## cr1869

## Source details

**Canonical URL:** [cr1869](https://www.imf.org/-/media/files/publications/cr/2018/cr1869.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2018/cr1869.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2018/cr1869.pdf.json)

---

### Executive summary — overview and scope
- Belgium’s financial landscape changed since the GFC and the European sovereign crisis: banking system assets fell from 470 percent of GDP in 2008 to 250 percent (driven by scaling back cross-border activities and NBB regulations); banks shifted toward domestic lending and deposit funding; insurance sector consolidated and moved partially from traditional guaranteed products to asset-management type products.
- FSAP stress testing comprised:
  - Macro-driven solvency stress tests for banks and insurance companies.
  - Hypothetical liquidity and asset market stress tests for banks including contagion analysis (interbank bilateral exposures and market-based contagion).
- Coverage:
  - IMF stress test covered six major Belgian SIs representing around 90 percent of total banking system assets: Argenta, AXA Bank Belgium, Belfius Banque, BNP Paribas Fortis, ING Belgium, and KBC Group.
  - Insurance stress test: eight undertakings representing 78 percent of total balance sheet assets.

### Macrofinancial background and key vulnerabilities
- Growth and outlook:
  - Real GDP grew by 1.2 percent in 2016 and expected to accelerate to 1.7 percent in 2017.
  - Output gap closing and expected to turn positive in 2018.
  - Absent further structural reforms and stronger growth in Europe, growth expected to remain at around 1.5 percent for the foreseeable future.
- Financial cycle and credit:
  - Bank credit growth accelerated since 2015; small positive credit gap opened in 2016.
  - Credit growth driven by residential mortgages and credit to the construction sector.
  - Financial cycle running ahead of the economic cycle — increased need for timely macroprudential policies.
- Banking sector structural facts:
  - Banks’ exposures to the non-financial private sector increased from 40 percent of assets in 2008 to about 50 percent in 2016.
  - Size of banks’ bond portfolio declined by half between 2008 and 2016.
  - Deposit funding rose from under 40 percent of total liabilities to 55 percent in 2016.
  - Market shares: BNP Paribas Fortis 30 percent of total system assets; KBC banks 24 percent; Belfius 15 percent.
- Insurance sector facts:
  - Total insurance sector assets represented 80 percent of GDP in 2016, up about 12 percentage points of GDP over 2012–16 (largely reflecting Solvency II market-based valuation).
  - Annual premiums in the life business declined by 31 percent between 2012 and 2016.
  - Sovereign bonds account for 43 percent of total assets of Belgian insurers; more than half of these sovereign bonds are Belgian sovereign bonds.

### FSAP stress-testing strategy and scenarios
- Four-pronged testing strategy:
  - Solvency stress tests (scenario-based and sensitivities).
  - Liquidity stress tests (LCR-based and implied cash-flow tests).
  - Contagion module (interbank bilateral exposures and network analysis).
  - Systemic risk module (market-based contagion using CoVaR).
- Scenario horizon: 2017Q1–2021Q4.
- Baseline: October 2017 WEO; Adverse scenario calibrated with IMF’s Global Macrofinancial Model (GFM) using a layers-of-shocks approach (Layer 1 through Layer 6).
- Key Layer 5 (Domestic Layer) assumptions (deviations from April 2017 WEO baseline):
  - Long term government bond yield: +100 basis points
  - Fiscal expenditure shocks: +1.0 percentage point to primary fiscal balance ratio
  - Wholesale funding spread: +50 basis points
  - Real housing price correction, peak-to-trough: −20 percent
  - CRE price correction, peak-to-trough: −30 percent
  - Private investment: −8.0 percent
  - Private consumption: −2.0 percent
- Tail/Adverse scenario outcomes (selected):
  - Belgium: nine quarters of negative growth, peak at −2.8 percent in 2018Q1; output falls 8.1 percent below baseline by 2019.
  - Consumption price inflation: falls 3.5 percentage points below baseline by 2020.
  - Money market rates widen by 120 basis points; long-term yields rise by 150 basis points by 2018.
  - Real equity prices fall by 20 percent; real estate prices collapse by 20 percent.
  - Scenario magnitude: 3.1 standard deviation move in two-year cumulative real GDP growth rate by 2018.

### Credit risk, PDs, LGDs, EADs — methodology highlights
- Two-step PD projection:
  - Aggregate PD proxies projected monthly by portfolio and geography using market-based PDs (Moody’s EDF series).
  - Statistical mapping from aggregate EDFs to bank-specific regulatory PDs using regressions on post-June 2014 supervisory data.
- PD proxies and inputs:
  - Corporate and retail unsecured PD proxies from Moody’s Analytics average one-year EDF by geography and portfolio (monthly data Jan 2005–Dec 2016).
  - Sovereign PDs extracted from spreads via a reduced-form structural/Merton approach assuming LGD=45 percent.
  - Mortgage PD shifts taken from ECB staff modelling tool (STAMP€) due to lack of credit register time series.
- LGD and EAD projections:
  - Adverse LGDs calibrated for first three years by portfolio and bank using estimated non-defaulted LGD ratios and portfolio shifts.
  - EAD projection driven by balance sheet growth, structural FX risk, and triggered credit lines/guarantees; shocks to triggered credit lines calibrated on maximum observed increases in off-balance sheet to EAD ratios in Pillar 3 disclosures.
- STA exposures and migration:
  - NPL definition: past due >90 days and defaulted loans; provisions projected for specific and collective allowances.
  - NPL coverage ratio floor in adverse scenario based on Basel crisis-era average of 65 percent.

### Bank solvency results and sensitivities (selected quantitative outcomes)
- System size and composition (end-2016):
  - Size of largest six Belgian banks: €917 billion.
  - RWAs: €349 billion (around 40 percent risk density).
  - Over 85 percent of capital required to cover unexpected credit losses; market risk just under 2 percent of capital requirements.
- Regulatory hurdle (CET1 hurdle rate):
  - CET1 hurdle rate: 8.5 percent for the Big 4 and 7.75 percent for the remaining two O-SIIs.
  - Hurdle also includes a 3 percent Tier 1 ratio based on CRR leverage framework.
- Aggregate CET1 (selected points):
  - Aggregate CET1 (end-2016): 15.1 percent.
  - Aggregate CET1 low point (2018, adverse): 11.4 percent (decline of 370 basis points).
  - Aggregate CET1 by 2021 (adverse): 13.2 percent.
  - RWAs rise by 290 basis points over the first two years (adverse).
  - Near-term contributions to CET1 decline: credit losses 140 basis points; valuation losses 210 basis points.
  - End-horizon contribution (5 years) of credit and valuation losses: 310 basis points each.
  - Total regulatory capital additional decline: 30 basis points to 17.3 percent by 2021 (due to elimination of no-longer-eligible Tier 2 instruments).
- Sensitivity test outcomes:
  - Low-for-long scenario: cumulative decline in interest rates of some 150-bps relative to baseline; CET1 erosion by 130 bps by 2021.
  - Mortgage collateral / LGD shock: increase of 25 percentage points in LGDs on Belgian mortgages → CET1 decline of 110 bps; banks would still post a 14 percent CET1 at the low point.
  - Inclusion of derivatives in repricing gap: aggregate effect positive; some banks build capital buffers between 50bps and 70bps.
  - Empirical pass-through of funding costs: raises CET1 up to 60bps in some cases.
- Baseline outlook: CET1 ratios stabilize around 15 percent later in horizon under baseline; deductions from Basel III leave CET1 comfortably above fully-loaded minimums.

### Liquidity stress testing, asset encumbrance, and contagion
- LCR and cash-flow testing samples and data:
  - LCR-based sample: six banks from solvency sample plus Degroof Petercam and Bank of New York Mellon, and all LSIs; average LCR ~140 percent in March 2017 for whole sector, Big six, and Big four.
  - Cash-flow test sample: six major Belgian banks and Bank of New York Mellon; ECB internal liquidity tool used across eighteen maturity buckets (1-day to 12 months).
- LCR scenario impacts (average LCR changes):
  - Systemic scenario: average LCR ratio would shed 40 basis points.
  - Idiosyncratic scenario: average LCR ratio would shed 46 basis points.
  - Combined scenario: average LCR would edge slightly above the 80 percent mark.
- Asset encumbrance:
  - Average asset encumbrance ratio (2016): 13.7 percent of total assets.
  - Asset encumbrance reached over one fourth of total central bank eligible assets for the sector.
- Contagion and network analysis:
  - Belgian interbank network density: low; core-periphery with largest banks at periphery.
  - Standard contagion assumptions: 40 percent LGD, 35 percent of lost funding non-replaceable, 30 percent fire sale discount.
  - Under these assumptions, no bank default would lead to consecutive defaults; results robust to exposure definitions and parameter variations.
  - Caveats: analysis excludes foreign counterparties and small exposures; assumes passive central bank; excludes valuation effects from common exposures and precautionary liquidity hoarding.

### Insurance solvency tests — design, results, and risks
- Scope and methodologies:
  - Solvency II used as framework; reference date end-2016.
  - Eight companies participated (78 percent coverage by assets); six used standard formula, two used partial/full internal models.
  - BU and TD exercises: BU included IMF adverse and NBB “low-for-long”; TD only IMF adverse.
  - IMF adverse scenario shocks: interest rates, equity/property prices, credit spreads, mortgage loan losses, and EUR exchange shocks; shocks front-loaded.
- Key stress assumptions (TD adverse scenario excerpts):
  - Equity shocks: Belgium and advanced economies −16.2 percent; Emerging/developing economies −25.0 percent.
  - Real estate: Belgium −20.0 percent; elsewhere −15.0 percent.
  - Sovereign bond spread changes: Belgium 1.47%; EA high-spread countries 2.00%; EA low-spread countries 0.50%.
  - Mortgage loans haircut: −2.0%.
- Results (BU median and aggregate):
  - Median solvency ratio pre-stress: 184 percent.
  - Median solvency ratio — IMF adverse scenario: 124 percent.
  - Median solvency ratio — NBB low-for-long scenario: 145 percent.
  - Aggregated solvency surplus: before stress €10.2 billion; IMF adverse €4.1 billion; NBB low-for-long €6.2 billion.
  - In each scenario, one company drops marginally below 100 percent SCR coverage; capital needed to restore full coverage amounts to less than €10 million in each case.
- Top-down results:
  - Two companies below 100 percent SCR in IMF adverse with aggregated capital shortfall of less than €400 million.
  - Market value of assets declines by 12 percent in IMF adverse; liabilities decline by 10 percent; excess of assets over liabilities down 29 percent.
- Drivers and vulnerabilities:
  - Main driver: increase in Belgian sovereign yields and higher Belgian credit spreads.
  - Volatility adjustment (VA) increase partially offsets credit spread effects for many companies.
  - Concentration: sovereign bonds and domestic sovereign positions are large; reinvestment risk material — 50 percent of fixed-coupon investments maturing between 2017 and 2025, average coupon 3.6 percent; assuming reinvestment at 1 percent, average coupon drops to 2.5 percent by 2025 and to 1.9 percent in 2030.
  - Average guaranteed interest rate end-2016: 2.6 percent (down from 3.1 percent five years earlier); significant long-maturity guaranteed liabilities remain.
- Underwriting and catastrophe sensitivities:
  - Longevity shock (permanent 20 percent decline in mortality): own funds decline <1 percent.
  - Severe pandemic: smaller effect.
  - Repetition of windstorms Lothar/Xynthia: median company loss in own funds <1 percent; reinsurance absorbs most.
- Financial conglomerates:
  - Three banking-led conglomerates: Belfius, KBC Group, Investar.
  - For the three, no insurer drops below 100 percent SCR; no additional capital needed from parent; all three remain profitable after stress and can sustain dividend payments.

### Institutional guidance and policy recommendations (timing and priority preserved)
- Expand stress testing and modelling:
  - Make Top-Down stress testing for all banks an integral part of the systemic risk monitoring toolkit. (NBB) — MT; Priority H
  - Develop in-house models in coordination with ECB to assess financial stability risks from solvency and liquidity interactions and contagion from domestic and cross-border counterparties. (NBB) — MT; Priority H
  - Expand the set of macroprudential stress tests in the insurance sector by introducing multi-period scenarios and projections. (NBB) — ST; Priority M
- Liquidity and intra-group vigilance:
  - Strengthen vigilance on liquidity pressures from non-maturing liabilities, derivatives, intersectoral flows, intragroup flows, and cross-border operations. (NBB) — I; Priority H
  - Continue to monitor asset encumbrance and risks from legacy portfolios. (NBB) — I; Priority M
- Insurance supervision and monitoring:
  - Use stress testing results to challenge companies’ ORSA and underlying projections for future business. (NBB) — ST; Priority M
  - Intensify monitoring of the mortgage loan portfolio of insurers and respective underwriting standards. (NBB) — I; Priority H
  - Continue close monitoring of the use and impact of long-term guarantee measures, particularly the volatility adjustment (VA). (NBB) — ST; Priority M
  - Set up comprehensive monitoring of risks arising from mortgage lending by insurers; analyze LTV and DTI metrics and explore regulatory arbitrage within conglomerates.
- Cross-sectoral and macroprudential recommendations:
  - Develop macroprudential stress tests incorporating feedback loops with the real economy for policy calibration.
  - Incorporate changes in policyholder behavior and companies’ assumed asset allocation shifts into cross-sectoral analysis.
  - Assess need for Pillar 2 add-ons if Solvency II standard formula inadequately captures insurer risk concentration (e.g., mortgage loans or sovereign bond concentration).

