## _cr16306

## Source details

**Canonical URL:** [_cr16306](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16306.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16306.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16306.pdf.json)

---

### PERFORMANCE OF THE BANKING SECTOR
- Structure and size
  - Bank assets amounted to 103 percent of the gross domestic product (GDP) at end-2015.
  - Pension funds, insurance, and mutual funds have assets of 3.6, 2.0, and 1.7 percent of GDP, respectively.
  - The largest 20 banks account for three quarters of system assets; the top 10 banks extend about 70 percent of total lending.
  - State-owned commercial banks accounted for 60 percent of system assets at end-2015.
  - Banks’ business model: credit intermediation; loans constitute 70 percent of banking assets.

- Capital and liquidity
  - Capital adequacy ratio (CAR) was about 13 percent in 2015 and fell to 12 percent after forbearance was reduced in early 2016.
  - Loan-to-deposit ratio decreased to 115 percent by end-2015 from a recent peak of 125 percent.
  - System averages mask considerable dispersion; many mainly small banks were closed.

- Asset quality and profitability
  - Household overdue loans reached 8.4 percent of total loans by February 2016; corporate overdue loans were 6.5 percent.
  - Return on assets reached 0.3 percent at end-2015.
  - Drivers: contracting net interest margins, falling net fees and commissions, non-interest expenses declining less than NII, and sharply rising provisions.

- Concentration, connected lending, and vulnerabilities
  - Large exposures stood at 261 percent of capital in February 2016.
  - Related party lending may be understated due to narrow related-party definitions and weak beneficial-owner implementation; stricter rules published in 2014 with implementation postponed twice until January 2017.
  - Banks in the 21–50 segment showed negative return on equity of about 25 percent in 2015Q4; two banks in this segment were undergoing resolution.
  - Net foreign asset position reached US$100 billion at end-2015.
  - Exposure of foreign banks to Russia about US$110 billion by end-2015; 40 percent concentrated in France and Italy (US$26 and US$19 billion).
  - Russian banks’ cross-border exposures about US$245 billion at end-2015 (below 20 percent of GDP); two thirds are loans and deposits.

- Risks identified
  - Asset quality is the largest risk; unsecured consumer lending in 2011–14 has increased retail NPLs.
  - Corporate-sector overdue loans rising in construction and real estate, mining, trade, and agriculture.
  - Supervisory inspections revealed violations: lending to shell companies, overvaluation of collateral, misreporting, unreliable financial statements.
  - Potential understatement of nonperforming assets due to: (i) lower quality of restructured loans; (ii) under-provisioned and under-collateralized portfolios; (iii) transfer of distressed assets to affiliated off-balance sheet entities not under consolidated supervision.

- Liquidity and market risk
  - System-wide liquidity risk contained, but small and medium banks are individually vulnerable; reliance on higher-yielding collateral, short-term secured markets, and CBR.
  - Securities portfolios 15 percent of total assets at end-2015, mainly debt securities; sovereign securities constitute one quarter of debt security portfolios.
  - Indirect sovereign risk from state-owned financial and non-financial corporate sectors likely greater than direct sovereign holdings.

### STRESS TEST DESIGN
- Coverage and modalities
  - Exercise comprised multiple tests: CBR top-down (TD), bank bottom-up (BU), staff top-down with asset quality adjustments, CBR single-factor tests, and CBR liquidity tests.
  - TD tests horizon: five years; BU tests horizon: one year.
  - Resilience assessed relative to regulatory capital adequacy minimums: 4 ½ percent for Common Equity Tier 1 and 8 percent for total CAR.
  - TD by CBR covered 681 banks on Russian accounting standards, stand-alone.
  - BU covered 12 banks (including 10 SIBs) accounting for two thirds of system assets.
  - Staff TD used publicly available data for 37 banks accounting for about 82 percent of system assets, consolidated groups and IFRS.

- Risks covered
  - Credit, market, interest rate, and liquidity risks modeled.
  - Staff analysis focused on credit and interest rate risks affecting net interest income, including FX-related deterioration.
  - CBR TD and BU also covered market and liquidity risks.

- Liquidity stress tests (CBR)
  - One-month horizon with three scenarios: mild, severe, very severe.
  - Scenarios: increasingly large outflows from liability categories, higher discounts on non-liquid assets; assume no CBR or interbank financing.
  - Assess whether liquidity outflows exceed available liquid assets.

