transmission channels; identify priorities for policy actions and informational/methodological gaps.
- Stress tests examine solvency, liquidity, and contagion risks, including a Top-Down (TD) exercise based on macroeconomic scenarios and sensitivity analyses.
- Time horizon for the TD macroeconomic scenario tests: three-year horizon (2016-2018), based on data available through December 2015.

### Scope and Coverage
- TD stress test covered the four largest significant institutions, representing:
  - 93 percent of monetary and financial institutions (MFI) assets
  - 84 percent of the banking sector retail deposits
- Included two subsidiaries of foreign banks.
- Due to restricted access to supervisory data, TD exercise focused on these four systemically important institutions.

### Stress Test Components and Methodology
- Tests included:
  - Top-Down macro tests: external and domestic shocks; forecasts of credit losses and other sources of profit and losses based on satellite models.
  - Sensitivity tests: domestic shocks.
  - Household Stress Test based on micro data and macro scenarios (HFCS).
  - CVA Stress Test based on data provided by banks.
- Liquidity tests included:
  - LCR-type liquidity stress test with different variants
  - NSFR-type liquidity stress test
  - Cash flow-based liquidity stress test using maturity buckets
  - Reverse liquidity sensitivity test
- Contagion analyses covered:
  - Pure interbank, banks-insurers and cross-border contagion models
  - Market-data-based network model inspired by Diebold-Yilmaz (2012) connectivity methodology
- Satellite models: map variables projected in macro scenarios into credit factors that determine individual banks’ gains or losses; link credit risk parameters (PDs) with various macrofinancial variables.
- TD liquidity tests used a maturity ladder analysis (cash flow-based with different maturity buckets) and supervisory information.
- Contagion tests covered domestic interbank exposures, interlinkages within the domestic financial system, and cross-border exposures between Finnish banks and foreign sectors.

### Key Vulnerabilities and Features of the Banking Sector
- Capitalization:
  - Equity-to-total assets ratio: 4.5 percent
  - Regulatory capital ratio: 21.7 percent of risk-weighted assets
  - Risk density (RWA to total assets ratio) of the four largest banks: 21.4 percent in 2015 (compared with 39 percent among euro area peers)
  - Credit risk RWA to gross loans ratio: 42 percent
- Liquidity and funding:
  - Loan-to-deposit ratio: 129 percent
  - Reliance on wholesale funding, in particular from external sources
- Cross-border risks:
  - Substantial financial and economic integration amongst Nordic countries
  - Cross-ownership of covered bonds widespread in the Nordic region
  - Investors in covered bonds are mostly banks, central banks and funds from the rest of the euro area
  - Interest rate risk in covered bonds: mortgage loans in cover pools primarily floating-rate while covered bonds issued as fixed-rate; banks hedge via derivatives

### Macroeconomic Scenarios and Assumptions
- All scenarios span three years (2016-2018); first year of shock is 2016.
- Two adverse scenarios (plus baseline):
  - Moderate stress scenario: risk aversion affecting the European “periphery”, adverse investment sentiment, sharp slowdown in emerging market economies; Finnish growth experiences a triple dip; one-notch downgrade of the sovereign rating.
  - Severe stress scenario: severe disruption in Finland’s European partners (large declines in house and other asset prices in Nordic neighbors) amplified by domestic shocks; size of overall shock larger than global financial crisis but in line with mid-1990s Nordic banking crisis; specific shocks include domestic demand confidence shock, decline in residential real estate prices, a 3-notch sovereign rating downgrade, procyclical fiscal consolidation; results include a 2.7 per cent per annum decline in GDP on average, significantly higher unemployment, decline in wages and social transfers, and persistent deflationary period.

- Key macroeconomic variable projections and scenario values (Est. in % 2015; values for 2016, 2017, 2018 shown exactly as in source):
  - Real GDP growth
    - Baseline: 0.4 0.9 1.2 1.3
    - Severe stress: 0.4 -5.0 -3.1 0.2
    - Moderate stress: 0.4 -1.8 -1.1 1.3
  - Real investment growth
    - Baseline: -2.1 2.6 3.6 3.4
    - Severe stress: -2.1 -17.0 -15.8 -4.0
    - Moderate stress: -2.1 0.2 -0.9 -2.1
  - CPI Inflation rate
    - Baseline: 0.0 1.3 1.6 1.8
    - Severe stress: 0.0 -0.2 -1.4 -2.4
    - Moderate stress: 0.0 0.6 -0.9 -1.3
  - Unemployment rate
    - Baseline: 9.5 9.5 9.1 8.6
    - Severe stress: 9.5 15.4 16.3 17.0
    - Moderate stress: 9.5 10.5 10.7 10.0
  - Nominal annual interest rate (10-year government bond rate)
    - Baseline: 0.7 1.0 1.1 1.3
    - Severe stress: 0.7 3.4 2.8 2.6
    - Moderate stress: 0.7 1.4 1.5 1.6
  - House price annual change
    - Baseline: -0.5 0.0 -0.1 0.1
    - Severe stress: -0.5 -15.0 0.0 0.0
    - Moderate stress: -0.5 -0.2 -0.9 0.0

### Hurdle Rates and Capital Framework
- Regulatory frameworks applied: Basel III, the European Union and the national frameworks, as defined by CRD IV, Regulation on Prudential Requirements, national law and FIN-FSA regulation.
- Hurdle rates set according to current minimum requirements; leverage ratio expected to become binding from 2018 onwards.
- Banks allowed to deplete a capital conservation buffer of 2.5 percent in the severe stress scenario.
- Ratios based on Basel III fully-loaded definitions; tests based on minimum capital ratios under Pillar I and did not take into account individual Pillar II requirements.
- Table of scenario hurdle rates and actual ratios (values preserved exactly):
  - Total Capital ratio (total capital to RWAs, in percent)
    - Baseline and Moderate Stress scenarios hurdle rate (including capital conservation buffer): 10.5 8 24.4
  - Tier I Capital ratio (Tier 1 capital to RWAs, in percent)
    - Severe Stress scenario hurdle rate (excluding capital conservation buffer): 8.5 6 23
  - Common Equity Tier I Capital ratio (CET1 capital to RWAs, in percent)
    - Actual ratio before stress: 7 4.5 21.9
  - Leverage ratio (Tier 1 capital to total assets, in percent)
    - 3 3 4.2

(Note: The table in the source presents columns titled "Scenario Baseline and Moderate Stress scenarios hurdle rate (including capital conservation buffer)", "Severe Stress scenario hurdle rate (excluding capital conservation buffer)", and "Actual ratio before stress". The numeric values above are preserved exactly as presented in the source.)

### Household Stress Test
- Conducted to assess vulnerabilities of Finnish households to significant drops in income, real estate prices and social transfers.
- Used household data from the Households Finance and Consumption Survey (HFCS) collected by the ECB.

### Interpretation and Limitations
- Stress tests use macroeconomic and satellite models estimated with historical data and are subject to estimation uncertainty.
- Limitations can be mitigated but not eliminated by state-of-the-art techniques.
- Choices in design include severity of shocks and selection of relevant historical episodes and data series length.
- Adverse scenarios reflect combinations of external and domestic shocks that ex ante have a very low probability of occurrence and should not be interpreted as macroeconomic “forecasts”.

*Prepared by Mrs. Kay Chung, Messrs. Mindaugas Leika and Cyril Pouvelle, with the collaboration of Messrs. Luis Brandao-Marques and Benjamin Huston, all Monetary and Capital Markets Department of the IMF.*

### 18.      In all the scenarios, a number of adjustments and assumptions were made to track the

### cr1706 - 18.      In all the scenarios, a number of adjustments and assumptions were made to track the

### Assumptions on balance sheet growth and projection mechanics
- Growth of banks’ balance sheets: projected to grow in line with nominal GDP so that the size of the banking system in terms of assets remains constant relative to the size of the economy.  
- Rationale: ensures banks that pass the tests remain sufficiently capitalized to support lending in a severe downturn and reduces the need to quantify second round effects from banks’ behavioral responses.  
- Note on net assets: in adverse scenarios, growth of net assets (total assets net of loan loss provisions) is usually lower than growth of total assets because provisions are higher.  
- Quasi-dynamic adjustment: banks’ balance sheets in the stress test model can shrink if credit growth is negative; the limit of deleveraging is aligned with the negative nominal GDP growth rate.