### Key quantitative figures and parameters (selected)
- Banking sector:
  - Size of largest six banks (end-2016): €917 billion.
  - RWAs: €349 billion.
  - Aggregate CET1 (end-2016): 15.1 percent.
  - Aggregate CET1 low point (2018, adverse): 11.4 percent.
  - Aggregate CET1 (2021, adverse): 13.2 percent.
  - Aggregate CET1 decline at low point: 370 basis points.
  - RWAs rise by 290 bps over first two years (adverse).
  - Near-term CET1 contributions: credit losses 140 basis points; valuation losses 210 basis points.
  - Mortgage pre-payment rates: around 22 percent in 2015–2016; around 14 percent in 2017.
  - IFRS9 expected increase in accounting provisions: around 20 percent (EBA Q1 2016 assessment); around 13 percent (Q4 2016 assessment).
  - Deposit funding share: 85 percent deposit funded; 15 percent debt funded (variation across banks 75 percent–95 percent).
- Liquidity:
  - Average LCR ratio (March 2017): around 140 percent.
  - Asset encumbrance (2016): 13.7 percent of total assets.
- Insurance sector:
  - Median solvency ratio pre-stress: 184 percent.
  - Median solvency ratio — IMF adverse: 124 percent.
  - Median solvency ratio — NBB low-for-long: 145 percent.
  - Aggregated solvency surplus pre-stress: €10.2 billion; IMF adverse: €4.1 billion; NBB low-for-long: €6.2 billion.
  - Average guaranteed interest rate as of end-2016: 2.6 percent.
  - Fixed-income portfolio maturing 2017–2025: 50 percent, average coupon 3.6 percent; reinvestment assumption at 1 percent → average coupon 2.5 percent by 2025 and 1.9 percent in 2030.
- Scenario shocks (selected layer and tail shock magnitudes):
  - Equity risk premium shocks (Layer 1): China, Euro Area, Japan, United Kingdom, United States −20 percent.
  - Duration risk premium shocks (Layer 1): High Spread EA +200 basis points; Low Spread EA +50 basis points; Japan/UK/US +100 basis points.
  - IMF adverse scenario: Belgian real estate prices −20 percent; long-term yields rise 150 basis points by 2018; money market rates widen by 120 basis points.
  - Banking sector sovereign shock (adverse): Belgian OLO rates +140 basis points in 2017 and +69 basis points in 2018 relative to Bund increases of 70 bps and 44 bps respectively.
  - Corporate spreads (adverse): rise by 358 basis points in 2017 and 179 basis points in 2018.

*International Monetary Fund — content unit cr1869 (extracted section).*

### EXECUTIVE SUMMARY __________________________________________________________________________ 7

### EXECUTIVE SUMMARY

### Overview
- Belgium’s financial landscape has changed since the global financial crisis (GFC) and the European sovereign crisis: the banking system has contracted driven by restructuring operations in entities that received government support; banks have adopted more traditional business models, with greater emphasis on domestic lending and deposit funding; the insurance sector has seen some consolidation and a gradual move away from traditional insurance products to asset management-type products; the scale of cross-border linkages in the financial system, while still significant, has declined markedly.
- The Financial Sector Assessment Program (FSAP) stress testing exercise comprises a comprehensive analysis of solvency and liquidity risks in the Belgian banking and insurance sectors.
  - One set of tests used macroeconomic scenarios to capture the impact of a drastic deterioration in macrofinancial conditions on the solvency of banks and insurance companies.
  - A second batch of tests used hypothetical deteriorations in liquidity and asset market conditions to gauge the impact of these developments on individual entities as well as on the risk of contagion. This second batch of tests was conducted only on banks.

### Key macrofinancial background and vulnerabilities
- Growth and outlook:
  - Real GDP grew by 1.2 percent in 2016 and is expected to accelerate to 1.7 percent in 2017.
  - The output gap is closing and is expected to turn positive in 2018.
  - Absent further structural reforms and stronger growth in Europe, growth is expected to remain at around 1.5 percent for the foreseeable future.
- Financial cycle and credit:
  - Bank credit growth has accelerated since 2015 and a small positive credit gap opened in 2016.
  - Credit growth is driven by residential mortgages and, to a lesser extent, credit to the construction sector.
  - The financial cycle in Belgium appears to be running ahead of the economic cycle, increasing the need for timely detection of mispricing of risk and excessive risk taking and for timely deployment of macroprudential policies.
- Structural changes in the banking sector:
  - Banking system assets shrunk from 470 percent of GDP in 2008 to 250 percent today, mainly due to scaling back of cross-border activities and NBB regulations limiting Belgian subsidiaries’ exposure to parent companies.
  - Belgium’s largest and fourth largest banks are subsidiaries of EA significant institutions.
  - Key crisis-era restructurings and support:
    - The Belgian government purchased for € 4 billion the Belgian subsidiary of Dexia Group (later rebranded as Belfius).
    - KBC Group received a total of € 7 billion of capital injections in 2008-09 and has entirely repaid the public support and started distributing dividends.
  - Market shares: BNP Paribas Fortis has a 30 percent share of total system assets; KBC banks have a 24 percent share; Belfius has a 15 percent share.

### FSAP stress testing: scope and approach
- Macroprudential stress tests were conducted to assess the financial system’s ability to withstand losses without amplifying shocks to the real economy.
- Tests covered:
  - Bank and insurance solvency under macroeconomic stress scenarios.
  - Bank liquidity and market stress using hypothetical deteriorations; contagion analysis was performed, including interbank channel contagion.

### Main solvency and liquidity findings
- Overall resilience:
  - Stress tests confirm that banks and insurance companies can absorb credit, sovereign, and market losses in the event of a severe deterioration in macrofinancial conditions.
  - All banks meet minimum capital requirements and none needs to draw down its capital conservation buffer over the stress horizon.
  - The risk of interbank contagion through direct exposures is low.
  - Insurance companies are generally resilient; losses in stress scenarios by insurers belonging to banking groups do not threaten the soundness of those groups.
- Solvency ratios (insurance sector median under scenarios):
  - In the IMF adverse scenario, the median solvency ratio drops from 184 to 124 percent.
  - In a low-for-long scenario, the median solvency ratio drops to 145 percent.
  - In each scenario, one out of eight companies drops marginally below a solvency ratio of 100 percent, but these shortfalls could rather easily be restored.
- Drivers of resilience:
  - Banks: relatively healthy loan portfolios and limited exposure to market and liquidity risks.
  - Insurers: sound solvency levels and reduced exposures to guaranteed rates.
- Risks identified that require monitoring:
  - Selected risky portfolios and idiosyncratic liquidity and concentration risks in certain banks.
  - Concentrated exposures towards domestic sovereign bonds among insurance companies.
  - Liquidity pressures from non-maturing liabilities, intersectoral flows, intragroup transactions, and cross-border operations.

### Policy recommendations (timing and priority preserved in intent)
- Banking sector recommendations:
  - Make Top-Down stress testing for all banks an integral part of the systemic risk monitoring toolkit. (NBB) — MT; Priority H
  - Develop in-house models in coordination with ECB to assess financial stability risks from solvency and liquidity interactions and contagion from domestic and cross-border counterparties. (NBB) — MT; Priority H
  - Strengthen vigilance on liquidity pressures from non-maturing liabilities, derivatives, intersectoral flows, intragroup flows, and cross-border operations. (NBB) — I; Priority H
  - Continue to monitor asset encumbrance and risks from legacy portfolios. (NBB) — I; Priority M
- Insurance sector recommendations:
  - Expand the set of macroprudential stress tests in the insurance sector by introducing multi-period scenarios and projections. (NBB) — ST; Priority M
  - Use stress testing results to further strengthen microprudential supervision by challenging companies’ Own Risk and Solvency Assessment (ORSA) and underlying projections for future business. (NBB) — ST; Priority M
  - Intensify monitoring of the mortgage loan portfolio of insurers and respective underwriting standards. (NBB) — I; Priority H
  - Continue the close monitoring of the use and impact of long term guarantee measures, in particular the volatility adjustment. (NBB) — ST; Priority M

### Institutional guidance
- The NBB is encouraged to expand its stress testing capabilities to strengthen monitoring of financial stability risks:
  - Develop a top-down stress testing framework for macroprudential purposes to assess financial stability risks from systemically important banks.
  - Strengthen vigilance on liquidity pressures from non-maturing liabilities, intersectoral flows, intragroup transactions, and cross-border operations.

*INTERNATIONAL MONETARY FUND — EXECUTIVE SUMMARY*

### 5.      Banks have reoriented their business models towards traditional financial activities but

### 5.      Banks have reoriented their business models towards traditional financial activities but 

### Reorientation of banks and insurers — key facts
- Banks’ exposures to the non-financial private sector increased from 40 percent of their assets in 2008 to about 50 percent in 2016.
- Size of banks’ bond portfolio declined by half between 2008 and 2016.
- Deposit funding rose from under 40 percent of total liabilities to 55 percent in 2016.
- Internationally active banks are concentrated on core markets: Czech Republic, France, Ireland, Luxembourg, Netherlands, and Turkey, with very limited exposure outside them.
- Some banks are shifting to digital banking platforms to rationalize networks, integrate branches, and reduce costs.
- Total assets of the insurance sector increased by about 12 percentage points of GDP over 2012-16 and represented 80 percent of GDP in 2016.
- The increase in insurance sector assets largely reflects the introduction of market-based valuation of assets under Solvency II.
- 12 percent of licensed insurance firms exited the sector since 2013.
- Several life insurers announced they will stop selling life insurance contracts with guaranteed rates.
- Several insurers moved from traditional insurance products to asset management type products.
- Annual premiums in the life business declined by 31 percent between 2012 and 2016.

### Banking sector balance sheet strength and vulnerabilities
- Capital ratios increased: average CET1 ratio rose by 200 basis points since 2012 to 15.7 percent in 2016.
- Non-performing loans reduced to 3.4 percent of gross loans in 2016.
- Loan loss provisioning rates estimated very low, reflecting improved macro conditions in foreign markets.
- Mortgage lending expansion:
  - Mortgage loans represent around 20 percent of banks’ balance sheet at end-2016.
  - Banks tightened lending policies over 2012–14, but a further build-up of high-risk segments occurred in the most recent two years, increasing vulnerability to Belgian real estate deterioration.
- Profitability pressures:
  - Net interest margins increased since 2014 despite falling interest rates because banks’ funding costs adjusted faster than asset yields.
  - Profitability challenges going forward due to large waves of mortgage repayments and refinancing at lower fixed rates in 2014-2016, and the lower bounds hit by funding costs.
  - Introduction of IFRS9 in 2018 will likely lead to higher provisions.
    - Based on the EBA impact assessment launched in the first quarter of 2016, IFRS9 is expected to increase accounting provisions by around 20 percent on average.
    - This estimate came down to around 13 percent in the second impact assessment launched in the last quarter of 2016.
  - Gradual implementation of additional capital requirements through 2019 will further squeeze profits.
    - Capital conservation buffer introduced in January 2016, starting off at 0.625 percent and rising to 2.5 percent by 2019.
    - Countercyclical capital buffer introduced in January 2016 set currently at 0 percent for Belgian exposures.
    - O-SII capital surcharge being gradually incremented from 0.5 percent (bucket 1) and 0.25 percent (bucket 2) to 1.5 percent and 0.75 percent, respectively as of January 2018.
- House loan pre-payment rates:
  - In 2015–2016, house loan pre-payment rates in Belgium reached around 22 percent, compared to a euro area average of 12 percent.
  - This rate decreased to around 14 percent in 2017.
- Risks from abrupt reversal of compressed risk premia: an abrupt and sustained increase in short-term and long-term interest rates would increase funding costs which could outpace repricing of assets and compress margins if banks are not adequately hedged.

### Insurance sector risks and concentrations
- Shift towards asset management-type products has changed insurers’ risk profile:
  - Interest rate risks from duration mismatches have declined due to better asset-liability matching.
  - Shift away from guaranteed products reduced exposure to interest rate and other market risks but increased liquidity risk as policyholders might redeem contracts, especially in times of rising interest rates.
  - Purchases of mortgage loan portfolios by some companies further expose them to liquidity risk.
- Profitability impact:
  - Sector’s return on equity averaged 20 percent before the financial crisis, declined to around 8-10 percent in 2015/16.
  - Returns tend to be higher for composite insurers pursuing both life and non-life businesses.
- Asset concentration:
  - Sovereign bonds account for 43 percent of total assets of Belgian insurers.
  - More than half of these sovereign bonds are Belgian sovereign bonds.
  - Real estate exposures are high and come mainly in the form of mortgage loans issued in the Belgian and Dutch market.
  - Some insurers in financial conglomerates have concentrated exposures towards banks, specifically deposits within the same group.