### TOP-DOWN SOLVENCY STRESS TESTS BY STAFF
- Methodology
  - Panel regression techniques adjusted for higher volatility in exchange rate, inflation, and interest rates.
  - Model focused on loan impairment charges to net income to incorporate changes in probabilities of default and loss given default and to mitigate differences in loan categorization and loss accounting across datasets.
  - Loans classified into five categories; categories IV and V defined as NPLs; provisions and charges also necessary for categories I–III.

- Asset quality adjustments by staff
  - Restructured loans ≈ 30 percent of large loans; staff estimated one third of restructured loans to be of weaker quality at end-2015.
  - Migration of weaker restructured loans increased loan impairment charges by 1 percentage point of total loans.
  - Staff accounted for under-provisioning in each loan category, increasing loan impairment charges by an additional 2.2 percentage points.
  - Overall effect of restructuring and under-provisioning assessed to be about 3.3 percentage points.

- Scenarios and horizons
  - Staff and CBR TD tests calibrated over five years to capture full shock-to-recovery cycles.
  - BU tests calibrated over one year.
  - Three macroeconomic scenarios used: baseline, V-shaped, and L-shaped.

### MACROECONOMIC SCENARIOS AND KEY PROJECTIONS
- Scenarios (selected projections 2015–20)
  - Baseline Scenario (Real GDP %): 2015 = -3.7; 2016 = -1.0; 2017 = 1.0; 2018 = 1.5; 2019 = 1.5; 2020 = 1.5
  - V-Shape Scenario (Real GDP %): 2015 = -3.7; 2016 = -7.8; 2017 = 2.6; 2018 = 2.9; 2019 = 1.1; 2020 = 1.0
  - L-Shape Scenario (Real GDP %): 2015 = -3.7; 2016 = -5.7; 2017 = -0.3; 2018 = 1.7; 2019 = 1.1; 2020 = 1.0
- Oil price paths
  - V-shaped: Oil price falls to US$19 per barrel; Brent recovers to US$40 in 2020.
  - L-shaped: Oil price falls to US$25 per barrel; Brent recovers to US$37 in 2020.
- Transmission channels and dynamics
  - Tail oil-price shocks calibrated to tail events in the oil price probability distribution.
  - Strong adverse effects on GDP and unemployment, large exchange rate depreciation, and strong responses of inflation and short-term interest rates to exchange rate depreciation.

### STRESS TEST RESULTS — HEADLINE FIGURES
- Two result sets: (1) reported capital ratios at end-Q3 2015, risk weights 90 percent, no asset-quality adjustment; (2) conservative set with asset-quality adjustment and risk weights 100 percent.
- First set (no asset-quality adjustment)
  - Estimated capital deficit of about ½ and 2 ½ percent of GDP in the baseline and stress scenarios respectively.
  - CBR and staff estimated aggregate capital deficit (peak during five-year horizon) of 0.3 and 0.4 percent of GDP, respectively, in the baseline scenario.
  - CBR estimated deficits of 2.5 and 2.1 percent of GDP in the V-shape and L-shape scenarios, respectively.
  - Staff estimated deficits of 2.4 and 2.7 percent of GDP in the V-shape and L-shape scenarios.
- Second set (with asset-quality adjustment)
  - Higher capital deficits of about 1 percent and 4 ½ percent of GDP in the baseline and in the stress scenarios, respectively.
  - Including restructuring and under-provisioning increases the capital deficit by 0.6 percentage points in the baseline and by about 2 percentage points in the stress scenarios.
  - Staff estimated capital deficits of 4.6 and 4.4 percent of GDP in the V-shape and L-shape scenarios respectively.
  - Peak deficits occur in 2017 in the V-shape scenario, and 2019 in the L-shape scenario.

### DETAILED SYSTEM AND BANK-LEVEL OUTCOMES
- V-shape scenario (with asset quality adjustment)
  - Banks with capital ratio breaches account for 87 percent of banking sector assets in 2017.
  - Median CET1 capital and total capital ratios fall to 0.8 percentage point and 3.5 percentage points, respectively, in 2017.
- L-shape scenario (with asset quality adjustment)
  - Cumulative capital deficit accounts for 4.4 percent of GDP.
  - Median total capital ratio bottoms out at 7.8 percent of RWA in 2018.
- Baseline with asset quality adjustment
  - Minimum regulatory capital requirements breached by 15 out of 37 banks in 2017, accounting for about 37 percent of banking sector assets.
  - Median CET1 capital and total capital ratios are 5.5 and 9.9 percent, respectively, in 2017.
- Medium-term recovery
  - Some weak banks accounting for about 40 percent of banking sector assets have difficulties returning to profitability.
  - CBR indicates weak large banks recover in the medium term; some medium- and small-sized banks (about 5 percent of banking system assets and not in staff’s sample) accumulate large capital deficits by 2020.