### Projection of risk-weighted assets (RWAs) and regulatory approach
- For banks operating under Basel II Internal Rating-Based (IRB) approach: risk weights are projected using the corresponding Basel II formula for credit risk where the capital requirement ratio depends on PD, LGD and asset correlation. This typically implies risk weights rise in a stress scenario.  
- Constraint: RWAs are exogenously constrained not to decrease over the stress period.  
- For exposures under the Basel II standardized approach (STA): RWAs were calculated assuming standardized risk weights and exposure migration from lower to higher risk weights during the stressed period.

### Evolution of profits and income projection rules
- Non-interest profit items and lines such as operational and administrative expenses, and net fee and commission income: projected to grow in line with nominal GDP.  
- Floor: zero set for the growth of operational and administrative expenses.  
- Extraordinary income: assumed not to recur during the 2016-2018 period in the baseline and adverse scenarios.  
- Non-performing loans: assumed to not provide any accrued income.  
- Lending rates: assumed to evolve in line with changes in the Euribor rate.

### Funding costs, interbank deposits and limitations
- Banks’ funding costs: projected based on an econometric OLS panel model with fixed effects estimating the annual change in Finnish banks’ average interest expense to total funding ratio, using annual data over 1999-2015.  
- Explanatory variables in the funding-cost model: the lagged dependent variable; the annual change in the 3-month money market interest rate; the annual change in the 10-year Finnish government bond rate; and the annual change in the amount of total funding in percent.  
- Finding: the change in the interbank rate was the variable with the most significant effect on the change in Finnish banks’ funding costs; the largest contribution in the severe stress scenario comes from the increase in the Finnish 10-year government bond rate due to its large increase.  
- Application: increase in funding costs applied to banks’ liabilities subject to funding shocks, including parent funding.  
- Two limitations noted: (i) absence of differentiation of the debt instruments by type and maturity; (ii) limited number of observations (40).  
- Projection of interbank deposits: aligned with the parameter set for the liquidity stress test to integrate liquidity stress test parameters and solvency stress test results.

### Dividend distribution rules in scenarios
- Distribution rules for after-tax profits:  
  i) Undercapitalized banks in any year of a given scenario were not allowed to distribute dividends;  
  ii) Banks that pass all minimum requirements distribute their dividends according to the official plans (as publicly announced by the end of Q4 2015).

### Estimated funding-cost model results (Table 4)
- Dependent variable: annual change in ratio of banks’ interest expenses to total funding (panel fixed-effects OLS model).  
- Coefficients and statistics preserved exactly as reported:  
  - annual change in Euribor rate: 0.6217*** (t-statistic 4.87)  
  - annual change in Finnish 10-year government bond rate: 0.2262* (t-statistic 2.20)  
  - annual change in total funding (in percent): -0.0107** (t-statistic -3.99)  
  - Constant: 0.1919** (t-statistic -3.54)  
  - R-square: 0.78  
  - # of observations: 40  
- Significance notation: * Denotes significance at the 10 percent level; ** at the 5 percent level; and *** at the 1 percent level.  
- Caution: obtaining a robust estimate for funding costs would require much longer time series and more observations; results of satellite models should be interpreted with appropriate care.

### Summary of scenario-specific modeling choices (Table 5 highlights)
- PDs: Satellite models used across Baseline, Moderate stress, Severe stress.  
- LGDs:  
  - Baseline: No changes.  
  - Moderate stress: 50% pass-through effect from the shock to real estate prices.  
  - Severe stress: 100% pass-through effect from the shock to real estate prices.  
- Growth rates of interest-bearing assets (credit growth), non-interest bearing assets, open positions: In line with nominal GDP growth rates in all scenarios.  
- AFS/HFT own sovereign spread shocks: applied as immediate downgrade shock for the first year in moderate and severe stress with numeric examples preserved as shown in Table 5 (for moderate stress: 0.822*0.63=0.517% plus shock to risk premium from the funding risk model; for severe stress: 0.539*0.63= 0.33957% plus shock to risk premium from the funding risk model).  
- Shocks to NPLs and provisions for STA portfolio: Bayesian VAR model with adjustments and corrections due to change in definition (break-up of data series in 2014); shock applied as a multiplier compared to the last observed value of NPLs (no changes for the baseline scenario).  
- Funding risk shock: projected based on the econometric OLS model using annual changes in Euribor rate, sovereign rate and growth in total funding in all scenarios.  
- Interest income growth: repricing gap model for shock to risk-free interest rates; effective interest rate model for loans and other interest earning assets; shock to Euribor minus shock to risk free interest rate on loans in all scenarios.  
- Net fee and commission income growth, other non-interest income growth, non-interest expenses growth, trading losses: Baseline uses average growth rates for the period 2008-2015; Moderate stress uses half of the worst decline in the period 2008 to 2015; Severe stress uses the worst decline in the period 2008 to 2015.  
- Tax rate: Actual statutory tax rate in all scenarios.  
- Pass-through note: pass-through rates of 50% and 100% assume two effects related to real estate market activity and discount effects when banks retain foreclosed assets.

### Credit risk structure and satellite modeling approach
- Composition of banking assets: Total loans represent 42 percent of total banking sector assets; debt securities represent 22 percent (most of which are marked to market); derivatives close to 19 percent (subject to CVA risk in the stress test). Exposures in loan books are predominantly domestic.  
- Distribution of exposures between approaches: Exposures of the four banks included in the stress test are almost evenly split between IRB and STA approaches. Banks apply the STA approach to sovereign and other public exposures. Although 47 percent of the loan portfolio is under STA approach, 76 percent of credit risk-related RWAs are reported and calculated under the IRB approach.  
- Retail loans: constitute the largest item in loan books but associated capital requirements are fairly low due to very low PDs and LGDs.  
- Satellite models: developed to transmit macroeconomic shocks to PDs and loan loss provisions; supervisory and public data used; satellite models constructed for five broad exposure classes subsequently applied up to 11 regulatory IRB exposure classes. CRR standard risk weights used for migration under STA portfolio. Time series started in 2008 and cover one credit cycle (defaults increased above their means in 2011-2013).  
- PD types used: Through-The-Cycle (TTC) and hybrid PDs (hybrid PDs being more volatile). Hybrid PDs are weighted average PDs corresponding to a combination of Point-in-Time and Through-the-Cycle PDs with a rough calculation indicating a weight of 30 percent of PIT PDs and a weight of 70 percent of TTC PDs.  
- Treatment of defaulted exposures in historic PD series: the FSAP team adjusted historic PD data series to remove double counting from defaulted exposures for provisioning purposes by calculating the ratio of PDs without defaulted exposures to PDs with defaulted exposures and applying it to TTC PDs with defaulted exposures to obtain shocked PDs without defaulted exposures used for expected loss calculation.

### Expected-loss calculation rules
- IRB portfolios: expected losses (ELs) = PIT or hybrid PDs × LGD × EAD.  
- STA portfolios (except sovereigns): flows of NPLs multiplied by the average provisioning ratio of 60 percent.  
- Sovereign exposures: RWAs were not stressed in line with CRR approach; only expected losses were calculated for own sovereign exposures. Sovereign PDs and LGDs referenced Moody’s data for migration shocks.  
- LGD projections: based on house and commercial real estate price projections; a pass-through of house price changes to LGD of 100 percent was assumed for the stress test based on granular data from the FIN-FSA covering the 3 systemically-important banks.

### Data sources and limitations
- Access: supervisory data at individual bank level, solo and consolidated bases, in a physical data room only (Table 6). Publicly-available data and commercial providers, EBA Transparency exercise, and banks’ annual reports were also used.  
- Cross-checks: pre-shock LGD data cross-checked with 2016 World Bank Doing Business report showing an average recovery rate of 90 percent for Finland. PD data cross-checked with additional information from banks and the ECB.  
- Use: these data were used in computation of RWAs and ELs in the stress test. Further modeling-choice details provided in Appendix II.

### Key stress-test finding on credit risk severity
- Expected change in system average PDs under scenarios (IMF model results):  
  - 2015 system average PD: 1.9 percent  
  - Severe stress scenario: system average PD increases to 6 percent in 2018  
  - Moderate stress (euro area-wide) scenario: peak PD of 3.9 percent  
- Interpretation: Top-down stress test results suggest significant increases in PiT PDs under the severe stress scenario, whereas PiT PDs remain flat in the baseline scenario. Potential credit risk losses in the loan book represent a large vulnerability of the banking sector.