### Key macrofinancial risks identified for the banking sector (RAM)
- A sudden increase in general risk aversion in global markets, leading to higher money market rates, a steepening of the yield curve, reductions in market liquidity, broad-based financial market dislocation, lower securities valuations and bank equity values, and slower economic activity.
- A large correction in the Belgian real estate market, leading to direct bank losses and potential weak consumption from household deleveraging; depressed mortgage loan valuations affecting asset encumbrance and funding costs.
- A reassessment of regional sovereign risk, with potential fiscal stress in the euro area triggered by political uncertainty or debt sustainability concerns, weakening banks’ balance sheets and access to wholesale funding.
- A prolonged period of low growth and low interest rates in the euro area (“low-for-long” scenario) with significant impact on interest margins and profits, exacerbated by credit losses, depressed corporate earnings, weaker stock returns, increased market leverage, and higher corporate defaults.

### Common macroeconomic scenarios and stress-test design
- Scenario horizon: starts in the first quarter of 2017 and extends through the fourth quarter of 2021.
- Baseline scenario based on the October 2017 World Economic Outlook (WEO).
- Projections include ten key macrofinancial variables in Belgium and ten material geographies for Belgian banks, including Bulgaria, Czech Republic, France, Hungary, Ireland, Luxembourg, Netherlands, Slovakia, Switzerland, Turkey, other EA countries, and other emerging markets.
- Core projected variables: real and nominal GDP, inflation, unemployment, short-term rates, long-term rates, public debt, and foreign exchange fluctuations.
- Adverse stress scenario covers the first three risks identified in the RAM; the low-for-long scenario assessed using single-factor shocks.
- Scenario calibrated using IMF’s Global Macrofinancial Model (GFM).
- Stress scenario design: layers-of-shocks approach (Layer 1 through Layer 6). Key layer assumptions and magnitudes (expressed as deviations from April 2017 WEO baseline):
  - Layer 1: Tightening of financial conditions in systemic economies, 2017Q1 – 2018Q2
    - Long term government bond yield; Duration risk premium shocks: High Spread Euro Area Economies +200 basis points; Low Spread Euro Area Economies +50 basis points; Japan, United Kingdom, United States +100 basis points
    - Real equity price; Equity risk premium shocks: China, Euro Area, Japan, United Kingdom, United States −20 percent
    - Money market interest rate spread; Credit risk premium shocks: China +100 basis points; Euro Area, Japan, United Kingdom, United States +75 basis points
    - Real bilateral exchange rate; Currency risk premium shocks: Euro Area +5.0 percent
  - Layer 2: Fiscal consolidation in the Euro Area, 2017Q1 – 2019Q2
    - Primary fiscal balance ratio; Fiscal expenditure shocks: High Spread Euro Area Economies +2.0 percentage points; Low Spread Euro Area Economies +1.0 percentage point
  - Layer 3: Credit cycle downturns in emerging market economies, 2017Q1 – 2019Q2
    - Loan default rate; Loan default shocks +0.0 to +6.4 percentage points
  - Layer 4: Suppressed economic risk taking worldwide, 2017Q1 – 2021Q4
    - Private investment; Investment demand shocks −4.0 percent
    - Private consumption; Consumption demand shocks −1.0 percent
  - Layer 5: Domestic Layer: Sovereign risk, funding stress, and housing correction in Belgium, 2017Q1 – 2019Q4
    - Long term government bond yield; Duration risk premium shocks +100 basis points
    - Primary fiscal balance ratio; Fiscal expenditure shocks +1.0 percentage point
    - Wholesale funding spread, Credit risk premium shocks +50 basis points
    - Real housing price correction, peak-to-trough −20 percent
    - CRE price correction, peak-to-trough −30 percent
    - Private investment; Investment demand shocks −8.0 percent
    - Private consumption; Consumption demand shocks −2.0 percent
  - Layer 6: Regional Layer in Germany and selected Emerging European Economies, 2017Q1 – 2019Q4
    - Private investment; Investment demand shocks −4.0 percent
    - Private consumption; Consumption demand shocks −1.0 percent
    - Market pressures in Selected Emerging European Economies (CZE, POL, TUR)
      - Long term government bond yield; Duration risk premium shocks +100 basis points
      - Real equity price; Equity risk premium shocks −10 percent
      - Real bilateral exchange rate; Currency risk premium shocks (only TUR) +20.0 percent

### Adverse scenario outcomes and severity
- Under the adverse stress test scenario:
  - Belgium experiences nine quarters of negative growth rates peaking at −2.8 percent in 2018Q1.
  - Output falls 8.1 percent below baseline by 2019.
  - Consumption price inflation falls 3.5 percentage points below baseline by 2020.
  - Money market rates widen by 120 basis points.
  - Long-term yields rise by 150 basis points by 2018.
  - Real equity prices fall by 20 percent.
  - Real estate prices collapse by 20 percent.
- International component:
  - Output loss in high spread euro area economies concentrated at −9.1 percent below baseline.
  - CESEE emerging markets experience around a 7.0 percent output loss at the low point in 2019.
- Severity comparison:
  - Severity of the adverse macro-financial scenario is comparable to the 2016 EU-wide stress testing exercise but generates greater market volatility from asset price adjustments.
  - The stressed GDP path constitutes a 3.1 standard deviation move in the two-year cumulative real GDP growth rate by 2018.

*Source: cr1869 - 5.      Banks have reoriented their business models towards traditional financial activities but*

### 22.      Macroprudential stress tests were conducted to assess the Belgian financial system’s

### 22. Macroprudential stress tests were conducted to assess the Belgian financial system’s ability to withstand losses without amplifying shocks to the real economy

### Purpose and scope
- Objective: ensure banks have the capacity to absorb shocks without engaging in deleveraging or amplification of market and liquidity stress triggered by:
  - rising cyclical vulnerabilities from rising leverage, abrupt asset prices corrections, or deterioration of underwriting standards in mortgages; and
  - structural shocks from shift towards more traditional business models, reduced activities in the Belgian interbank market with interbank activities conducted with foreign counterparties, and greater reliance on non-maturing domestic deposits.
- Coverage:
  - IMF stress test covered the six major Belgian SIs, representing around 90 percent of total banking system assets.
  - The six major Belgian SIs are Argenta, AXA Bank Belgium, Belfius Banque, BNP Paribas Fortis, ING Belgium, and KBC Group.
  - Assessment conducted at the highest level of consolidation in Belgium including the insurance arm for KBC Group and Belfius and the consolidation of banking associates and foreign subsidiaries.
  - NBB analysis covered the same sample at a solo level, excluding foreign subsidiaries.

### Data, limitations, and caveats
- Data sources and consistency:
  - FSAP stress test results based on end-2016 supervisory data, complemented by NBB’s pre-June 2014 supervisory data for a core set of variables (e.g., default rates, funding cost by instrument, net trading income).
  - Post-2014 SSM reporting templates were merged with publicly available sources for other variables (e.g., banks’ sovereign exposures drawn from annual reports).
  - ECB shared supervisory returns including ITS files (FINREP and COREP) and STE files (maturity ladder) for the solvency test.
- Key data challenges and modelling caveats:
  - Matching and reconciliation of risk data extracted from multiple sources is complex.
  - Use of proxy variables where direct data unavailable (e.g., duration of fixed income proxied by remaining maturity from the 2016 EBA transparency template; PDs projected drawing on EDF series).
  - EDF series tend to display greater volatility than regulatory PDs and are based on a country-based representative portfolio.
  - Stress test scenarios calibrated using historical data identify extreme “tail events” from historical distribution, but crises may involve unanticipated shocks and unexpected interrelationships not reflected in history.

### FSAP stress testing strategy (four-pronged approach)
- Solvency stress test:
  - Fully-fledged scenario-based test and range of sensitivity tests to assess capital adequacy.
  - Macrofinancial scenario explored domestic and global risks and included a traded risk scenario to capture business risk from tighter financial conditions, changes in market valuations, and ALM vulnerabilities.
- Liquidity stress tests:
  - Two complementary frameworks:
    - Liquidity Coverage Ratio (LCR)-based approach implemented under four different scenarios.
    - Implied cash-flow test applied under six alternative scenarios, simulating stress factors over eighteen maturity buckets from one-day through twelve months.
- Contagion module (interbank bilateral exposures):
  - Examined transmission of shocks via Belgian banks’ bilateral matrix of large exposures, including intragroup transactions.
  - Amplification mechanisms through fire sales and cascading defaults assessed using network analysis on the Belgium interbank market.
- Systemic risk module (market-based contagion):
  - Assessed contagion through equity markets using a CoVaR approach to evaluate correlated losses through equity markets unrelated to credit or funding exposures.

### Scenario specification and major shocks
- Sovereign yields:
  - Sovereign shocks calibrated for twenty-three sovereigns to which Belgian banks were exposed at end-2016.
  - Baseline: shocks to the risk-free rate proxied by the 10-year Bund increase by an average 36 basis points per year close to the average 37 basis points for the Belgian OLO rate.
  - Adverse scenario (Belgian OLO): rates rise by 140 basis point in 2017, and a further 69 basis point in 2018, relative to Bund rate increases of 70 basis points in 2017 and 44 basis points in 2018.
  - Duration of sovereign holdings proxied by residual maturity reported in 2016 EBA transparency templates.
- Corporate yields:
  - Corporate yields proxied by Moody’s Baa Index.
  - Baseline: spreads rise in line with shifts to the OLO curve.
  - Adverse: spreads rise by 358 basis points in 2017, and 179 basis points in 2018.
  - Duration of corporate securities proxied by average maturity of all active corporate bonds as of October 2017, yielding an average duration of 4 years weighted by face value.
- Money markets, real estate, and credit:
  - Adverse scenario assumes abrupt decompression of asset risk premia, low secondary market liquidity, and re-emergence of euro area financial stress.
  - Euribor spreads over Bund T-bill rise around 100 bps over the stress test horizon.
  - Projected path for money market rates in Belgium reaches a peak in 2018 with a 120bps increase relative to baseline.
  - Real residential property and real commercial real estate prices fall by 20 percent peak-to-trough in the adverse scenario.
  - Credit to the private sector reaches negative rates starting in Q1 2018 at a peak of -4.45 percent.
- Equity markets and volatility:
  - Heightened volatility reflected in equity markets (volatility path generated for Eurostoxx 50, BEL-20, and S&P 100).
  - Changes in investors’ risk preferences and aggregate liquidity expected to influence portfolio allocation and willingness to bear sovereign risk.

### Bank solvency, exposures, and regulatory hurdle
- System-wide figures and risk composition:
  - Size of the largest six Belgian banks at end-2016: €917 billion.
  - RWAs: €349 billion, representing around 40 percent risk density.
  - Over 85 percent of capital is required to cover unexpected losses from credit risk.
  - Market risk represents just under 2 percent of capital requirements.
- Regulatory hurdle (“hurdle rate”):
  - Hurdle rate set at the fully loaded Pillar I minimum common equity, the fully loaded O-SII buffer, and the fully implemented capital conservation buffer.
  - CET1 hurdle rate: 8.5 percent for the Big 4 and 7.75 percent for the remaining two O-SIIs.
  - Hurdle rate also includes a 3 percent Tier 1 ratio based on the CRR leverage framework.

### Credit risk modelling approach
- Role of credit risk:
  - Credit risk accounts for the largest regulatory capital requirement for Belgian banks.
- Modelling methodology:
  - Separate credit risk analysis calibrated for five Basel asset classes, two regulatory books (IRB and STA), and eleven material geographies: Belgium, Bulgaria, Czech Republic, France, Hungary, Ireland, Luxembourg, Netherlands, Slovak Republic, Switzerland, Turkey.
  - For IRB exposures, credit risk evolves with EaD, PD, and LGD.
  - For STA exposures, credit risk reflected in higher provisioning rates and credit risk migration to substandard categories.
- IRB usage and portfolio granularity:
  - Aggregate non-defaulted EaD for large Belgian banks reached €876 billion in 2016, of which 90 percent was booked under the IRB approach.
  - The split is the same for exposures to central governments or central banks which represent around fifth of total exposures.
  - Regulatory credit risk parameters (PD, LGD, EaD) for approved IRB models are “hybrid” estimates lying between ‘pure PiT’ and ‘pure TTC’ estimates.
- Provisioning calculation:
  - Impact of stress on regulatory capital through increased provisions computed as the level of expected losses: EAD * PD * LGD (notation provided in source where i denotes bank, j denotes asset class, and t is the time dimension).

### Practical considerations and interpretation
- FSAP stress test results should be interpreted with caution due to:
  - Data vintage (end-2016 supervisory data) and need to merge multiple reporting templates and public sources.
  - Use of proxies for several risk parameters and volatility in proxy series (e.g., EDFs).
  - Historical calibration of tail events which may not capture unanticipated or nonlinear crisis dynamics.