### COMPARISONS: CBR VS STAFF AND BU VS TD
- CBR and staff top-down tests show broadly similar capital deficits; main modeling differences:
  - Staff projects assets and equity (liabilities residual); CBR projects liabilities and equity (assets residual).
  - Credit losses peak in first year in staff’s tests; CBR shows peak in second year.
- Bottom-up versus top-down (12 banks comparison, bn RUB)
  - Baseline: BU aggregate capital deficit = RUB 3 billion; TD aggregate capital deficit = RUB 65 billion.
  - V-shaped: BU aggregate capital deficit = RUB 199 billion; TD aggregate capital deficit = RUB 522 billion.
  - L-shaped: BU aggregate capital deficit = RUB 70 billion; TD aggregate capital deficit = RUB 426 billion.
  - Market risk appears to be underestimated by banks, alongside credit risk.

### SINGLE-FACTOR AND LIQUIDITY TESTS (CBR)
- Single-factor credit test
  - Default of each bank’s top five non-bank borrowers: 338 banks with 82.5 percent of banking system assets experience an aggregate capital deficit of RUB 1,519 billion.
- Market risk single-factor results (selected)
  - Exchange rate depreciation of 20 percent: five banks comprising 0.1 percent of system assets would have a capital deficit of RUB 0.2 billion.
  - 30 percent decline in value of equity holdings: seven banks comprising 0.5 percent of system assets would have a capital deficit of RUB 0.9 billion.
  - Interest rate shocks (corporate +1,000 bps and government +400 bps): 13 banks comprising 6.7 percent of system assets would have an aggregate capital deficit of RUB 20 billion.
- Liquidity stress tests (assume no CBR or interbank financing)
  - Mild scenario: 36 banks (1.7 percent of system assets) show liquidity deficits; total shortage less than 0.1 percent of GDP.
  - Severe scenario: 111 banks (11 percent of system assets) with shortage of 0.4 percent of GDP.
  - Very severe scenario: 181 banks (41.8 percent of system assets) with shortage of 2.9 percent of GDP.
  - Large state-owned banks hold special liquidity positions but sanctions restrict access to term external funding; lower-rated small and medium banks rely on higher-yielding collateral, short-term secured markets, and CBR.

### REGRESSION, PROJECTIONS, AND MODELING NOTES
- Estimation sample period for balance-sheet and income relationships: 2005Q1 to 2015Q3.
- Net charge offs (NCOs) modeled as share of gross loans; NCOs negatively related to real output, house prices, and loan growth; positively associated with higher interest rates and term premium.
- Net interest income depends on interest rates and growth in loans and deposits (assumed to grow at same rate).
- Risk weights assumed constant at two levels: 90 percent and 100 percent.
- Dividend payouts and taxes fixed at 20 percent over the stress testing horizon.

### ASSET QUALITY REVIEW RECOMMENDATION
- An asset quality review is recommended to assess risks stemming from uncertainty about asset quality, concentrated lending, and related party lending.
- Stress tests should be supplemented by a granular and comprehensive review of banks’ asset portfolios to gauge capitalization needs.
- Implementation note: "The Bank of Russia has already begun to implement the recommendations of FSAP mission: as part of banking supervision improvement the Risk Analysis Service has been organized. The Service is expected to assess assets and transactions of credit institutions on ongoing basis."

### STRESS TESTING RECOMMENDATIONS (PRIORITIES)
- Implement modern, comprehensive, and integrated TD and BU methodologies covering banks and affiliated entities.
- Develop an “expected loss approach” for credit risk using probability of default, loss given default, and exposure at default.
- Develop a credit registry to provide additional data.
- Expand use of TD bank-by-bank stress test results in supervision, including capital planning and dividend policy.
- Consider possible publication of aggregate TD stress test results in the Financial Stability Review.
- Enhance system-wide stress tests for all banks, beginning with large and medium-sized banks, using macro scenarios and key bank stress test parameter assumptions over a three-year horizon.
- Conduct stress tests on a consolidated basis to capture downstreaming of problematic assets by D-SIBs to unconsolidated subsidiaries.
- Conduct liquidity stress tests by currency to identify shortages in particular currencies that aggregate tests may overlook.