*Source: IMF staff calculations as presented in the provided text.*

### 25.      It is worth noting that expected losses in the adverse scenario are mostly driven by

### cr1706 - 25. It is worth noting that expected losses in the adverse scenario are mostly driven by

### Retail-driven expected losses and household vulnerability
- Expected losses in the adverse (severe) scenario are mostly driven by losses in the retail portfolio; households’ solvency is particularly affected.
- Contrast with 1990s Nordic banking crisis: that crisis was characterized by bankruptcy of large companies following the collapse of trade with the former Soviet Union; those big companies have largely been restructured and have diversified international exposures and business lines.
- Household-specific stress test outcome:
  - Implied household PDs rise from the current 3.9 percent (as estimated using micro data) to almost 13.6 percent under a scenario of a significant drop in incomes combined with a 15 percent shock to real estate prices and without any decline in average expenditure (multiplier of 3.5).
  - The PiT PD multiplier estimated by the FSAP team is described as "on a very conservative side" given Finnish households’ room to reduce expenditure.
  - Bank of Finland (BoF) model result: share of households with negative income margin increases from 14.9 to 16.2 percent under a shock to employment related income by -15 percent.

### Provisions, profitability, and quantified credit losses
- Credit losses over the three-year horizon in the loan book:
  - Severe stress scenario: EUR6.9 bn, equivalent to 1.5 percent of total banking system assets.
  - Baseline scenario: EUR0.8 bn, equivalent to 0.2 percent of total banking system assets.
- New provisions in the baseline scenario are more than offset by net income before losses (2 percent).
- Note: expected loss models used for provisioning purposes will be changed in 2018 when new IFRS standards become binding; it is expected that banks will form more provisions for expected losses. Top-down stress testing results partially accounted for that by using stressed PDs and LGDs.

### Household stress-testing methodology (micro-data approach)
- Standard portfolio stress tests may underestimate non-linear effects from large shocks (correlations between household PDs and macro variables can be weakened by public safety nets).
- To capture non-linear effects and household financial-condition shocks, the FSAP used micro data: 2011 micro data from the ECBS Household Finance and Consumption Survey (HFCS).
- The implied PDs reported are not realized or true forward-looking PDs; see Appendices for methodological details.

### Credit Valuation Adjustment (CVA) and derivatives exposures
- Derivatives represent a significant share of banks’ assets; CVA stress tests estimated additional capital requirements due to downgrades (increased default risk) of banks’ counterparties.
- Methodology highlights:
  - Exposures broken down by country segments: Denmark, Finland, Sweden, United Kingdom and the rest of the World.
  - Exposures grouped into Standardized (STA) and Advanced Method (AM) categories.
  - STA exposures grouped into 6 regulatory rating classes with risk weights ranging from 0.7 (class 1) to 10 (class 6) percent.
  - AM exposures grouped into 10 rating classes by Credit Default Swaps (CDS).
  - STA: exposure migration by one notch for each regulatory rating class.
  - AM: stressed scenario used the worst shock to the respective country’s CDS observed in the last 12 years (for the residual region, the shock equaled Denmark’s).
  - CDS shock pass-through assumed non-linear: 100 percent for first grades, 150 percent for middle ones, 200 percent for last ones.
  - Stressed VaR calculations assumed the market turmoil would last for one month (25 days).
- Average shock to RWAs for all banks included in the CVA stress tests was 30 percent. This result would be significantly higher (4 to 5 times) if all exposures exempted by CRR were included.

### Cross-border derivatives concentration and contagion risks
- Analysis highlights cross-border contagion risks among Finland, Denmark, Sweden and the UK due to regional concentration of derivative exposures.
- Many derivatives, including OTC ones, are cleared through companies in the UK; reliance on CCPs is noted but their ability to fully compensate/minimize potential losses is not tested in a severe stress scenario.
- Table 7 (CDS spreads for 5 year sovereign bonds in bps, 2003-2015) summary:
  - Denmark: Maximum 201
  - Finland: Maximum 94
  - Sweden: Maximum 161
  - UK: Maximum 175
  - Mean for Denmark 34; Finland 37; Sweden 29; UK 56
  - Standard Deviation for Denmark 37; Finland 20; Sweden 28; UK 30

### Market risks: interest rate risk and valuation effects
- Interest rate risk (time-to-repricing approach):
  - Three of four banks display a negative time-to-repricing gap (liabilities repriced faster than assets).
  - Aggregate repricing gap for the four banks as of December 2015 for maturities below one year: EUR -1.9 bn.
  - In the severe stress scenario, aggregated loss directly due to the change in interest rates amounts to EUR134 mn, with no material effect on the CAR over the entire stress horizon.
- Valuation effects on debt instrument holdings:
  - Potential valuation losses on own-sovereign debt in the severe stress scenario: EUR293 mn, contributing by 0.3 percentage points to the decline in the CAR over the entire stress horizon.
  - Contributing factors:
    - Moderate exposure to own sovereign: average AFS exposure of 0.9 percent of total assets; individual bank ratios between 0.7 and 3.5 percent.
    - Limited average maturity of Finnish banks’ own sovereign AFS bond portfolio: averaging 4 years.
  - Severe stress scenario includes a large increase in the Finnish government bond 10-year rate of 2.7 percentage points.

### Foreign exchange risk
- Net open FX position and effects:
  - Net open FX position for the banking system as of December 2015: EUR 130 mn, equivalent to 0.03 percent of assets and 0.6 percent of Tier 1 capital.
  - In the severe stress scenario (euro slightly rising against the US dollar), small loss of EUR4 mn.

### Results of solvency stress tests based on macro scenarios
- Under a fully-loaded Basel III definition, in the severe scenario:
  - Aggregate Common Equity Tier 1 (CET1) ratio would drop by 14 percentage points compared with the starting point.
  - CET1 would be 15percentage points below the baseline in 2018.
  - Main drivers of the change in capitalization (contributions over the stress horizon):
    - Funding costs: -6.4 percentage points of RWAs.
    - Loss provisions: -6.1 percentage points of RWAs.
    - Changes in risk-weighted assets: -2.8 percentage points of RWAs due to shocked credit risk parameters.
  - Two banks would see their CET1 ratios fall below the hurdle rates of 4.5 percent, translating into a capital shortfall of 0.2 percent of GDP.
- In the baseline and moderate stress (euro area-wide) scenarios, every bank would stay above the regulatory minima.
- Note: capital shortfalls would be larger on the basis of bank-specific Pillar II requirements which are not publicly-available.

_Italic: Source: IMF staff (from cr1706 - 25. It is worth noting that expected losses in the adverse scenario are mostly driven by)_.

### 43.      Results measured in terms of the leverage ratio are more severe. In the severe stress

### Results measured in terms of the leverage ratio are more severe.

### Leverage ratio stress-test results
- In the severe stress scenario, the leverage ratio at the aggregate level of the four largest banks would drop from 4.2 percent in 2015 to 2.3 percent at end-2018.
- Each of the four largest banks would see its leverage ratio fall below the hurdle rate of 3 percent in 2018.
- This would translate into capital shortfalls equivalent to 2 percent of GDP.
- Table summary (severe stress scenario):
  - Banking system's CET1 ratio (in percent): 8.3
  - Banking system's leverage ratio (in percent): 2.3
  - Number of banks with 6%<T1<8.5% (conserv. buffer): 1
  - Number of undercapitalized banks (CET1<4.5%): 2
  - Number of undercapitalized banks (leverage ratio<3%): 4
  - Max. capital shortfall in terms of CAR, T1, CET1 or leverage ratio (as a percent of GDP): 2

### Caveats on stress-test estimates
- Credit loss estimates and solvency projections in the severe stress scenario may be biased down- or upwards due to:
  - The top-down stress test did not take into account loan write-offs and cures due to data unavailability.
  - Some assumptions had to be made in terms of Loss Given Default and other supervisory parameters.

### Recommendations on IRB models and regulatory response
- Ensure that banks’ Internal Ratings-Based models take into account losses in stressed environments.
- Under current CRR regulation, banks may put higher weights on recent PDs if they consider recent data to better reflect long-term risks; in Finland, recent benign market conditions disproportionately lower long-term PDs.
- To avoid this shortcoming:
  - Use regular micro-data based household stress tests to benchmark PDs and LGDs for retail portfolios and to challenge aggressive modeling techniques that lower risk parameters.
  - Finalize and commend the authorities’ plan to introduce regulatory floors for the risk-weights calculated by IRB banks shortly.