*Source: IMF staff (chapter content provided).*

### 40.      The FSAP team used a two-step process to project stressed PDs. The approach of the

### cr1869 - 40. The FSAP team used a two-step process to project stressed PDs

### Methodology overview
- The FSAP team used a two-step approach:
  - Based on the stress test scenario, a time series of aggregate PD proxies was projected by portfolio and geography using market-based PDs. These series are available at the monthly frequency and cover all material geographies and feed the FSAP team satellite models for credit risk.
  - A statistical approach was used to forecast bank-specific PDs using post-June 2014 supervisory data to adjust aggregate projections to banks’ IRB regulatory PDs.
- A regression analysis of bank-level regulatory PDs on country-level EDFs was conducted to adjust the path of projected EDFs at the bank level.

### PD proxies by portfolio and geography
- Corporate and retail unsecured exposures:
  - PD proxy obtained from Moody’s Analytics using the average one-year expected default frequency (EDF) by geography and portfolio.
  - Portfolios used by geography: corporate group, construction and real estate development group, financials group, consumer nondurables and services group.
  - These categories were mapped to COREP portfolios: corporate and SME exposures, specialized lending exposures, exposures to institutions, and retail unsecured, respectively.
  - Analysis used monthly data over January 2005 through December 2016.
- Exposures to central governments:
  - PD proxy extracted from sovereign yields using the Merton approach.
  - FINREP data on banks’ securities portfolios used to back out implied COREP exposure to sovereigns from central bank exposure.
  - Composition of sovereign portfolio from banks’ annual reports for AFS and HTM portfolios (subject to impairment charges) used to proxy breakdown of sovereign exposures by issuer.
- Sovereign PD extraction from spreads:
  - A reduced-form structural model extracted PD estimates from sovereign spreads projected under the scenario.
  - The approach assumes the difference between a risk-free security and a risky security is the put option on the value of the assets which includes the loss induced by the stressed PD and LGD of the bond.
  - Using the credit spreads for sovereign i linked to the scenario S_{i,Tt}, time to maturity (T-t), and assuming LGD=45 percent, the implied risk-neutral PD is backed-out as (expression in source).
- Mortgages (retail exposures secured by real estate):
  - PD shifts were taken from the ECB staff modeling tool (STAMP€).
  - FSAP did not have time series of default rates for mortgages from credit registers; instead relied on ECB estimates using data provided by national competent authorities in the 2016 EBA stress testing exercise and calibrated under the FSAP scenario.

### Statistical modeling and transformations
- To address truncated distribution of default rates, a logit transformation was applied prior to analysis to ensure projected rates remain within the 0-1 bound.
- The FSAP team ran a battery of over 300 credit risk specifications to obtain PD projections covering eleven geographies using multiple approaches:
  - (i) panel-based regression by geography using fixed effects and robust standard errors;
  - (ii) dynamic panel-based regression using Arellano and Bond lagged structure as instrumental variables;
  - (iii) individual portfolio-based estimation by geography;
  - (iv) dynamic individual estimation by portfolio and geography with robust errors.
- Each approach was tested with lag structures between one and three lags and alternative specifications capturing country, regional, and global factors.
- Model selection criteria included goodness of fit, forecasting performance, and expert judgment based on behavior during the 2008-09 global financial crisis and the 2012 European sovereign debt crisis.
- Estimated projections were applied over the forecast window 2017Q1-2021Q4 under each scenario.
- The estimation period cited for several models is 2005Q1 through 2016Q4 (and for Moody’s EDF monthly series, January 2005 through December 2016).

### Determinants and drivers of corporate credit risk
- For Belgian corporate exposures:
  - Credit risk is heightened by stressed conditions in Belgium, the EA, and global markets.
  - Higher unemployment and a credit crunch contribute to rising corporate EDFs.
  - Default rates increase with deteriorating financial conditions in euro area debt sovereign markets captured by soaring spreads of 10y OLO rates to Bund rates.
  - Strains in money markets captured by widening spreads of USD Libor over T-bill rates also contribute.
- For overseas exposures (example: Turkey):
  - Key drivers include rising inflation and currency depreciation, weakening of growth outlook in emerging markets, and tightened global money market conditions.

### Application to bank-level PDs, LGDs, and EADs
- Shifts to PDs were applied to the FSAP team’s estimated non-defaulted bank regulatory PDs using COREP supervisory data combined with geographical breakdowns of exposures by residence of obligor to obtain average PDs and LGDs by portfolio.
- Adjustments were required because not all exposures are assigned to obligor grades or pools by obligor; key assumption: distribution of credit quality steps of exposures not subject to obligor grades has the same structure as exposures included in the obligor pool.
- Adverse LGDs:
  - Calibrated for the first three years of the scenario by portfolio under the adverse scenario provided by the FSAP team.
  - LGD projections were derived by bank, using estimated non-defaulted LGD ratios and shifts to LGD by portfolio, where j denotes portfolio, i denotes bank, and t is time (formula as provided in source).
- EAD projection drivers and formula:
  - Projection of EAD driven by balance sheet assumptions, structural FX risk in foreign geographies, and triggered credit lines and guarantees.
  - Changes to EAD in the IRB portfolio governed by the provided expression in the source:
    - variables include ti g, the growth rate of the IRB portfolio; j FXi f , fraction of foreign currency loans; t FX Δ is the shock to foreign currency under the macro scenario; (j ti PD 1, 1 − ) represents the non-defaulted portfolio; j ti L Δ is the shock to triggered credit lines and guarantees; j ti UCL 1, − is the amount of undrawn guarantees.
  - To calibrate shock to triggered credit lines and guarantees, FSAP used the maximum increase in the off-balance sheet exposures to EAD ratio for corporates reported by banks in their Pillar 3 disclosures.

### Balance sheet and scenario assumptions
- FSAP scenario assumed constrained dynamic balance sheets by geography, excluding banks’ management actions.
  - Under the FSAP scenario, credit growth determined by credit demand shocks triggered by negative consumption and investment shocks that lead to a slowdown of credit under the adverse scenario.
  - Credit supply effects or banks’ management actions to restore capital ratios were disallowed.
- Given geographic footprints, growth rate of credit computed as the EAD-weighted nominal GDP growth across relevant jurisdictions.
- Data on FX structural effects projected under assumption that loans are denominated in domestic currency of the overseas exposure.

### Capital requirements and regulatory parameter treatment
- To compute capital requirements, ‘hybrid’ regulatory risk parameters were projected and applied to Basel III formula for IRB exposures.
- RWAs derivation depends on PD, LGD, EAD, correlation assumptions, and effective maturity for each exposure.
- According to the Basel III framework:
  - RWAs were computed after applying the scaling factor of 1.06 to credit RWAs.
  - A multiplier of 1.25 was applied to the correlation parameter of all exposures to large regulated financial institutions and to all unregulated financial institutions.

### STA exposures, NPLs, and provisions
- Expected losses for STA exposures are expected to rise due to migrations from performing loans to NPLs and migration effects within performing and non-performing buckets.
- NPL definition includes past due loans over 90 days as well as defaulted loans; amount included loans and receivables as well as HTM investments.
- Provisions projected for specific and collective allowances for incurred but not reported losses.
- Credit impairments are reported under FINREP and not disaggregated between IRB and STA exposures, which may introduce biases for internationally active banks.
- Credit impairments for NPLs were estimated using supervisory data and calibrated to banks’ core markets (Table 5).
  - Data sources merged: post-June 2014 SSM data, 2008Q1–2014Q2 NBB data, and Fitch data over 2005Q1–2007Q4.
  - A range of bank-specific regressions was run with many specifications and lag structures; panel estimations with fixed effects run for robustness.
  - A logit transformation on NPL values was applied to ensure projected rates lie within the [0-1] range.

*International Monetary Fund — content unit cr1869 (extracted section).*

### 57.      To compute the impact of migration on capital requirements, the FSAP team used a

### cr1869 - 57.      To compute the impact of migration on capital requirements, the FSAP team used a

### Impact of migration on capital requirements (methodology)
- Three-pronged approach:
  - The risk weight of NPL was set at an average 100 percent, following Basel III’s regulatory framework which specified that a 100 percent risk weight should be applied on past-due loans when specific provisions are no less than 20 percent of the outstanding amount of the loan.
  - The difference between 100 percent and the average risk weight of each STA exposures (excluding sovereign and central bank exposures) was multiplied by the nominal amount of NPLs projected under the scenario to forecast the increase in risk weights attributed to the STA exposures in default.
  - The non-defaulting portfolio was assumed to downgrade one notch under the adverse scenario. The asset-weighted risk weight of Belgian banks’ STA exposures (net of sovereign and central bank claims) corresponds to a Basel III credit assessment for corporate claims between A+ to A- (50 percent) and BBB+ to BB- (100 percent). A one-notch downgrade in the BBB+ to BB- category represents an increase of 8.3 percent in the average risk weight of the non-defaulted portfolio.

### Drivers of capital requirements for STA exposures
- Capital requirements for STA exposures were driven by:
  - Changes in provisioning rates.
  - Growth of EAD.
  - Structural FX risk.
  - Triggered credit lines and guarantees.
  - Migration effects.
- Four main components driving shifts to RWAs in the STA portfolio:
  - Component 1: motion of RWAs generated by the flow of provisions, the growth rate of the portfolio, and FX effects.
  - Component 2: increase in risk weights resulting from triggered off-balance sheet credit lines and guarantees.
  - Component 3: increase in risk density from the transition of loans from the performing to nonperforming category.
  - Component 4: change in risk density from the transition matrix estimated for performing exposures.

### Coverage ratio projection
- Coverage ratios projected using a joint econometric approach and historical evidence:
  - Baseline scenario: end-2016 reported ratios by bank were used.
  - Adverse scenario: coverage ratio estimates forecast based on econometric projections and using the floor of Basel findings on the behavior of coverage ratios during the global financial crisis which stood at an average 65 percent.

### Market risk approach — Fair valuation impact (securities and sovereigns)
- Valuation losses on trading assets, financial assets designated at fair value, and available-for-sale (AFS) securities were estimated using a granular full revaluation approach on all debt securities held in Belgian banks’ portfolios.
- Sovereign portfolio: granular analysis included all sovereign exposures by issuer and regulatory book; Belgian banks are exposed to sovereigns domiciled in 23 jurisdictions.
- AFS portfolio composition (aggregate, end-2016):
  - Belgian government sovereign exposure: 40 percent of AFS portfolio.
  - Foreign government sovereign exposure: 37 percent of AFS portfolio.
  - Under one fifth of total exposures to financial and non-financial corporates (corporate exposures proxied using the Moody’s Baa corporate index).
- Valuation losses under stress were transferred to regulatory capital through the statement of other comprehensive income (OCI).
- Asset duration and repricing:
  - Corporate portfolio average duration used as proxy: 4 years.
  - Repricing risk benchmark: 10-year Bund rate.
  - Under the adverse scenario, the 10-year Bund rate increases by 70bps in 2017 (relative to the initial point) and a further 44bps in 2018.
- Securities portfolio evolution under constrained balance sheet assumption: notional values grow according to provisions, growth of interest-bearing assets, currency composition, and FX shocks; no portfolio rebalancing or liquidation of positions allowed throughout the stress test horizon.

### Derivatives, counterparty risk, and valuation
- Size and evolution of derivative portfolio:
  - Notional amount declined from as high as 8 times the balance sheet at end-2008 to € 3.2 trillion (or 3.2 times total on-balance sheet assets of the sector) by 2016.
  - In terms of market value, relative size shrank from 16 percent of the balance sheet to 6 percent at end-2016.
- Accounting and hedge features:
  - About 85 percent of derivatives are classified as IFRS accounting category “held for trading”.
  - Of the derivative portfolio, 10 percent are identified in prudential reporting as economic hedges and another 10 percent as macro-hedges used to hedge a specific portfolio.
- Counterparty credit risk concerns:
  - Potential difficulty hedging open positions if a counterparty fails, including intra-group counterparties.
  - CVA capital charges around 1.5 percent of RWAs reflect these concerns.

### Traded risk stress test mechanics
- Loss calculation for fixed income instruments uses modified duration multiplied by shifts to credit spreads (shocks shown in Table 4 of source).
- Hedges for interest rate risk assumed ineffective under stress; fair value of hedges not stressed under the adverse scenario.
- Disallowed flight-to-quality effects in bond markets to increase scenario severity.

### Other risk factors
- Market risk from FX and commodity price shocks considered negligible for Belgian banks (no material short or long net positions under standardized approach).
- Basis risks and optionality risks excluded due to data limitations:
  - Basis risk sources include diverging paths between interbank rates and sight/saving deposit rates, and differing reference indices (OLO-rates vs swaps) for mortgage repricing and hedging.
  - Optionality risk from early mortgage prepayment penalties legally limited to three months of interest; risk reduced after the 2014-16 refinancing wave.

### Net Trading Income (NTI)
- NTI projections reflect size of derivative books; fair value of trading derivatives reached 85 percent of trading assets at end-2016.
- NTI was a volatile component of operating income; P&L impact was low at -5 percent in 2016 but can be exacerbated under market duress.
- Econometric framework:
  - Merged supervisory series (NBB pre-2014, SSM post-2014).
  - Drivers: (i) changes to fair value of trading assets; (ii) shocks to benchmark money market rates; (iii) shocks to equity realized volatility using projected paths for EURSTOXX 50, BEL-20, and S&P 100.
  - Bayesian Model Averaging (BMA) technique used; trading income and fair value of assets treated as endogenous.
- Results: shocks to NTI are negative though contained; effects vary across banks (Figure 10 shows point estimates of NTI in percent of trading assets).