### Market risks — sensitivity analysis overview
- Sensitivity tests assessed vulnerabilities to individual shocks: decline in domestic sovereign securities prices; increase in interest rates affecting net interest income; increase in interest rates worsening credit quality; decline in house prices lowering recovery rates; depreciation/appreciation of the euro nominal effective exchange rate affecting net open FX positions; nominal depreciation worsening credit quality for some borrowers; decline in stock prices.
- Sensitivity tests are static: they assess the instantaneous impact of shocks on banks’ balance sheet positions as of December 2015, assuming risk-weighted assets stay constant after shocks.

### Sovereign bond shock
- A 500 bp increase in interest rates was applied to domestic sovereign AFS and HFT portfolios (using average maturity proxies).
- Result: the CAR in the system would decline by 0.8 percentage points due to these losses taken in isolation.
- Conclusion: Finnish banks are little exposed to domestic sovereign bond risks and initial capitalization on a transitional basis provides a large buffer.

### Interest rate risk — net interest income effects
- A 500 basis point increase in interest rates (maturity ladder/gap analysis) would:
  - Lower the total banking system’s net interest income by EUR94 mn euro.
  - Lower the aggregate Tier 1 capital ratio by 1.9 percentage points.
  - Not cause undercapitalization in any of the four banks.

### Interest rate risk — effects on credit quality
- A 500 basis point increase in domestic real interest rates would:
  - Increase the average PD in the system by 2.4 percentage points (from 1.3 percent to 3.6 percent).
  - Result in credit loss provisions expenses equal to EUR0.8 bn.
  - Decline total CAR by 3 percentage points.
  - Not cause undercapitalization in any of the four banks.
- Caveats: the test assumes no pre-impairment profits under stress, the increase is sustained for one year and only affects PDs and credit losses directly, and indirect/second-round effects are not captured.

### Foreign exchange rate risk
- A 30 percent appreciation of the euro nominal effective exchange rate would cause a loss of EUR39 mn, with no material impact on the aggregate Tier 1 ratio and no undercapitalization in any of the five banks.
- A 30 percent depreciation of the euro nominal effective exchange rate would cause a gain of EUR39 mn.
- Indirect effects of nominal depreciation on credit quality were not assessed because the share of foreign currency lending in total loans is less than 1 percent in Finland.

### Real estate price shock (credit quality)
- A 30 percent decline in domestic real estate (housing and CRE) prices would:
  - Raise weighted average LGD in the system by 21 percentage points (from 27 percent to 48 percent).
  - Cause aggregate loss provisions of EUR0.8 bn.
  - Cause a loss of CAR in the system equivalent to 6 percentage points.
  - Not lead to undercapitalization of any of the banks included in the stress test.
- Note: a key determinant is the increase in risk weights due to very low risk weight density, especially for residential mortgages.

### Stock price decline
- A 50 percent decline in stock prices would cause a loss of EUR 1.1 bn, translating into a 1.1 percentage point decline in the aggregate Tier 1 ratio.
- This shock, taken in isolation, would not cause undercapitalization in any of the four banks.
- Marked-to-market equity investments make up between 0 and 0.7 percent of banks’ total assets.

### Concentration risk: failure of large corporate exposures
- Gross largest exposure (before exemptions and credit risk mitigation): on average 11½ percent of Tier 1 capital.
- Net largest exposure (after exemptions and credit risk mitigation): limited at 6.2 percent.
- Under national regulation, the default of up to the ten largest exposures of each of the four systemically-important banks would not cause any undercapitalization.
- Under extreme scenario with zero recovery rate on credit risk mitigation:
  - Default of the largest exposure of each bank would imply a capital shortfall of EUR268 mn (0.1 percent of GDP) relative to the 8.5 percent Tier 1 capital hurdle rate.
  - Default of the five largest exposures in each bank: total Tier 1 capital shortfall EUR1.5 bn (0.7 percent of GDP).
  - Default of the ten largest exposures in each bank: total Tier 1 capital shortfall EUR2.3 bn (1.1 percent of GDP).
- The 100 percent haircut on collateral valuation is very conservative and implies double default for guarantees; correlation between obligor and guarantor defaults could not be estimated due to lack of guarantor identity information.

### Liquidity stress tests — methodology and calibration
- Tests based on national transposition of Liquidity Coverage Ratio (LCR), Net Stable Funding Ratio (NSFR), and cash flow-based analysis by maturity bucket.
- LCR requirements: banks must maintain an LCR above 60 percent from October 2015 and 100 percent from 2018.
- National transposition differences from Basel III LCR:
  - Covered bonds included in level 1 assets in EU legislation.
  - Larger range of high quality liquid assets included but subjected to high haircuts (e.g. equity haircut of 50 percent).
  - Higher granularity of deposits; accelerated phase-in timetable reaching 100 percent in 2018 rather than 2019.
- NSFR will require banks to maintain a stable funding profile at a one-year horizon from 2018.
- Top-down liquidity stress tests (FIN-FSA and FSAP team) used supervisory maturity structures as of December 2015 for the same sample of four banks; tests run in aggregate combining all currencies and separately for USD, GBP, SEK.
- Foreign currency-denominated liabilities make up 17 percent of total liabilities.
- Scenarios included alternative more severe deposit withdrawals and a third scenario featuring a dry-up of unsecured wholesale funding calibrated to very severe stress conditions.

*Source: IMF staff calculations and IMF report content provided in the source content unit.*

### 62.      The LCR-based stress test was based on three scenarios, with various parameters in

### The LCR-based, NSFR-based, and Outflow Analysis Liquidity Stress Tests (Finland, December 2015)

### A. LCR-based stress test: methodology and baseline funding structure
- Test structure:
  - Based on three scenarios with various parameters for deposit run-off rates, roll-off rates for cash inflows and asset haircuts.
  - First carried out at the aggregate currency level (combining bank positions in every currency).
  - Banks allowed to use sovereign securities as collateral to obtain liquidity through standard ECB lending facilities (weekly repo operations or overnight loans).
  - Liquid assets at the banking system level: 17 percent of total assets (including both on- and off-balance sheet items).
- Funding structure of the banking system (excluding capital and including contingent credit lines) as of December 2015:
  - 45 percent of funding comes from retail sight and time deposits;
  - 28 percent comes from secured wholesale funding, e.g. funding from the Eurosystem and covered bonds;
  - 27 percent is unsecured wholesale funding, coming in particular from credit institutions (20 percent), and operational deposits generated by clearing, custody, and cash management activities.
  - 8 percent of funding comes from undrawn but committed credit and liquidity facilities.

### A. LCR-based stress test: assumptions on run-off, roll-off rates and haircuts
- General principles guiding run-off rate choices:
  - More informed and sophisticated depositors withdraw funding more rapidly than less informed ones → higher run-off rates for wholesale funding than retail.
  - Secured funding sources have lower run-off rates than unsecured funding sources.
- Cash inflows:
  - Derivatives: 28 percent of one-month inflows.
  - Maturing loans from financial institutions and central banks: 54 percent of one-month inflows.
- Roll-off assumptions applied to maturing amounts:
  - 50 percent rates applied to inflows from retail and nonfinancial wholesale counterparties (performing loans to non-financial customers).
  - 100 percent rates applied to maturing loans to financial institutions.
  - Interpretation in tests: cash inflows from performing loans to non-financial customers decline by 50 percent in a month; cash inflows from maturing loans to financial institutions dry up completely.

### A. LCR-based stress test: results by scenario (aggregate/system-wide)
- Baseline (European Commission Regulation parameters):
  - Aggregate LCR at December 2015: 124 percent.
  - Every bank passed the 60 percent hurdle rate (national regulation in 2015) and the 100 percent hurdle rate (binding level in 2018).
- Adverse Scenario 1 (retail shock with higher run-off rates, higher roll-off rates, higher haircuts):
  - Aggregate LCR would fall to 84 percent.
  - Liquidity shortfall: EUR 17 bn.
  - Shortfall expressed in GDP: 8 9 percent of GDP.
  - Note: scenario severity exceeds the 2008/2009 crisis peak monthly retail withdrawal of 4.1 percent since 2003 in Finland; under the scenario banks lose 20 to 30 percent of retail and small business deposits in a month.
- Adverse Scenario 2 (dry-up of unsecured wholesale funding):
  - Assumed 100 percent run-off on unsecured wholesale funding (including corporate deposits other than from SMEs and parent funding).
  - Aggregate LCR would fall to 66 percent.
  - Total liquidity shortfall: EUR45 bn.
  - Shortfall expressed in GDP: 22 percent of GDP.
- Combined shock (retail shock + dry-up of unsecured wholesale funding):
  - System-wide LCR: 61 percent.
  - Liquidity shortfall: EUR57 bn.
  - Shortfall expressed in GDP: 27.5 percent of GDP.
- Foreign-currency LCR stress tests (U.S. dollar, British pound, Swedish krona):
  - Resulting liquidity shortfalls range between EUR1.1 bn and 3 bn, i.e. 0.5 percent and 1.4 percent of GDP.
- Reverse liquidity stress test (retail withdrawals only):
  - To bring the aggregate liquidity ratio below 100 percent (other parameters unchanged), withdrawal rates would have to reach:
    - 30 percent for retail deposits covered by the deposit guarantee scheme in a month;
    - 40 percent for uncovered retail deposits in a month.

### B. NSFR-based stress test: results and implications
- Aggregate NSFR computation (December 2015):
  - Available stable funding for the 4 largest banks: EUR192.2 bn.
  - Required stable funding: EUR168.7 bn.
  - Aggregate NSFR: 113.9 percent (above the minimum requirement of 100 percent).
- System/individual results:
  - Aggregate ratio comfortably above 100 percent, but some individual banks show liquidity shortfalls reflecting excessive maturity transformation.
  - Liquidity mismatches at the one-year horizon: EUR25.4 bn, equivalent to 12.2 percent of GDP.
  - Drivers: size of derivatives portfolios and long-term loans to financial institutions at some banks (these items require stable funding and appear on both sides of the balance sheet).

### C. Outflow analysis stress test: methodology and results
- Maturity buckets used: 1 to 7 days, 8 to 15 days, 16 to 30 days, 31 to 90 days, 91 to 180 days, and more than 180 days.
- Assumption: pace of deposit outflows slows as the time horizon increases.
- For each bucket, net outflows compared to counterbalancing capacity (liquid assets subject to haircuts).
- Key result:
  - For the 16 to 30 days bucket, the gap would equal EUR1.7 bn, equivalent to 0.8 percent of GDP.
- Representative run-off/roll-off/haircut assumptions preserved from Table 11 (selected examples):
  - Retail funding: sight deposits (unstable) run-off rates for buckets: 12%, 6%, 6%, 2.5%, 0%, 0% (1–7 days to >180 days).
  - Retail funding: term deposits run-off rates across buckets: 35%, 35%, 35%, 30%, 25%, 20%.
  - Secured wholesale funding from other financial institutions: 100% across all buckets.
  - Unsecured wholesale funding from other financial institutions: 75%, 75%, 75%, 30%, 30%, 20%.
  - Undrawn committed credit/liquidity lines: 10%, 5%, 5%, 10%, 20%, 0%.
  - Roll-off on loans maturing: 37% across all buckets.
  - Haircuts on liquid assets: Cash items 0%; Securities in trading book 20%; Securities available for sale 20%; Securities held to maturity 20%.

### D. Overall assessment of short-term liquidity risks and policy implications
- Tests confirm exposures to short-term liquidity risks due to:
  - Heavy reliance on wholesale funding.
  - Foreign currency funding exposures.
- Policy recommendations and monitoring options (as identified in the analysis):
  - Improve authorities’ liquidity monitoring, including:
    - Thorough examination of banks’ liquidity positions in foreign currencies in stressed environments;
    - Performing liquidity stress tests using the structure of cash flows at various maturities;
    - Applying customized versions of the LCR along the maturity ladder.
  - Consider options to discourage cross-ownership of covered bonds.

### E. Interconnectedness analysis and contagion risks
- Domestic interbank contagion (network model based on Espinosa-Vega and Solé (2010), using bilateral domestic gross credit exposures for the four largest banks as of end-December 2015):
  - Model assumptions: pure contagion from defaults plus fire-sale funding shocks (50 percent haircut in fire sale of assets and 65 percent roll-over ratio of interbank debt).
  - Findings:
    - Domestic interbank positions are small relative to banks’ capitalization.
    - For the four largest banks, the sum of gross domestic exposures to the other three banks is smaller than their regulatory capital.
    - No single domestic bank failure would trigger another bank’s failure via direct domestic interbank exposures → no cascade effect through the four-bank market.
    - As of end-December 2015, none of the four largest banks is undercapitalized relative to the Tier 1 regulatory minimum after shocks on domestic interbank exposures.
    - Two of the four largest banks show significantly higher vulnerability to spillovers from the other two (index of vulnerability higher); two banks show a higher index of contagion.
- Domestic bank–insurer contagion (matrix of bilateral domestic gross exposures between 4 banks and 6 largest insurance companies, including capital cross-participation, bonds, credit and deposits):
  - Findings:
    - Contagion risks associated with bank–insurer interlinkages are very limited.
    - For every bank and for the six insurance companies, the sum of domestic exposures to other banks and insurance companies is smaller than regulatory capital.
    - No single domestic entity failure would trigger another entity’s failure via this network → no cascade effect.
    - Index of contagion ranges between 0.2 and 44.2 percent (highest index associated with a bank).
    - Index of vulnerability ranges between 1.6 and 13 percent (highest vulnerability associated with an insurance company).
    - Explanation: exposures of insurance companies to banks are larger relative to insurers’ balance sheets (13.7 percent of total assets on average) than exposures of banks to insurers (0.2 percent of banks’ total assets); average insurer exposure to banks is skewed by one insurance institution with a significant exposure to one bank.

*Source: cr1706 - 62.*

### 81.       Cross-border balance sheet analysis reveals strong linkages with the Nordic countries.

### Cross-border balance sheet analysis reveals strong linkages with the Nordic countries.

### Overview
- Cross-border exposures between credit institutions account for 63 percent of total cross-border exposures and principally take the form of loans, advances and debt securities; these typically involve intra-group exposure within Nordic banking groups.
- Derivatives constitute a large share of cross-border exposures for the Finnish banking system: 42 percent of cross-border claims and 36 percent of liabilities.
- Nordea group concentrated derivatives on Nordea Bank Finland (NBF) in 2010, greatly increasing NBF’s cross-border exposure to the London clearing house in the UK; around 75 percent of the derivatives are interest rate-related over the counter products cleared in central counterparty clearing houses.
- Planned conversion of Nordea Bank Finland into a branch (planned for 2017) will move the unique derivatives exposure from NBF to the Nordea group parent bank’s balance sheet and will cause NBF’s cross-border exposure to decline significantly.
- Significant exposures also exist outside the Nordic region, notably with the UK (derivatives in London Clearing House), Germany (debt securities, derivatives) and the USA (Federal Reserve).

*Sources: The Finnish authorities; supervisory and BIS data as cited in the original text.*

### Data and Methodology
- Two supervisory data sets from FIN-FSA were used:
  - Bank exposures to credit institutions in five countries: Denmark, Sweden, Norway, Germany and Russia (channels: loans, debt securities on assets; deposits and debt instruments on liabilities).
  - Large exposure data from COREP showing each bank’s exposure by counterparty (single borrower), using consolidated group-based exposures.
- Analysis used gross data before credit risk mitigation.
- Derivatives exposures to the UK (84 percent of total foreign exposure) at central counterparty clearing houses (CCP) are noted but not included in the bilateral balance sheet analysis.
- Two contagion scenarios were simulated using bilateral exposures of the four largest banks:
  - Credit risk: partial default on loans from Finnish banks following severe stress in counterparties.
  - Funding risk: contraction of cross-border funding from parent banks and other credit institutions leading to potential fire sales.
- Simulation assumptions:
  - Loss-given-default parameter equals 100 percent; banks unable to recover any of their loans.
  - Funding-side test focused on each bank’s ten largest single borrowers and five countries with large exposures.

### Stress-test Scenarios and Modalities
- Credit risk scenario tests how increases in nonperforming exposures affect banks’ CAR and the share of nonperforming exposures required to push CAR below regulatory minima.
- Funding risk scenario simulates dramatic contraction of cross-border funding to estimate changes in liquid assets and potential need for liquidity assistance.

### Key Findings — Credit and Funding Risks
- Maximum credit exposures of the four largest Finnish banks to another Nordic country range from 75 to 180 percent of capital.
- A 39 percent share of payment obligations from one country being unmet would cause the minimum Tier 1 capital requirements (including the capital conservation buffer) of the most exposed banks to fall below the hurdle rate of 8.5 percent.
- Domestic banks would still meet minimum capital requirements under these shocks; subsidiaries of foreign banks would be more affected due to intra-group exposures.
- The four largest banks have significant gross exposures to the same foreign entities; hypothetical default of these counterparties could bring down the solvency ratio of three banks below minimum requirements simultaneously.
- Significant intra-group exposures or exposures to central government exceed the 25 percent of eligible capital large exposure regulatory limit in gross data, though these are exempted from the limits.
- Historical impaired share of intra-group exposures is consistently low at 0.02-0.08 percent over the past 8 quarters.
- A contraction of 34 percent of funding from one country would deplete one bank’s liquid assets; a contraction exceeding 34 percent may lead to fire sales and deteriorated capital positions via the liquidity-solvency nexus.
- Banks’ cross-border funding is geographically and instrument-wise diversified for most banks, but subsidiaries of foreign banks are concentrated in deposits from parent banks.