### Pre-Provision Net Revenue (PPNR) and funding costs
- PPNR modelling decomposes shocks to margins into drivers: reference rates, volume of credit, newly defaulted balances, and banks’ commercial strategies; ‘specific’ drivers differ from ‘systemic’ drivers.
- Funding costs methodology:
  - Split into a reference date and a spread by bank and instrument using supervisory data (merged NBB and SSM series).
  - Bank-specific deposit rate spread over benchmark forecasted using a BVAR approach with macrofinancial variables and bank-specific determinants (including capital ratios and asset quality).
  - Funding benchmark: 3m Euribor path generated by the scenario and treated as exogenous.
- Projected bank spreads over Euribor:
  - Bank spreads over Euribor reach around 75bps in 2017 and 100bps in 2018.
  - Relative to the starting point, spread shocks are around 18bps in 2017 and 12bps in 2018.

*Source: IMF staff analysis as presented in the provided content unit.*

### 80.      A broader strategy was applied to forecast bank spreads for debt instruments. Data on

### 80.      A broader strategy was applied to forecast bank spreads for debt instruments.

### Forecasting bank spreads for debt instruments
- Liability structure by re-pricing maturity bucket sourced from the STE maturity ladder template.
- Three approaches used to project bank spreads over reference rate for outstanding debt liabilities:
  - Behavior of yield-to-maturity spreads over benchmark for all single-name issued bonds active during the GFC and the sovereign debt crisis. Spreads were projected as the spike observed in crisis episodes on active bonds as of end-2016. The data was sourced from Bloomberg over 2005-May 2017.
  - Evolution of spreads at issuance of bank debt excluding short-term paper. The dataset covered all the issuing subsidiaries that belong to the same financial group at the Belgian highest level of consolidation. The data was sourced from Dealogic over 2005-May 2017.
  - A BVAR approach drawing on the effective rate over debt reported by banks in their supervisory returns. Stress test median outcome results are shown under this approach.

### Funding structure and funding costs
- Belgian banks’ liability structure: 85 percent deposit funded, and only 15 percent debt funded; variation across banks between 75 percent and 95 percent.
- Funding costs determined largely by benchmark rates and bank spreads over deposit rates.
- Strong retail deposit base limits rise in funding costs for Belgian banks under stressed conditions given historical behavior of deposits.

### Interest income and lending spreads
- Lending rates on loans and advances decomposed into the 5y OLO reference rate and a bank specific spread over benchmark.
- Implicit lending rates sourced from banks’ quarterly reports (publicly available).
- Determinants of lending spreads: core macrofinancial scenario conditions and projected path of funding cost.
- A BVAR econometric strategy used to estimate pass-through of idiosyncratic funding costs to customers at the bank level. Main stress testing results presented under non-pass through; pass-through shown as a sensitivity.
- Estimated pass-through impact: moderate, ranging between 9bps and 11bps over the first two years of the stress.
- Implicit interest rate on securities portfolio computed over the last two reporting periods by asset class (own sovereign bonds, foreign sovereign bonds, corporate bonds, other securities) for AFS and HTM regulatory books; implicit interest rate projected forward assuming no changes in portfolio composition during the stress test period (conservative assumption).

### Interest rate risk in the banking book (IRRBB)
- Interest rate gaps computed across six maturity buckets: <1m, 1m-2m, 2m-3m, 3m-6m, 6m-12m, and >12m.
- Net income from repricing gaps calculated using repricing gap, bucket mid-point in days, and shock to risk free rates.
- Interest sensitivity assessed by assuming 200 bps increases or decreases in interest rates and measuring impact on on- and off-balance sheet exposures; resulting changes in CET1 relatively small on average but can be severe for some banks.

### Other P&L items and dividend policy
- Bank tax rate driven by macroeconomic scenario; econometric analysis suggests effective bank tax rate over net profit not very sensitive on average; panel-based regression adjusted to banks’ reported tax rates at the reference date.
- Other P&L items projected using panel-based estimation with fixed effects; time series data sourced from Fitch applied to FINREP supervisory returns. Main variables included:
  - Fees and commissions relative to total assets;
  - Non-interest income relative to total assets;
  - Non-interest expenses relative to total assets including administrative and operating expenses;
  - Tax rate over net profits.
- Fees and commissions can represent up to 70bps of total assets for some banks; other non-interest income around 30 bps for some banks.
- Dividend payout ratio linked to net profits:
  - Under positive profits, dividend payout floor set at 30 percent subject to dividend restrictions if banks breach capital conservation buffer.
  - If profits exceed 50 percent of profits at the reference date, dividend payout ratio set at 50 percent.
  - If banks are making losses, no dividend payout is assumed.

### Solvency results — scenario-based outcomes
- Baseline scenario:
  - Capital ratios decline slightly because of full implementation of Basel III deductions and balance sheet expansion.
  - Aggregate CET1 ratios stabilize at around 15 percent in the later years.
  - Deductions required by Basel III leave CET1 ratios comfortably above the fully-loaded minimum of 8.5 percent for the four largest O-SIIs and at 7.75 percent for the remaining two O-SIIs.
  - Total regulatory capital projected to decline by an additional 30 basis points to 17.3 percent by 2021 due to elimination of capital instruments no longer eligible as Tier 2 capital (replacement of Tier 2 grandfathered instruments disallowed in the test). Minimal impact from AT1 grandfathered instruments.
- Adverse scenario:
  - All banks meet minimum capital requirements and none needs to draw down its capital conservation buffer at the low point of the stress.
  - Aggregate CET1 ratios fall by 370 basis points from 15.1 percent at end-2016 to a low point of 11.4 percent in 2018 before trending back to 13.2 percent by 2021.
  - RWAs rise by 290 bps over the first two years.
  - Contribution to lower CET1 from credit and valuation losses estimated at 140 basis points and 210 bps, respectively, in the near term.
  - At the end of the 5-year horizon, credit and valuation losses depress CET1 equally by 310 basis points, partially offset by robust pre-tax profits driven by resilient interest margins.
  - Limited impact of market risk on P&L due to smaller trading assets and small equity exposures; significant valuation losses on AFS portfolio given assumed correction in fixed income markets and long duration of securities portfolio.

### Sensitivity tests and stress variants
- Low-for-long scenario:
  - Policy rates and lending rates remain constant at end-2016 levels, amounting to a cumulative decline in interest rates of some 150-bps relative to the baseline.
  - Interest margins and profitability fall, pushing down CET1 capital ratios by 130 bps by 2021.
  - Long duration of assets mitigates impact.
- Concentration risk:
  - Default of each bank’s three largest exposures leaves CET1 ratios above regulatory minima in all cases.
  - Average CET1 would tilt toward regulatory minimum following default of five largest counterparties; wide dispersion across banks.
- Mortgage collateral / LGD shock:
  - Increase of 25 percentage points in LGDs on Belgian mortgages would lead to a decline of 110 bps of CET1.
  - Banks would still post a 14 percent CET1 ratio at the low point of the stress.
- Inclusion of derivatives in repricing gap:
  - Aggregate effect positive due to hedging strategies; some banks build capital buffers by between 50bps and 70bps relative to benchmark excluding derivatives.
- Empirically estimated pass-through:
  - Allowing for pass-through effects estimated via BVAR would raise CET1 ratios up to 60bps in some cases.
  - Caveat: estimation does not account for volume effects on consumer credit from rising lending rates; credit path used to estimate CET1 impact is exogenous to banks’ lending practices and determined entirely by the stress test scenario.

### Key quantitative results and parameters (selected)
- 85 percent deposit funded; 15 percent debt funded; range: between 75 percent and 95 percent.
- Data coverage for spreads: 2005-May 2017.
- Aggregate CET1 (end-2016): 15.1 percent.
- Aggregate CET1 low point (2018, adverse): 11.4 percent.
- Aggregate CET1 by 2021 (adverse): 13.2 percent.
- Aggregate CET1 decline at low point: 370 basis points.
- RWAs rise by 290 bps over first two years (adverse).
- Near-term contributions to CET1 decline: credit losses 140 basis points; valuation losses 210 basis points.
- End-horizon contribution of credit and valuation losses: 310 basis points each.
- Total regulatory capital decline: additional 30 basis points to 17.3 percent by 2021.
- Low-for-long cumulative decline: some 150-bps relative to baseline; CET1 erosion by 130 bps by 2021.
- Pass-through lending spread impact: 9bps to 11bps over first two years.
- Interest rate shock used in IRRBB sensitivity: 200 bps up/down.
- Mortgage LGD shock: +25 percentage points → CET1 decline 110 bps.
- Derivatives inclusion effect: capital buffers increase by between 50bps and 70bps.
- Empirical pass-through raises CET1 up to 60bps in some cases.

*Source: IMF staff estimates.*

### 100.      The liquidity stress test covered the entire banking sector. For the LCR-based test, the

### The liquidity stress test covered the entire banking sector. For the LCR-based test, the

### Scope and samples
- LCR-based test sample: the six banks in the solvency stress test sample, plus two other SIs (Degroof Petercam and Bank of New York Mellon), and all LSIs.
- Cash-flow based test sample: the six major Belgian banks and Bank of New York Mellon.
- LCR and cash-flow stress tests were conducted jointly with NBB using supervisory data as of June 2017.
- Cash-flow tool sample coverage for contagion through large exposures supervision: eleven banks (Argenta, Axa Bank Europe, Degroof, Belfius, BNPPF, INGB, Investar, KBC Bank, KBC Group, and Bank of New York Mellon).

### LCR-based stress-test scenarios (calibrated by ECB staff)
- Three scenarios with assumptions on haircuts, inflow rates, and cash-flow run-off rates more severe than Basel:
  - Systemic stress scenario:
    - Applies lower market values and higher haircuts to HQLA assets.
    - Assumes higher outflow rates from drawdown of liquidity lines under system-wide distress of counterparties.
    - Assumes reduction of some inflow rates from financial counterparties.
  - Idiosyncratic scenario:
    - Applies higher run-off rates to unsecured funding, including stable retail deposits.
    - Motivated by credit risk sensitive investors.
  - Combined scenario:
    - Applies the most stressed assumptions on HQLA, inflow rates and run-off rates from the previous two scenarios.

### Cash-flow based tests (ECB liquidity tool)
- Uses ECB’s internal liquidity tool covering six stress scenarios and simulating stress factors over eighteen maturity buckets ranging from one-day to twelve months.
- Calibration motivated as a diagnostic tool to identify potential pressure points rather than as a forecasting tool to project liquidity shortfalls.
- Caveats:
  - Analysis performed using mapping from EBA’s old liquidity templates to the new COREP liquidity templates; subject to potential misreporting of some items.
  - Severity of some scenarios is extreme (e.g., the ‘contractual scenario’ assumes run-off rates for non-maturing liabilities reach 100 percent).
  - Calibration of the ‘combined scenario’ exhibits amplification effects in run-off rates compounded across maturity buckets.
  - Results from the cash-flow-based analysis should be interpreted with caution due to potential inconsistencies in the data used in those tests.

### Liquidity results and key statistics
- Average LCR ratio stood at around 140 percent in March 2017 for the whole sector, the Big six, and the Big four.
- Stress-test impacts on average LCR ratio:
  - Systemic scenario: average LCR ratio would shed 40 basis points.
  - Idiosyncratic scenario: average LCR ratio would shed 46 basis points.
  - Combined scenario: average LCR ratio would edge slightly above the 80 percent mark.
- Drivers of resilience:
  - Strong deposit base, high quality of counterbalancing capacity (high quality assets), and early implementation of a prudential liquidity requirement ratio in 2012.
- Vulnerabilities:
  - Stability of retail deposits varies across banks and volatility may change, making banks relatively vulnerable to idiosyncratic liquidity stress; this is captured by the idiosyncratic scenario of the LCR-based and cash-flow-based stress tests.
- LCR addresses 30-day liquidity pressures from repo transactions, but its look-back approach is not well suited to capture forward-looking stressed flows from collateralized derivatives associated with large shifts to market risk factors.

### Asset encumbrance
- Average asset encumbrance ratio of Belgian banking system: 13.7 percent of total assets in 2016.
- Asset encumbrance reached over one fourth of total central bank eligible assets for the sector.
- Main sources of asset encumbrance:
  - Large derivative books against cash collateral (in some cases related to legacy portfolios).
  - Covered bonds against mortgage loans.
  - Reliance on repo transactions supported by general collateral.
- Monitoring implication: although asset encumbrance for the sector is manageable, it represents a potential source of liquidity risk for some large banks and warrants continued monitoring.