### Market-based (Implicit Linkages) Analysis
- Diebold-Yilmaz (2014) variance decomposition of weekly equity returns shows strong implicit linkages:
  - Finnish banks share strong linkages or spillovers with insurers.
  - Banks and central government show moderate linkages, reflecting strong fiscal buffers and low implicit sovereign guarantees.
  - Finland’s nonfinancial sectors are strongly interconnected, notably technology and goods sectors.
- Within the Nordic region, Sweden and Finland share the strongest implicit linkages; Iceland shows the weakest linkages to others.
- Finland is the Nordic country most strongly connected to developed economies in continental Europe and shows implicit links with metal commodity markets.
- Implicit linkages corroborate explicit ties from cross-border trade, banking, and investment activity; regional banking activity, wholesale funding reliance, and large cross-holdings of covered bonds contribute to integration.

### Conclusions and Policy Implications
- The Finnish banking system features extremely high concentration, heavy reliance on wholesale funding, strong Nordic linkages, and low risk-weight asset density; stress tests assessed resilience to foreign and domestic macrofinancial shocks.
- Main stress-test results:
  - Severe scenario would very significantly affect the banking system; two of the four largest banks’ capital would fall slightly below minimum regulatory Capital Adequacy Requirements.
  - Recapitalization needs under the severe scenario would be around EUR 550 mn or 0.3 percent of GDP.
  - When measured by leverage ratio, all four banks would fail to meet the 3 percent hurdle rate in 2018; recapitalization needs based on the leverage ratio would be equivalent to 2 percent of GDP, indicating possible underestimation of risks by internal models for risk weighted assets.
  - Funding and liquidity risks are the primary vulnerabilities, including exposure to large deposit withdrawals, a dry-up of unsecured wholesale funding, and material exposure to funding risks in foreign currencies.
- Recommendations and areas warranting consideration:
  - High vigilance by authorities given structural strong interconnectedness within the Nordic region.
  - Consider expanding existing protocols with other Nordic supervisors to include formal, region-wide sharing of supervisory data.
  - Consider a Nordic-wide coordinated stress test exercise to account for interlinkages and spillovers, including liquidity-solvency interactions and coverage beyond banks to cross-border regional insurance companies and financial market infrastructure.

*Italicized source attribution: IMF staff report as extracted from the provided content unit.*

### 97.      Recommendations include reviewing banks’ Internal Ratings-Based models, finalizing

### cr1706 - 97.      Recommendations include reviewing banks’ Internal Ratings-Based models, finalizing

### Key Recommendations
- Review banks’ Internal Ratings-Based (IRB) models.
- Finalize plans to introduce regulatory floors to the risk-weights calculated by IRB banks.
- Improve liquidity monitoring by performing liquidity stress tests.
- Conduct a Nordic-wide coordinated stress test exercise to:
  - take into account interlinkages and spillovers;
  - include liquidity–solvency interactions;
  - expand coverage beyond banks to cross-border regional insurance companies and financial market infrastructure.

### Risk Assessment Matrix — Principal Risks and Expected Impacts
- I. Euro area bond market contagion / Higher-than-expected fallout from the UK referendum result on EU membership.
  - Overall Level of Concern: Medium
  - Relative Likelihood: Medium
  - Expected Impact: Medium
  - Notes: Finland could be affected if sovereign and financial sector stress reemerges across the Euro area. Direct impact of UK leaving EU expected to be limited, but increased financial volatility possible due to transactions with UK central clearing counterparties.
- II. Structurally weak growth in key advanced and emerging economies.
  - Overall Level of Concern: High
  - Relative Likelihood: High
  - Expected Impact: High
  - Notes: Finland’s exports tightly linked to Euro area markets; weak external demand could push Finland into stagnation.
- III. Heightened risk of fragmentation/security dislocation in parts of the Middle East, Africa, and Europe.
  - Overall Level of Concern: Medium
  - Relative Likelihood: Low
  - Expected Impact: Low–Medium
  - Notes: Russia is Finland’s fifth largest export market; a severe downturn in Russia could shave “a couple tenths of a percent” off GDP growth.
- IV. Severe stress macroeconomic and house price shock in an interconnected neighboring Nordic country.
  - Overall Level of Concern: Medium
  - Relative Likelihood: Medium
  - Expected Impact: Medium
  - Notes: Household debt high in the Nordics; largest banks in Finland are Swedish and Danish—spillovers could curtail lending in Finland.
- V. Persistently lower energy prices, triggered by supply factors reversing only gradually.
  - Overall Level of Concern: Low
  - Relative Likelihood: Low
  - Expected Impact: Medium
  - Notes: Long-lasting low oil prices could reduce inflation and investment; an increase in commodity prices from disruptions would dent purchasing power and profitability.

### Stress Test Matrix (STeM) — Banking Sector: Solvency, Liquidity, Contagion
- General
  - Institutions included (solvency, liquidity, contagion): 4 banks
  - Market share covered: 93 percent of the MFI assets
  - Baseline / latest data date: end- December 2015 (solvency baseline); Latest data: December 2015 (liquidity & contagion)
  - Scope: bank consolidated level data for banks headquartered in Finland and sub-consolidated data for subsidiaries of foreign banks; supervisory and market data sources.

- Solvency — Methodology and Assumptions
  - Methodologies: satellite models developed by the FSAP team; balance sheet-based approach; market data-based approaches.
  - Satellite models: credit losses, pre-impairment income, credit growth; models to integrate solvency–funding interactions; methodology to calculate sovereign risk; losses from bonds and money market instruments using modified duration haircuts.
  - Income projection: net fee income and commission income projected based on nominal GDP growth and expert judgment.
  - Stress test horizon: 3 years (2016/2018).
  - Finland-specific severe stress scenario GDP growth path: -5.0, -3.1 and +0.2 percent, in 2016, 2017 and 2018 respectively.
  - Market shocks: calibrated to magnitudes close to those observed in 2008/2009.
  - Outputs: system-wide capital shortfall; number of banks and percentage of banking assets falling below certain ratios.

- Liquidity — Methodology and Assumptions
  - Methodologies: Basel III-LCR and NSFR type proxies (based on European Commission Delegated Act); cash-flow based liquidity stress test using maturity buckets; reverse liquidity test by banks.
  - Risks assessed: funding liquidity (liquidity outflows); market liquidity (price shocks).
  - Buffers: counterbalancing capacity; central bank facilities.
  - Tail shocks: run-off rates calculated following historical events and LCR/NSFR rates; bank run and dry up of wholesale funding markets, accounting for haircuts to liquid assets.
  - Regulatory standards referenced: European Commission Delegated Regulation (EU) 2015/61; Basel Committee on Banking Supervision (2014), “Basel III: The Net Stable funding ratio”.

- Contagion — Methodology and Assumptions
  - Methodologies: Network interbank model by Espinosa-Vega and Solé (2010); Diebold–Yilmaz variance decomposition connectedness methodology; data-driven correlation networks.
  - Tail shocks: pure contagion via default of institutions; spillover index and transmission.
  - Outputs: number of undercapitalized and failed institutions and their asset shares; evolution and direction of spillovers within the network.