### Contagion and network analysis
- Network analysis sample: all Belgian SIs using large exposures supervisory data.
- Channels assessed for transmission:
  - Bilateral exposures: cascading defaults from counterparties with large exposures to a distressed firm.
  - Market contagion: equity losses from revised market expectations on solvency of banks with similar business models.
- Network characteristics:
  - Density of the Belgian interbank network is low.
  - Network displays negative assortativity (nodes of a similar degree are associated with each other).
  - Network follows a core-periphery model with the largest banks at the periphery rather than at the core, reflecting reduced interbank activity and deleveraging between 2007 and 2014.
- Contagion simulation assumptions and results:
  - Standard assumptions: 40 percent LGD, 35 percent of lost funding non-replaceable, and 30 percent fire sales discount.
  - Under these assumptions, there are no banks whose default would lead to consecutive defaults of other institutions.
  - Results are robust to definition of exposures (original or after exemptions and credit risk mitigation), LGD parameters, rollover assumptions on lost funding, and fire-sale haircuts.
- Limitations and caveats of contagion analysis:
  - Data drawn from large exposures template excludes small exposures.
  - Analysis limited to domestic interbank market (exposures to foreign counterparts excluded) and assumes a passive role from the central bank.
  - Active fire-sale assets calibrated exogenously.
  - Valuation effects from common exposures to stressed assets by banks holding similar assets are excluded.
  - Contagion effects from bear-market sentiment to banks with similar business models are excluded.
  - Does not explore precautionary liquidity hoarding or Knightian uncertainty about rollover behavior.

### Systemic risk assessment (market contagion)
- CoVaR methodology applied to capture risks from market contagion and propagation through equity markets.
- Financial system definition for CoVaR: large internationally active European banks (the “EBA” sample).
- Sample construction:
  - Initial list draws from the 2018 EU-wide stress test sample of 49 banks; two large financial institutions from the 2016 EU-wide stress test sample were added back.
  - Excludes non-listed banks because the methodology requires stock market data.
  - Resulting sample formed by a total of 35 large banks.
  - Weekly data used from June 2005 through May 2017; selection criterion based on availability of comparable data over a long period.

*Source: IMF staff estimates and National Bank of Belgium supervisory data as presented in the original text.*

### 119.      Results are robust to the definition of Belgian banks’ peer group. Three additional peer

### Results are robust to the definition of Belgian banks’ peer group

### Peer-group construction, data, and sample
- Three alternative peer groups to Belgian banks were constructed: “Peer_local,” “Belgium,” and “Peer_CESEE,” capturing European banks with local presence in material geographies for listed Belgian banks; European banks with local presence in Belgium; and European banks with the same core markets as listed Belgian banks, respectively.
- The results are consistent with those obtained using the broader “EBA” sample.
- The analysis uses weekly data from June 2005 through May 2017.
- The sample includes the 32 largest listed EU banks from the 2018 EU-wide stress testing sample (analysis based on weekly data from June 2005 through May 2017).

### VaR and CoVaR definitions and conditioning variables
- A bank is defined as distressed when it reaches its VaR returns.
- CoVaR is defined as the maximum loss to be expected in a portfolio representative of the whole financial system for a given confidence level and time horizon, conditional on the maximum loss expected in another portfolio at a specific confidence level and time horizon.
- Estimation of systemic risk measures is conditional on a set of macro-financial state variables sampled weekly from U.S., European, and Belgian markets:
  - U.S. variables: Volatility Index (VIX) of the CBOE; change in the credit spread between the 10-year Moody’s seasoned Baa corporate bond and the 10-year U.S. Treasury bond.
  - European variables: liquidity spread (difference between the 3-month Euro repo rate and the 3-month German T-bill yield); change in the German Treasury bill secondary market 3-month rate; change in the slope of the German Bund (yield spread between the German benchmark 10-year bond and the German 3-month T-bill); Eurostoxx 500 Composite Index return.
  - Belgian variable: spread of the 5y OLO rate over the swap curve.
- All state variables sampled weekly; data obtained from Bloomberg.

### Empirical findings on market-based systemic risk and spillovers
- Market perceptions of systemic risk through equity markets have eased following crisis episodes:
  - VaR of financial system returns, controlling for macrofinancial drivers, edged down to around -12 percent of quarterly equity returns in May 2017 from a peak of -65 percent during the GFC and -40 percent during the European sovereign crisis.
- The CoVaR measure of comovement in tail returns is affected by distress in individual banks:
  - The coefficient related to dynamics of individual bank returns is always significant and improves model forecasting of the financial system portfolio tail performance.
  - Market volatility and liquidity pressures exhibit the strongest predictive power among state variables.
  - The median elasticity of the system to individual tail returns reaches 0.96 when the distress bank is deleveraging.
  - The amplification effect from bank deleveraging increases with a bank’s systemic importance, measured by its delta CoVaR.
- Risk of cross-border spillovers between listed Belgian banking institutions and other peer banks appears moderate:
  - CoVaR analysis suggests that, on average, the risk that severe distress affecting the listed Belgian bank is transmitted to large European banks is moderate.
  - The contribution to systemic stress is time-varying; the ranking of delta CoVaR for the Belgian bank is robust when calculated for the entire period June 2005–May 2017 and separately for the global financial crisis (Aug 2007–Aug 2009) and the European sovereign debt crisis (May 2010–Dec 2012).
- Empirical diagnostics (median statistics reported):
  - Pseudo-R2 reported values include 0.557 (with t-statistic 8.722) and 0.995 (with t-statistic 14.634) in different specifications.
  - Other displayed coefficients and t-statistics include (examples from figure): Symmetric/Asymmetric coefficients such as -0.008 (−1.423), 0.001 (0.177), -0.002 (−7.931), -0.002 (−7.792), -0.055 (−1.674), -0.012 (−0.598), and others as shown in the analysis.

### Box 2 — Overview of the CoVaR methodology (key formulas and concepts)
- CoVaR definition: the CoVaR of system j given the VaR of bank i, denoted CoVaRλ|VaRλ(i), is the λ quantile of the conditional loss function Pr( Xj ≤ CoVaR | Xi ≤ VaR ).
- ΔCoVaR approximates a bank’s individual contribution to systemic risk:
  - ΔCoVaR = CoVaRλ,j|VaRλ,i − CoVaRλ,j|VaRmedian.
  - Operationalized in the note as Δ = CoVaR − VaR (see formal expression in text).
- Banking system returns given bank i’s distress constructed as a weighted average of remaining banks’ returns, excluding the distressed bank, with weights proportional to book value of total assets.
- Econometric specifications include symmetric and asymmetric formulations to capture reinforcing deleveraging effects:
  - Symmetric specification: system tail predicted as function of lagged system, bank returns, and conditioning variables.
  - Asymmetric specification: allows separate coefficients for negative and non-negative conditioning events to capture stronger effects when bank returns are negative.
- Final contribution metric computed via percentiles and the relation %50ˆt,i VaRVaRCoVaRˆ−=Δλλ (see equation in text).

### Insurance solvency tests — scope, scenarios, and shocks
- Solvency II was implemented in the European Union in 2016 and forms the basis for the insurance stress test; reference date for ST exercises set to end-2016.
- Participation and coverage:
  - Eight insurance companies participated in the stress test.
  - These eight undertakings account for 78 percent coverage in terms of total balance sheet assets in the Belgian insurance sector.
  - Six companies apply the Solvency II standard formula; two companies use a partial or full internal model.
- Scope:
  - Stress test run at the solo entity level and excluded business performed by foreign subsidiaries; foreign activities for Belgian groups are mostly limited.
- Scenario specification:
  - Bottom-up (BU) stress test included two scenarios: the NBB “low-for-long” scenario and an “IMF adverse” scenario aligned with the banking sector stress test. Top-down (TD) exercise included only the IMF adverse scenario.
  - NBB “low-for-long” scenario specifics:
    - Foresees a drop in the Euro interest rate term structure to the lowest level observed during the last two years plus an additional drop of 15 basis points for the first 20 years of the term structure.
    - After 20 years (the “last liquid point”), instead of extrapolating towards an ultimate forward rate of 4.2 percent, the extrapolation is made towards a rate of 2 percent.
  - IMF adverse scenario narrative centers on:
    - an increase in general risk aversion in global markets,
    - a large correction in the Belgian real estate market,
    - a reassessment of regional sovereign risk.
    - For the insurance stress test, shocks are assumed to occur instantaneously at the beginning of the first year (front-loaded).
  - Market risk shocks for insurers are defined granularly and include shocks to:
    - risk-free interest rate, equity and property prices, credit spreads of corporate and sovereign bonds, losses on mortgage loans, and a shock to the external value of the Euro.
    - Given increased credit spreads, the Solvency II volatility adjustment (VA) increases, partially offsetting credit spread shock effects.
- Additional sensitivity exercises (not additive to the macrofinancial scenario):
  - Longevity: modeled as a permanent 20 percent decrease in mortality rates applied to all policies.
  - Mortality shock: modeled as a permanent 15 percent increase in mortality rates applied to all policies.
  - Pandemic event: modeled as a temporary 35 percent increase in disability-morbidity rates and a temporary 10 percent increase in mortality rates for the next 12 months.
  - Two historical windstorm events (Lothar (1999) and Xynthia (2010)) were used as catastrophic event templates; insurers provided reinsurance recoverables from the participant’s five largest reinsurers (on a group basis).
- Capital standards and modelling:
  - Main output is the effect on available own funds eligible for coverage of the solvency capital requirement (SCR).
  - Participating undertakings were requested to recalculate the SCR after stress, as stresses may affect capital requirements.

*Source: IMF staff estimates and analysis based on Bloomberg data and participating supervisory exercises.*

### 134.      Insurance companies have a broad range of risk-mitigating mechanisms in place which

### Insurance companies have a broad range of risk-mitigating mechanisms in place which

### ST design and methodology
- Management actions:
  - Could be included in BU calculations only if they were non-discretionary rules already in place at the reference date.
- Projection requirements for undertakings:
  - Three-year projection of business development under the baseline and the IMF’s adverse scenario.
  - Key projected figures: gross written premiums, gross claims, lapse rates, investment returns, net earnings, gross technical provisions, own funds, and the SCR.
  - Projections had to be made in line with the macrofinancial scenario.
  - Market value of investments was assumed to stay constant after the occurrence of shocks at the reference date (instantaneous shock); recovery driven solely by underwriting business and recurring investment income from interest, dividends and rents.
- Top-Down (TD) stress test inputs:
  - Granular breakdown of investment assets: geographical breakdown of sovereign and corporate bond holdings, rating distribution, maturities and coupon rates of fixed-income investments.
  - Cash-flow projections for the upcoming 50 years in line with the Solvency II contract boundaries.
- TD modelling features:
  - Haircuts applied to market value of assets per scenario; re-valuation of fixed-income assets with the stressed term structure (per currency).
  - Technical provisions re-valued with the stressed term structure including the volatility adjustment.
  - Re-calculation of the SCR limited to the market risk component; other components (including capital charge for life underwriting risks) assumed unchanged.
  - For companies using full or partial internal models, TD calculations used a simplified approach broadly in line with the standard formula.

### Key stress scenarios, haircuts and spread changes (as provided in source table)
- Equity shocks:
  - Equity, Belgium and advanced economies: -16.2%
  - Equity, Emerging/developing economies: -25.0%
- Haircuts and spread changes for fixed-income and real estate:
  - Corporate bonds and securitisations: AAA 0.50%
  - AA 0.80%
  - A 1.20%
  - BBB 1.80%
  - BB or lower 3.00%
  - Unrated 2.00%
  - Real estate, Belgium: -20.0%
  - Real estate, elsewhere: -15.0%
- Mortgage loans haircut: -2.0%
- Sovereign bond spread changes and country buckets:
  - Belgium 1.47%
  - EA high-spread countries 2.00%
  - EA low spread countries 0.50%
  - Other advanced economies 1.00%
  - Emerging/developing economies 2.00%
- Exchange rate changes:
  - USD -5.7%
  - GBP -9.4%
  - CZK -7.1%

### Results — solvency and capital impacts
- BU (bottom-up) results:
  - Median solvency ratio:
    - Pre-stress: 184 percent
    - IMF adverse scenario: 124 percent
    - NBB low-for-long scenario: 145 percent
  - Aggregated solvency surplus of participating companies:
    - Before stress: €10.2 billion
    - IMF adverse scenario: €4.1 billion
    - NBB low-for-long scenario: €6.2 billion
  - Companies breaching 100 percent SCR coverage:
    - In each scenario, one company drops marginally below 100 percent; capital needed to restore full coverage amounts to less than €10 million in each case.
- TD (top-down) results:
  - Broad confirmation of BU results with some differences:
    - Interquartile distribution and mean solvency ratio slightly below BU numbers, mainly due to companies with larger derivative holdings (their hedging effect not modeled in TD).
    - Two companies would face SCR coverage below 100 percent in the IMF adverse scenario with an aggregated capital shortfall of less than €400 million.
- Asset and liability value changes in IMF adverse scenario (aggregated):
  - Market value of assets declines by 12 percent.
  - Value of liabilities declines by 10 percent.
  - Excess of assets over liabilities is 29 percent lower after stress.
- Asset and liability changes in NBB low-for-long scenario (aggregated):
  - Assets increase by 3 percent.
  - Liabilities increase by 5 percent.
  - Excess of assets over liabilities lowered by 14 percent.