### Key Numerical and Design Parameters (preserved exactly)
- Market share covered: 93 percent of the MFI assets
- Baseline / latest data date: end- December 2015; Latest data: December 2015
- Stress test horizon: 3 years (2016/2018)
- Finland-specific severe stress GDP growth: -5.0, -3.1 and +0.2 percent, in 2016, 2017 and 2018 respectively
- Multiplier a for updating TTC PDs with shocked PIT PDs: 10%
- Quantified PD sensitivities:
  - Mortgage: unemployment +1 percentage point → PD +0.6 percentage points; GDP growth -1 percentage point → PD +0.2 percentage points
  - Non-financial corporates: GDP growth -1 percentage point → PD +0.1 percentage points; Euribor +100 basis points → PD +0.2 percentage points
  - SMEs: unemployment +1 percentage point → PD +0.7 percentage points

*Source: cr1706 - 97.      Recommendations include reviewing banks’ Internal Ratings-Based models, finalizing*

### Appendix III. Table 1. Finland: Results from the Estimation of the IMF Credit Risk Satellite

### Appendix III. Table 1. Finland: Results from the Estimation of the IMF Credit Risk Satellite

### Estimated model results by asset class (Dependent variable: logit transform of the PD)
- Asset classes reported: Mortgages, Corporate, SMEs, Institutions, Retail(oth.)
- Coefficients, t-statistics (in parentheses), significance markers preserved as in source:

- YoY real GDP growth (in log)
  - Mortgages: -0.018** (-2.18)
  - Institutions: -0.02** (-2.45)
  - Retail(oth.): -0.02* (-1.98)

- YoY real GDP growth (in log) (lagged by 2 quarters)
  - Corporate: -0.01* (-1.83)

- Unemployment rate (in percent)
  - Corporate: -0.05** (2.05)
  - Mortgages: 0.099** (2.44) (lagged by 1 quarter)
  - Retail(oth.): 0.07* (1.76) (lagged by 7 quarters)

- Euribor, 6M rate (in percent)
  - Mortgages: -0.06** (2.5)
  - Corporate: -0.06* (1.6)

- Dummy variable
  - Mortgages: -0.06** (2.29)
  - Corporate: 0.18 (1.6)*
  - Retail(oth.): 0.19** (2.46)

- Lagged dependent variable
  - Mortgages: 0.32* (3.50)
  - Corporate: 0.88*** (10.7)
  - SMEs: 0.24*** (2.45)
  - Institutions: 0.63*** (7.49)
  - Retail(oth.): 0.46*** (2.75)

- Constant
  - Mortgages: -3.50* (-6.59)
  - Corporate: -0.44* (-1.67)
  - SMEs: -2.05*** (-5.5)
  - Institutions: -2.46*** (-4.38)
  - Retail(oth.): -2.15*** (-3.59)

- Estimation method by asset class
  - Mortgages: OLS
  - Corporate: OLSR
  - SMEs: RLS
  - Institutions: OLS
  - Retail(oth.): OLS

- Goodness-of-fit and sample size by asset class
  - R-square:
    - Mortgages: 0.51
    - Corporate: 0.83
    - SMEs: 0.53
    - Institutions: 0.88
    - Retail(oth.): 0.74
  - # of observations:
    - Mortgages: 243
    - Corporate: 302
    - SMEs: 273
    - Institutions: 128
    - Retail(oth.): 3128

- Notes:
  - t-statistics in parentheses.
  - * Denotes significance at the 10 percent level; ** at the 5 percent level; and *** at the 1 percent level.
  - Dependent variables: Logit transformed PDs

### Household stress-test inputs, methodology, and key results
- Data and update approach
  - Representative ECB Household Consumption and Finance Survey sample: detailed data about 10900 households (income, financial/non-financial assets, monthly payments for primary residential mortgage).
  - Original survey year: 2009; updated to approximate 2015 using Statistics Finland annual surveys and mapping by deciles (income) and quintiles (expenditure).
  - Income and expenditure updates:
    - Income: mapped using deciles; expenditure: mapped using quintiles (average expenditure per quartile used, not minimum).
    - Expenditure data available for 2012; income data used for 2014 (survey frequency differences noted).

- Income dynamics: 2009-2014 (by income decile)
  - Columns preserved exactly: Growth rates: wages; Growth rates: disposable income; Growth rates: transfers*
  - Decile 1: 6.59% ; 2.78% ; 0%
  - Decile 2: -1.01% ; -1.43% ; -2%
  - Decile 3: -10.52% ; 0.22% ; 10%
  - Decile 4: -10.75% ; -1.06% ; 11%
  - Decile 5: -8.96% ; 2.19% ; 10%
  - Decile 6: -0.29% ; 5.57% ; 18%
  - Decile 7: 7.95% ; 5.07% ; 9%
  - Decile 8: 1.80% ; 7.65% ; 20%
  - Decile 9: 3.91% ; 5.68% ; 10%
  - Decile 10: 0.91% ; 4.56% ; 27%
  - Note: "While high income households got biggest increase in transfers, these constitute relative low share of total income (compared to low and middle income households)."

- Descriptive characteristics of households with mortgages
  - Are unemployed (i.e. no one member of household is employed): 3.4%
  - Households which receive social transfers: 27%
  - Have negative income: 5%

- Net income margin (IM) and default rule (exact formulas preserved in structure)
  - Net income margin IM computed as:
    - IM = [Employee income*(1+g_ie)*(1+α) + Regular Social Transfers*(1+g_it)*(1+β) + Other Income*(1+g_io)*(1+γ)] − Expenditure_decile_i*(1+δ) − Residential Mortgage Payments*(1+ε)
    - All income and expenditure data annualized; regular social transfers exclude pensions.
    - α, β, γ, δ, ε are shock parameters; g are growth rates.
  - Net assets (NA) computed as:
    - NA = Total real assets*(1+μ) + Total financial assets*(1+θ) − Total outstanding balance of household's liabilities
    - μ and θ are shocked growth rates of real estate prices and financial assets respectively.
  - Default rule for implied PD:
    - PD = {1, if IM<0, and NA<0; 0 otherwise}

- Pre-shock and calibration observations
  - Pre-shock data: around 2.2% of defaulted households in sample.
  - Implied weighted average PD for mortgage portfolio computed as:
    - PD_s^t_P = Σ_{i=1}^{10} PD_s_i * Share_i  (formula preserved)

- Stress scenario assumptions and outcomes
  - Severe stress scenario used for household ST simulations.
  - Additional assumption: incomes (mostly social transfers and wages) decline by 20% (instantaneous shock).
  - Final distribution of PDs after simulation (qualitative: illustrated in Figure A4 in source).
  - Average implied PD for the mortgage portfolio:
    - Will increase from 3.9 to 13.6 percent.
  - Counterfactual without decrease in wages and social transfers:
    - Defaults will increase by 3.8 percentage points to 7.6 percent on average.
  - Caveats:
    - Income and expenditure update by deciles and quintiles contain additional error margins.
    - Using average expenditure (conservative) yields higher estimates of households with negative margins; assuming reduced expenditure post-shock would lower PD shocks but increase macroeconomic shocks to consumption/GDP.

- Comparison with Bank of Finland household stress test
  - Bank of Finland result cited: increase share of households with negative income margin from 14.9 to 16.2 percent (shock to employment related income by -15 percent).
  - Key methodological differences (as reported):
    - FSAP team used 2009 data updated by deciles; Bank of Finland uses actual micro data for 2015.
    - FSAP definition of default: negative income AND negative assets; Bank of Finland: negative income only.
    - FSAP uses average expenditure per quartile; Bank of Finland uses minimum income.
  - Implication: FSAP team obtained higher multipliers assuming higher-income households have higher expenditure; allowing households to reduce expenditure would lower PD shocks in FSAP model.

### Additional methodological notes relevant to credit risk and stress calculations
- PDs from household model used in broader loss provisioning and capital adequacy (summary of links preserved):
  - Loan loss provisions for mortgages: Shocked quasi PIT PDs * Risk factor from Households ST model * LGDs * EADs.
  - Loan loss provisions for NPLs: Shocked NPLs * 60% provisioning ratio [NPLs for mortgage portfolio * risk factor from HH model].
  - Losses on AFS and HFT securities: haircuts [Moody's PDs * LGD + risk premium] * Amount of own sovereign bonds.
  - RWAs adjustments:
    - IRB exposure classes: TTC PDs + 10% of PiT PDs; AIRB: LGDs + 10% of shocked LGDs; FIRB: No shock for LGDs.
    - STA exposure classes: 50% of exposure migration by one step (excluding zero risk weighted exposures).
  - Other elements preserved from source flow: interest income, funding cost shocks, interest rate risk on banking book (gap), trading/fee income assumptions, taxes, dividends, credit growth in line with nominal GDP projections, CVA RWAs adjustments.