### Drivers of solvency impact and role of the volatility adjustment (VA)
- Main driver:
  - Increase in Belgian sovereign bond yields is the most relevant factor in the IMF adverse scenario, combined with substantially higher credit spreads for the Belgian sovereign.
- Volatility adjustment:
  - Most companies benefit from the assumed increase of the VA in the IMF adverse scenario; VA immunizes them against excessive spikes in credit spreads not driven purely by deterioration in underlying credit risks.
- Composition notes:
  - SCR composition figures include LAC_DT and LAC_TP whenever a separate module exists or when companies provided estimates; scope of internal model SCR modules can differ from the Solvency II standard formula.

### Sector behavior, market spillovers and asset allocation changes after stress
- General capacity to absorb shocks:
  - Insurance sector appears capable of withstanding modeled losses without materially amplifying market risks or tightening market liquidity.
- Expected asset allocation adjustments (types of assets likely divested or increased):
  - Assets likely divested: equity and corporate bonds below investment grade.
  - Assets likely increased: sovereign bonds.
- Timing and market impact of adjustments:
  - Actions would unlikely be disruptive; likely spread over several weeks or months, especially as the need for action is limited for the majority of undertakings.

### Profitability and medium-term recovery projections
- Return on equity under IMF adverse scenario:
  - Mean return on equity reduced from around 15 percent to less than 7 percent in the first year of the projection horizon.
  - Insurers expect net income to improve to levels slightly below 10 percent in the following years, assuming no recovery of financial markets after the instantaneous stress event.
- Solvency coverage recovery:
  - Profits remain positive even after stress; solvency coverage is recovering steadily, but mostly only in 2019.

### Underwriting sensitivities and catastrophe exposures
- Life underwriting shocks:
  - Longevity shock (permanent 20 percent decrease in mortality rates): decline of own funds by less than 1 percent.
  - Severe pandemic: would affect own funds even less.
  - Permanent 15 percent increase in mortality rates: marginal positive impact on the sector.
- Non-life underwriting shocks:
  - Repetition of historic windstorms (Lothar 1999 or Xynthia 2010): median company would suffer a loss in own funds of less than 1 percent.
  - Most impact expected to be borne by primary insurers as losses are below retention levels agreed with reinsurers.
  - Exposure to catastrophic events appears well managed via diversified reinsurance contracts.
  - Climate change increases exposure to a higher frequency of loss events.

### Challenges in a prolonged low-yield environment
- Duration and sensitivity:
  - Life insurance liabilities have very long durations; under market-consistent valuation they are highly sensitive to changes in interest rates used to discount future expected claims.
- Asset-liability matching:
  - Exact match usually not possible due to lack of very long-dated investment assets.
  - Belgian insurers have closed duration gaps by investing more in long-running fixed-income assets, specifically domestic sovereign bonds.
- Product and guarantee adjustments:
  - Aggregate decline in guaranteed interest rates driven by product changes: reducing guarantees in new business and increasing unit-linked policy sales.
  - One large company withdrew from offering any new policies with guaranteed rates.
  - Average guaranteed interest rate as of end-2016: 2.6 percent (down from 3.1 percent five years earlier).
  - Significant buckets of liabilities with guaranteed rates well above 2 percent and maturities of 30 years and more remain on balance sheets.

### Reinvestment risk and projected coupon declines
- Concentration of maturing fixed-income assets:
  - For the aggregate of insurance companies included in the stress test, 50 percent of all investments with a fixed coupon will mature between 2017 and 2025.
  - These maturing assets pay a coupon of 3.6 percent on average.
- Reinvestment assumptions and projected average coupon:
  - Assuming expiring instruments can only be reinvested at a coupon of 1 percent:
    - Average coupon for the fixed-income portfolio will drop to 2.5 percent by 2025.
    - Average coupon will further drop to 1.9 percent in 2030.
- Aggregate position and profitability implications:
  - On aggregate, the Belgian insurance sector still records positive spreads of investment returns over guaranteed interest rates.
  - Sector expected to remain profitable after the IMF adverse scenario, but significant differences exist across companies.
  - Companies more active in non-life and unit-linked life business are less affected and could sustain prolonged low yields; companies with a high stock of guarantees likely to experience a drain on profitability.

### Effect on financial conglomerates and stress test perimeter
- Structure and scope:
  - Three banking-led financial conglomerates operating in Belgium; banking sector is the most important sector within Belgian financial conglomerates.
  - Three groups mentioned: Belfius, KBC Group, Investar (Argenta Spaarbank belongs to Investar).
  - Belfius, KBC Group, and Investar are designated as financial conglomerates with significant banking and insurance activities and must comply with the Financial Conglomerate Directive.
- Regulatory treatment:
  - Belfius, KBC Group, and Investar received a waiver from the NBB allowing use of the Danish Compromise method.
  - For regulatory reporting, solvency is reported on the basis of a 370 percent risk weighting applied to holdings of own fund instruments of the insurance company after deconsolidation of the insurance entity from group figures.
  - Intragroup exposures and risk concentration treated as supplementary risk categories for Pillar 2 capital requirements.
- Stress testing approach:
  - Stress tests conducted separately for banks and insurers but based on a common stress scenario for consistency.
  - Spillover effects from the insurance firm to the bank-led conglomerate were assessed to identify potential amplification effects.

*Source: IMF staff calculations and company submissions (excerpts from chapter).*

### 155.      Different transmission channels for risks exist in a financial conglomerate. The most

### cr1869 - 155.      Different transmission channels for risks exist in a financial conglomerate. The most

### Transmission channels of risk within a financial conglomerate
- Capital channel: the parent undertaking may be obliged to recapitalize a subsidiary after material stress.
- Upstream payments channel: regular payments by the subsidiary to the parent (e.g., dividends or distribution fees) which could be lower in a stress situation.
- Intra-group exposures channel: on- and off-balance exposures subject to counterparty default risk.

### Conglomerate stress-test findings (insurance and banking interactions)
- Insurance stress test outcomes for the three entities within a financial conglomerate:
  - None of the three drops below the 100 percent SCR coverage.
  - No additional capital would be needed from the parent or other group entities.
  - All three remain profitable after stress, supporting continued dividend payments at previous years’ levels.
- Insurers’ operational and funding links to banks:
  - Insurers hold sizable deposits with related banks.
  - Operational links can extend beyond distribution channels, e.g., KBC has highly integrated cross-sector risk management.
  - Stress tests suggest the insurance sector could play a stabilizing role on the funding structure of the financial conglomerate.
- Fees and commission vulnerability:
  - Profitability could be challenged through subdued fees and commission income.
  - Premiums, especially in life insurance, have been declining in recent years; the IMF adverse scenario is not seen as a disruptive event that would significantly reduce premium volumes and distribution fees further.

### FSAP concluding findings on banks and systemic resilience
- Solvency and capital:
  - Under the FSAP macroeconomic stress scenario, the aggregate CET1 ratio falls from 15.1 percent to a low point of 11.4 percent.
  - All banks would meet minimum capital requirements and none would need to draw down their capital conservation buffer.
  - Belgian banks have almost doubled their risk-weighted capital ratios from pre-crisis levels.
- Liquidity:
  - Positive liquidity stress test results reflect a shift to traditional funding patterns and holdings of high quality assets.
  - Reliance on deposits and high quality liquidity buffers reduce sector vulnerability to systemic stress, but retail deposit stability varies across banks.
  - Asset encumbrance is manageable overall but represents a potential source of liquidity risk for a few large banks and warrants continued monitoring.
- Contagion and interbank exposures:
  - Risk of contagion from a bank default through interbank exposures in Belgium is contained due to small interbank exposures relative to capital levels and low network connectivity.
  - Caveat: analysis excludes cross-border exposures among credit institutions domiciled in other jurisdictions, and knock-on effects from behavioral responses by non-bank financial institutions.

### Caveats and stress-test limitations
- Stress-test calibration:
  - Scenarios are calibrated using historical data identifying extreme “tail events” based on a historical distribution; crises can produce unanticipated shocks and unexpected interrelationships.
  - Nonlinear effects can be partially captured, but unknown patterns may emerge, especially in extreme shocks.

### Policy recommendations to the NBB (microprudential and macroprudential)
- Expand stress testing capabilities:
  - Develop macroprudential stress tests that incorporate feedback loops with the real economy to strengthen monitoring of financial stability risks and support macroprudential policy calibration.
- Strengthen vigilance on intersectoral flows and cross-border operations:
  - Continue monitoring intragroup transactions, collateral flows, and cross-border operations to detect potential reversal of eased systemic risk indicators.
- Expand insurance stress testing:
  - Expand regular macroprudential stress tests for insurance companies with multi-period perspectives and feedback loops.
  - Incorporate changes in policyholder behavior and companies’ assumed asset allocation shifts into cross-sectoral macroprudential analysis.
  - Use stress test results to strengthen microprudential supervision, challenging companies’ Own Risk and Solvency Assessment (ORSA) and projections for future business (premium growth and investment returns).
  - As an intermediate step, undertake horizontal analyses of companies’ business plans and growth forecasts.
- Monitor insurer mortgage lending risks:
  - Set up a comprehensive monitoring framework for risks arising from mortgage lending by insurers.
  - Regularly analyze loan portfolio risk parameters such as loan-to-value and debt-service-to-income ratios.
  - Explore potential regulatory arbitrage between banks and insurers within a financial conglomerate.
  - If Solvency II Standard Formula inadequately captures a company’s risk profile (e.g., large holdings in mortgage loans or high concentration of the sovereign bond portfolio), assess the need for a Pillar 2 capital add-on.
- Monitor LTG measures and the volatility adjustment (VA):
  - Continue close monitoring of the use and impact of LTG measures, particularly the VA.
  - Regularly assess insurers’ liability valuation and solvency sensitivity to changes in the VA.
  - Monitor how insurers adjust portfolios in response to any perceived malfunctioning of the VA.

### Risk Assessment Matrix — selected risk highlights and quantitative points
- Sudden increase in global risk aversion:
  - Overall Level of Concern: Low
  - Likelihood of severe realization in 1–3 years: Medium
  - Expected impact: sizeable pick-up in money market rates, steepening of the yield curve, reductions in market liquidity, mounting losses in fair value instruments, and higher funding costs.
- A large correction in the Belgian real estate market:
  - Overall Level of Concern: Low
  - Likelihood of severe realization in 1–3 years: High
  - Key points:
    - Overvaluation estimates have stabilized at around 8 percent.
    - Exposures to construction and real estate have grown dynamically over the last decade with 10 percent of the loans granted to Belgian regulated real estate investment companies (REITS).
    - Insurers hold sizable mortgage loan portfolios compared to EU peers, amounting to almost 10 percent of assets.
    - The macroprudential measure adding a 5pp risk weight add-on for real estate mortgages by IRB banks has increased Belgian banks’ capacity to absorb losses.
- Reassessment of regional sovereign risk:
  - Overall Level of Concern: Low
  - Likelihood of severe realization in 1–3 years: High
  - Key points:
    - Sovereign bond portfolio relative importance: banks’ fixed income securities portfolio about 75 percent sovereign in 2016; insurers hold nearly 60 percent of their assets in sovereign bonds.
- A prolonged period of low growth and low interest rates in the EA:
  - Overall Level of Concern: Medium
  - Likelihood of severe realization in 1–3 years: Medium
  - Expected impact: compression of bank net interest income, greater defaults and impairments, potential amplification of adverse net income via large currency fluctuations for internationally active banks, and persistent profitability pressures on insurers despite partial shifts away from guaranteed life business.

### Volatility estimation (equity market realized volatility methodology)
- Realized volatility computed as a 20-day moving average on EURSTOXX 50, BEL-20, and S&P 100.
- Notation and estimators preserved:
  - Classical estimator using closing price changes: sigma-hat^2 = (C1 - C0)^2.
  - Inclusion of opening price reduces variance of estimate via formula involving f (fraction of the day that trading is closed).
  - Garman and Klass (1980) best scale invariant estimator under uninterrupted trading:
    - sigma-hat^2 = 0.511*c^2 - 0.019*ud - 0.383*d^2 + ... (preserve original estimator expression as presented).
  - For the S&P index, trading hours 9.30h through 16.00h yield fraction of market trading of 0.27.
  - Under intermittent trading, the composite ratio estimator is given preserving the original formula and coefficients as presented.
- Empirical example referenced: sharp drop of 5.5 percent in JPMorgan Chase shares after a $2 billion trading loss announcement; opening gap of 8.8 percent lower following news.

### Banking sector solvency test — scope and data points
- Institutional perimeter:
  - Six major banks included: Argenta, AXA Bank Belgium, Belfius Banque, BNP Paribas Fortis, ING Belgium, and KBC Group.
  - Market share covered: About 90 percent of total banking sector assets.
  - Effective date: December 2016.
- Data sources and scope:
  - Supervisory data; NBB’s pre-2014 reports and ECB post-2014 ITS templates; public sources including bank Pillar 3 disclosures, Bloomberg, Dealogic, Haver Analytics, Moody’s KMV, Fitch, International Financial Statistics (IFS), IMF Global Assumptions (GAS), and IMF WEO.
  - Scope of consolidation: consolidated group basis; projection of consolidated ratios based on solo estimates.
  - Treatment of insurance activities: insurance-generated dividends and commissions are recognized in financial conglomerates’ income statement; bank P&L projections incorporate upstreaming of dividends and fees/commissions from insurance stress test results.
- Stress-testing methodologies:
  - NBB conducted a TD macroprudential stress test based on the WEO forecast (baseline) and IMF’s Global Macrofinancial Model with additional shocks to residential and commercial real estate prices (adverse), using satellite models for credit risk and P&L impact.
  - FSAP team conducted an independent TD macroprudential stress test with detailed calibration:
    - For IRB exposures, a separate credit risk model calibrated for five Basel asset classes, core industry sectors, and eleven material geographies (Belgium, Bulgaria, Czech Republic, France, Hungary, Ireland, Luxembourg, Netherlands, Slovakia, Switzerland and Turkey).
    - For STA exposures, projections of stressed NPL ratios, stressed coverage ratios, and a stressed transition matrix.
    - Market risk: stress to all sovereign issuers to which Belgian banks are exposed modeled for twenty-three issuers and a BAA corporate index; corporate benchmark index modeled separately.
    - Net trading income projected using a Bayesian VAR at the bank level.
    - Funding costs linked to euribor and bank-specific spreads estimated via Bayesian VAR; lending rates linked to OLO benchmark rates and commercial spreads.
- Additional stress-test elements:
  - TD stress test included assessment of a ‘low-for-long’ environment impact and a sensitivity test of the mortgage portfolio.
  - For credit risk, over 55 credit risk models and around 500 econometric specifications were estimated for PDs, using bank-specific regressions, panel regressions, and dynamic equations.

*Source: cr1869 - 155.*

### 3. Tail shocks     Scenario analysis

### 3. Tail shocks — Scenario analysis

### Scenario calibration and narrative
- Calibrated using the IMF’s Global Macrofinancial Model and auxiliary models to estimate stressed paths for residential and commercial real estate prices.
- Follows a layer-of-shocks approach combining simultaneous exogenous shocks and endogenous structural response amplifying initial macrofinancial impacts.
- Narrative features:
  - Re-emergence of financial stress and sovereign default risk in the euro area: money market spreads rise by 125 bps and 10y Belgium government bond yields increase by 150 bps.
  - Tightening of global financial conditions: long-term yields raise by an average of 100bps and equity prices fall by 20 percent in the EA, Japan, the U.K. and the U.S.
  - Autonomous domestic demand contraction in Belgium triggered by sovereign risk concerns and large property market corrections:
    - Private investment drop by 12 percent.
    - Private consumption drop by 3 percent.
    - Real residential property prices and commercial real estate prices fall by 20 percent peak-to-trough.
- Regional layer of shocks from de-globalization initiatives affecting Germany and Central and Eastern European Economies material for Belgian internationally active banks.
- Scenario magnitude: constitutes a 3.1 standard deviation move in two-year cumulative real GDP growth rate by 2019, calculated over 1990–2016.

### Macroeeconomic impacts and key statistics
- Belgium output falls 8.1 percent below baseline by 2019.
- Consumption price inflation falls 3.5 percentage points below baseline by 2020.
- Unemployment rate rises 2.2 percentage points above baseline by 2019.
- Within the euro area at the low point in 2019:
  - High spread economies experience output fall of 9.1 percent below baseline.
  - Low spread economies experience output fall of 7.2 percent below baseline.
- Government debt ratio in Belgium rises 17.9 percentage points above baseline by 2021 despite fiscal consolidation, driven by lower nominal output and higher debt service costs.

### Banking sector: solvency test — scenario design and shocks
- Includes an additional idiosyncratic and system-wide funding risk shock from money market dislocation linked to banks’ spreads over benchmark under stress.
- Sensitivity analysis includes:
  - Shocks to Belgium residential house prices impacting stressed LGDs.
  - Shocks to the euro swap curve.
  - Shocks to FX rates.
  - Low-for-long environment.
  - Shocks to concentration risk to single obligors.
  - Interest sensitivity gap of the ALM book under a range of market scenarios.
  - IRRBB excluding derivatives from the repricing gap schedule.
- Credit risk estimated according to national implementation of the CRD IV framework, covering retail, corporate, sovereign/public sector, and financial institution exposures; includes covered bonds and securitization exposures.
- Off-balance sheet exposures included using baseline and stressed Credit Conversion Factors (CCFs).
- Sovereign risk: mark-to-market valuation of securities for all sovereign issuers relevant for Belgian banks (list included in source).
- Market risk: stress from shocks to interest rates, exchange rates, credit spreads, commodities, and equity prices.

### Banking sector: profits, behavioral adjustments, and regulatory treatment
- Profits:
  - Stressed revenues from insurance activities for financial conglomerates projected under insurance stress test.
  - Net trading income from equity positions, debt instruments, and trading derivatives.
  - Interest income declines due to lost income from defaulted loans; interest income from non-defaulting loans estimated by satellite models.
  - Interest expenses increase due to rising funding costs with empirically estimated pass-through, plus add-on funding stress with no pass-through to lending rates.
  - Net fee and commission income and other income evolve with macro conditions and banks’ balance sheets.
  - No change in business models (no rebalancing of portfolio).
- Behavioral adjustments and dynamic balance sheets:
  - Credit supply effects are disallowed to calibrate credit risk projections; balance sheets evolve with key macro aggregates adjusting for credit demand.
  - EAD under stress from off-balance sheet exposures increases about 5-10 percent on average due to higher use of undrawn credit and liquidity facilities; all facilities conservatively treated as contractually irrevocable (“committed”).
  - Maturing assets replaced by exposures of same type and risk.
  - Dividend policy: under positive profits, dividend payout floor at 30 percent subject to restrictions if banks breach the capital conservation buffer; if profits exceed 50 percent of profits at the reference date, dividend payout ratio set at 50 percent; if banks make losses, no dividend payout assumed.
  - Effective tax rate evolves with macro scenario.
  - Losses are recognized in the same year a shock hits.
  - If banks’ capital ratio falls below regulatory minimum during the stress horizon, no prompt corrective action is assumed.
- Regulatory impact:
  - Effects of phase-out of no-longer-eligible additional Tier 1 and Tier 2 capital included.
  - No conversion of additional Tier 1 capital assumed during the stress horizon.
  - Capital definition follows Belgium implementation of CRD IV rulebook, including CET1, Tier 1, and total CAR; transitional path for capital components no longer eligible.
  - CET1/Tier 1/CAR ratio hurdle rates:
    - Big-4 Belgian banks: 6.0/7.5/9.5 percent of RWAs (including the O-SII fully implemented capital surcharge).
    - Other two O-SII Belgian banks: 5.25/6.75/8.75 percent of RWAs.
    - Additional 2.5 percent hurdle rate for the capital buffer breach.
  - Leverage ratio hurdle rate: 3 percent met with Tier 1 capital using Belgian implementation of CRD IV.

### Banking sector: reporting format for results
- Outputs presented:
  - Evolution of CET1, Tier 1, CAR, and leverage ratio for the aggregate banking system.
  - Contribution of key drivers to aggregate net profits and aggregate CET1 capital ratios.
  - Number of banks and share of total assets below hurdle rates.
  - Capital shortfall in terms of nominal GDP.

### Liquidity stress testing matrix — scope and scenarios
- Institutional perimeter:
  - LCR-based test: all Belgian banking institutions including seven SIs, two significant subsidiaries of EU SIs, ten core LSIs, ten other LSIs, and four subsidiaries of EU SIs with presence in Belgium; Dexia excluded as a run-off vehicle.
  - Cash-flow based test: all eight Belgian SIs.
  - Market share covered: Between 95 and 100 percent of banking sector total assets.
  - Data and base date:
    - Contractual based tests: supervisory data as of June 30, 2017.
    - Cash-flow liquidity stress tests: based on June 30, 2017 except two significant subsidiaries of EU STs based on end-2016 data.
- Channels and methodology:
  - Basel III measures of liquidity risk—the LCR conducted on four calibrated scenarios.
  - Two implied cash flow tests under six alternative scenarios.
  - Evolution and drivers of asset encumbrance ratio.
- Risks and buffers:
  - Risks: funding risk, rollover risk, market liquidity risk, liquidity risk related to systemic and idiosyncratic events, margin requirements for derivatives and repo operations.
  - Buffers: HQLA securities assessed at market values net of haircut on a security-by-security basis.
- LCR tail shock scenarios (four separate scenarios):
  - LCR Scenario under standard assumptions calibrated by BCBS.
  - Systemic stress scenario: lower market values and higher haircuts to HQLA; higher drawn on liquidity lines; reduced inflows from financial counterparts.
  - Idiosyncratic stress scenario: higher run-off rates to unsecured funding (10 percent run-off for stable deposits) and lower inflow rates.
  - Combined stress scenario (LCR “Belgium wholesale stress”): takes most stressed haircuts/run-off rates from previous scenarios; characterized by:
    - Freeze of wholesale funding on interbank, secured funding via repo and covered bonds, and commercial paper markets (operational deposit run-off rate of 75 percent and not-fully covered corporate deposits run-off rate of 100 percent).
    - Liquidity risk from sizeable margin calls related to secured funding, derivatives and foreign currency funding due to market liquidity shocks, derivative assignments, and unwinds and disruptions in the FX swap market (rollover of secured funding backed by other than Level 1 and Level 2A assets of up to 0 percent).
- Implied cash-flow based tests (six scenarios over eighteen maturity buckets from one-day through twelve months cumulative):
  - S0: baseline.
  - S1: Contractual (excludes modeling adjustments to non-maturity assets and liabilities).
  - S2: Macroeconomic (includes lower inflows from debt issuance and monies due; incorporates higher haircuts on banks’ counterbalancing capacity).
  - S3: Idiosyncratic (higher run-off rates to liabilities, including 10 percent run-off rates for stable deposits and 30 percent for other deposits).
  - S4: Idiosyncratic (no inflows): excludes most contractual inflows except for central banks and derivative receivables.
  - S5: Combined scenario including S2 and S3 assumptions.
- Regulatory standards:
  - Counterbalancing capacity above net cash outflows under stress scenario.
  - CRD IV full implementation for the LCR ratio at 100 percent.
- Reporting format for liquidity results:
  - Changes in average liquidity position and counterbalancing capacity for each scenario.
  - Distribution of banks’ liquidity position for each scenario.
  - Number of banks with counterbalancing capacity below net cash outflows.
  - Banks’ post-shock net liquidity position.
  - Liquidity shortfall in terms of banking system total liabilities.

### Insurance sector: solvency risk — framework and tail shocks
- Institutional perimeter:
  - 8 composite insurers included; market share: 78 percent of total balance sheet assets.
  - Reference date: December 31, 2016.
- Methodology and time horizon:
  - Investment assets: market value changes after price shocks affecting solvency position.
  - Sensitivity analysis of effect on available capital and solvency position.
  - IMF top-down uses instantaneous shock; also a 3-year projection (only in the baseline and the adverse scenario) in bottom-up.
- Adverse scenario (tail shocks) parameters:
  - Interest rates +50 bps (EUR, parallel shift up to a maturity of 20 years).
  - Sovereign bond spread: +147 bps (Belgium), +50 bps (other low-yield euro area countries), +200 bps (high-yield euro area countries).
  - Stock prices: -16.2 percent (advanced economies), -25 percent (emerging and developing economies).
  - Property prices: -20 percent (Belgium), -15 percent (other countries).
  - Corporate bond spreads between +50 bps (AAA) and +300 bps (BB and lower).
  - 5.7 percent depreciation of EUR against USD.
  - IMF top-down variant also included a 2 percent haircut on mortgage loan portfolio.
- NBB “low-for-long” scenario:
  - Interest rate term structure (up to 20 years) at minimum of the last 2 years minus 15 bps; beyond 20 years extrapolation towards an ultimate forward rate of 2.0 percent (instead of 4.2 percent as prescribed by Solvency II).
- Sensitivity analyses:
  - Longevity shock: permanent 20 percent decline in mortality rates.
  - Mortality shock: permanent 15 increase in mortality rates.
  - Pandemic event: temporary 35 percent increase in disability/morbidity rates, temporary 10 percent increase in mortality rates.
  - Variations of the interest rate term structure.
  - Default of largest financial counterparty.
  - Catastrophic events considered: Windstorm Lothar (December 1999) and Windstorm Xynthia (February/March 2010).
- Risks and buffers:
  - Risks/factors assessed: market risks (interest rates, share prices, property prices, FX rates, credit spreads), credit risks (default in mortgage loans; default of largest financial counterparty), underwriting risks (longevity, mortality, pandemic, catastrophic events).
  - Summation of risks with no diversification effects in aggregation.
  - Buffers: product-specific in bottom-up; none in top-down.
  - Behavioral adjustments: management actions limited to non-discretionary rules in place at the reference date (bottom-up); none in top-down.
- Regulatory standards and reporting:
  - Regulatory/accounting standards: Solvency II and National GAAP.
  - Output presentation includes:
    - Impact on solvency ratios (including and excluding the effect of long-term guarantee measures).
    - Impact on net income.
    - Contribution of individual shocks.
    - Dispersion measures of solvency ratios and net income.

*Source: cr1869 - 3. Tail shocks     Scenario analysis (IMF country report chapter).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2018/cr1869.pdf_