### Key statistical and model-summary points
- Pre-shock mortgage portfolio implied PD: 3.9 percent
- Post-severe-stress mortgage portfolio implied PD: 13.6 percent
- Alternative (no wages/transfers decline) average PD: 7.6 percent
- Pre-shock fraction of defaulted households in sample: around 2.2%
- Household sample size in the ECB survey: 10900 households
- Income dynamics table and decile-level growth rates preserved exactly (see section above)

*Source: IMF staff calculations*

### Appendix VI. Table 4. Financial Variables Used in Analysis

### Appendix VI. Table 4. Financial Variables Used in Analysis

### Implicit cross-border financial linkages — country-level
- Equity Indices:
  - OMX Iceland All shares, Sweden DS Market, Denmark DS Market, Finland DS Market, Norway DS Market, North America DS Market; Developed Europe DS Market; Developed Asia DS Market; Developing Europe DS Market; Developing Asia DS Market; Latin America DS Market; Middle East and Africa DS Market
- Global Factors:
  - GSCI Four Energy Commodity, Industrial Metals, Precious Metals, and Agriculture indices; Citibank Global Bond Index; MSCI World Real Estate Index; HFRX Global Hedge Fund Index; JPM Trade-weighted USD REER Index; US Economic Policy Uncertainty Index; Citi Short- and Long-Term Macro Risk indices; Citi Global Economic Surprise Index
- Controls:
  - CBOE S&P500 Skew, S&P500 VIX, and OVX indices; Merrill Lynch MOVE Index; ASR Global Equity Risk Premium Index

Notes:
- Changes in bond yields were calculated in the same manner as equity returns. All bond index data was of daily frequency and sourced from Thomson-Reuters Datastream.
- For robustness, Sovereign-to-Other implied financial linkages were also estimated using single name 5-year CDS on senior sovereign debt; the results were analgous to those based on bond yields.

### Implicit cross-border banking and insurance financial linkages — sector-level
- Nordics:
  - Sweden Banks and Insurance; Denmark Banks and Insurance; Finland Banks and Insurance; Norway Banks and Insurance
- Non-Nordic:
  - United States Banks and Insurance; United Kingdom Banks and Insurance; France Banks and Insurance; Germany Banks and Insurance; Japan Banks and Insurance; China Banks and Insurance; Brazil Banks; India Banks and Insurance; Turkey Banks; Russia Banks

Controls (sector-level models, also listed for Finland specifically):
- Sweden DS Market; Denmark DS Market; Finland DS Market; Norway DS Market; S&P500; FTSE 100, CAC 40, DAX 30, Topix, Shanghai SE A Shares, CNX 500, BIST National 100, Russia MICEX; GSCI Four Energy Commodity, Industrial Metals, Precious Metals, and Agriculture indices; Citibank Global Bond Index; MSCI World Real Estate Index; HFRX Global Hedge Fund Index; JPM Trade-weighted USD REER Index; US Economic Policy Uncertainty Index; Citi Short- and Long-Term Macro Risk indices; Citi Global Economic Surprise Index; CBOE S&P500 Skew, S&P500 VIX, and OVX indices; Merrill Lynch MOVE Index; ASR Global Equity Risk Premium Index

### Implicit domestic financial linkages — sector-level (selected countries)
- Denmark:
  - Banks; Construction Materials; Consumer Goods; Consumer Services; Food Beverages; Food Producers; Health Care; Industrials; Insurance; Oil and Gas; Real Estate; Technology; Telecom, Media, IT
- Finland:
  - Banks; Basic Materials; Construction Materials; Consumer Goods; Consumer Services; Electricity; Forestry Paper; Health Care; Industrials; Industrials Metals Mines; Oil Gas; Real Estate; Retail; Technology; Telecom Media IT
- Iceland:
  - OMX All Shares
- Norway:
  - Banks; Basic Materials; Chemicals; Consumer Goods; Food Beverages; Food Producers; Industrials; Industrials, Metals, and Mines; Insurance; Oil and Gas; Real Estate; Telecom, Media, and IT; Farming and Fishing
- Sweden:
  - Banks; Basic Materials; Construction Materials; Consumer Goods; Health Care; Industrial Metals and Mines; Insurance; Oil and Gas Producers; Real Estate; Retail; Speciality Finance; Technology, Hardware, and Equipment

- Nordic Sovereigns:
  - Iceland 10-Year Government Benchmark Bid Yield Index; Sweden 10-Year Benchmark Government Bond Index; Finland 10-Year Benchmark Government Bond Index; Norway10-Year Benchmark Government Bond Index; Denmark 10-Year Benchmark Government Bond Index

Controls (sovereign/sector models):
- Sweden DS Market; Denmark DS Market; Finland DS Market; Norway DS Market; Russia MICEX; GSCI Four Energy Commodity, Industrial Metals, Precious Metals, and Agriculture indices; Citibank Global Bond Index; MSCI World Real Estate Index; HFRX Global Hedge Fund Index; JPM Trade-weighted USD REER Index; US Economic Policy Uncertainty Index; Citi Short- and Long-Term Macro Risk indices; Citi Global Economic Surprise Index; CBOE S&P500 Skew, S&P500 VIX, and OVX indices; Merrill Lynch MOVE Index; ASR Global Equity Risk Premium Index

### Direct / indirect trade, banking, and investment ties (data sources and vintages)
- Trade:
  - IMF Direction of Trade Statistics (DOTS), Total Trade, as of 2014Q2
- Banking:
  - BIS Consolidated Banking Statistics, International Claims over All Sectors, as of 2015Q2
- Investment:
  - IMF Coordinated Portfolio Investment Survey (CPIS), Net Inward Equity Investment and Net Inward Position in Debt Instruments, as of 2015Q2
  - IMF Coordinated Direct Investment Survey (CDIS), Net Inward Foreign Direct Investment, as of 2015Q2

### Concurrent sovereign and financial sector credit risk
- Sovereigns:
  - 5-year single-name CDS spreads on senior debt for Denmark, Finland, Iceland, Norway, and Sweden
- Financials:
  - 5-year single-name CDS spreads on senior debt for Danske Bank, DnB, Nordea, and Sampo Bank

### Implicit sovereign guarantee of financial sectors
- Moody’s CreditEdge 5-year fair value CDS and aggregate CDS spreads for Dankse Bank, DnB, and Nordea

### Notes on implicit financial linkages (data availability, modeling choices, and definitions)
- All sector-level bond and equity indices were sourced from Datastream.
- High-quality sector-level equity indices were not available for Iceland; the OMX All Shares Index was used as a proxy to measure implied sector-to-sector financial linkages.
- Data availability precluded including Iceland in analysis of cross-border banking and insurance linkages.
- Insurance sector equity indices of sufficient length were not available for Brazil, Turkey, or Russia.
- Publicly available equity price data on Swedish insurers was limited, with constituents comprised of at most three insurers depending on the time period; thus linkages inferred for Swedish insurers should be viewed cautiously.
- Motivation for choice of controls:
  1. Implicit financial linkages (country-level): minimize influence of global changes in investor risk appetite for equities and commodities.
  2. Implicit financial linkages (sector-level): minimize influence of global changes in investor risk appetite for equities and commodities; minimize influence of country-specific changes in investor risk appetite for equities; minimize influence of country-specific broad equity prices trends; minimize influence of idiosyncratic trends in behavior of global factors.
- Endogenous variables in each VAR specification were ordered by descending degree of perceived exogenity — global factors were included first, when applicable, and then equity and bond indices.
- The choice of ordering of global factors was determined by the authors’ subjective perception of these variables' degree of exogeneity; alternative orderings had no discernable impact on results.
- In specifications where global factors were treated as endogenous, the following ordering was used:
  - Citi Short- and Long-Term Macro Risk indices; Citi Global Economic Surprise Index; US Economic Policy Uncertainty Index; JPM Trade-weighted USD REER Index; GSCI Four Energy Commodity, Industrial Metals, Precious Metals, and Agriculture indices; Citibank Global Bond Index; MSCI World Real Estate Index; HFRX Global Hedge Fund Index.
- Country and sector bond and equity indices were ordered in terms of total market value, from largest to smallest.

Other data notes:
- 5-year CDS spread data for Sampo Bank begins in 2012; thus all bivariate copula modeling between Sampo Bank and other entities reflects only the period of 2012 onward.
- 5-year CDS spread data was not available for the Iceland Housing Finance Fund (HFF); yields on long-dated HHF bonds were used as a proxy.
- Entity-level 5-year fair value CDS spreads and aggregate CDS spreads are pre-computed financial variables sourced directly from Moody’s CreditEdge.
- In instances where fair-value CDS spread data was missing, predictive means matching was used to impute values (see van Buuren and Groothuis-Oudshoorn, 2011).
- DS = Datastream; CDS = credit default swap; other acrony​mes are considered common knowledge and not explictly defined.

*Appendix VI. Table 4. Financial Variables Used in Analysis — IMF Country Report (selected excerpts).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2017/cr1706.pdf_