*Source: EXECUTIVE SUMMARY and selected chapters (IMF staff note prepared as part of the 2016 Financial Sector Assessment Program for the Russian Federation).*

### EXECUTIVE SUMMARY __________________________________________________________________________ 5

### EXECUTIVE SUMMARY

### PERFORMANCE OF THE BANKING SECTOR
- Structure and size
  - Bank assets amounted to 103 percent of the gross domestic product (GDP) at end-2015.
  - Pension funds, insurance, and mutual funds have assets of 3.6, 2.0, and 1.7 percent of GDP, respectively.
  - The largest 20 banks account for three quarters of system assets; the top 10 banks extend about 70 percent of total lending.
  - State-owned commercial banks accounted for 60 percent of system assets at end-2015.
  - Banks’ business model: credit intermediation; loans constitute 70 percent of banking assets.

- Capital and liquidity
  - Capital adequacy ratio (CAR) was about 13 percent in 2015 and fell to 12 percent after forbearance was reduced in early 2016.
  - Loan-to-deposit ratio decreased to 115 percent by end-2015 from a recent peak of 125 percent.
  - System averages mask considerable dispersion; many mainly small banks were closed.

- Asset quality and profitability
  - Household overdue loans reached 8.4 percent of total loans by February 2016; corporate overdue loans were 6.5 percent.
  - Return on assets reached 0.3 percent at end-2015.
  - Drivers: contracting net interest margins, falling net fees and commissions, non-interest expenses declining less than NII, and sharply rising provisions.

- Concentration, connected lending, and vulnerabilities
  - Large exposures stood at 261 percent of capital in February 2016.
  - Related party lending may be understated due to narrow related-party definitions and weak beneficial-owner implementation; stricter rules published in 2014 with implementation postponed twice until January 2017.
  - Performance across the system is uneven: banks in the 21–50 segment showed negative return on equity of about 25 percent in 2015Q4; two banks in this segment were undergoing resolution.
  - Net foreign asset position reached US$100 billion at end-2015.
  - Exposure of foreign banks to Russia about US$110 billion by end-2015 (halved since 2014); 40 percent concentrated in France and Italy (US$26 and US$19 billion).
  - Russian banks’ cross-border exposures about US$245 billion at end-2015 (below 20 percent of GDP), two thirds are loans and deposits; subsidiaries mainly in Austria, Turkey, and Cyprus.

- Risk assessment
  - Asset quality is the largest risk; unsecured consumer lending in 2011–14 has increased retail NPLs.
  - Corporate-sector overdue loans rising in construction and real estate, mining, trade, and agriculture.
  - Supervisory inspections revealed violations: lending to shell companies, overvaluation of collateral, misreporting, unreliable financial statements.
  - Potential understatement of nonperforming assets due to: (i) lower quality of restructured loans; (ii) under-provisioned and under-collateralized portfolios; (iii) transfer of distressed assets to affiliated off-balance sheet entities not under consolidated supervision.

- Liquidity and market risk
  - System-wide liquidity risk contained, but individual vulnerabilities remain; small and medium banks reliant on higher-yielding collateral, short-term secured markets, and CBR.
  - Securities portfolios 15 percent of total assets at end-2015, mainly debt securities; sovereign securities constitute one quarter of debt security portfolios.
  - Indirect sovereign risk from state-owned financial and non-financial corporate sectors is likely greater than direct sovereign holdings.

### STRESS TEST DESIGN
- Coverage and modalities
  - Exercise comprised multiple tests: CBR top-down (TD), bank bottom-up (BU), staff top-down with asset quality adjustments, CBR single-factor tests, and CBR liquidity tests.
  - Bank-focused due to bank-dominated financial sector.
  - TD tests horizon: five years; BU tests horizon: one year.
  - Resilience assessed relative to regulatory capital adequacy minimums: 4 ½ percent for Common Equity Tier 1 and 8 percent for total CAR.
  - TD by CBR covered 681 banks on Russian accounting standards, stand-alone.
  - BU covered 12 banks (including 10 SIBs) accounting for two thirds of system assets.
  - Staff TD used publicly available data for 37 banks accounting for about 82 percent of system assets, consolidated groups and IFRS.

- Risks covered
  - Credit, market, interest rate, and liquidity risks modeled.
  - Staff analysis focused on credit and interest rate risks affecting net interest income, including FX-related deterioration.
  - CBR TD and BU also covered market and liquidity risks.

- Liquidity stress tests (CBR)
  - One-month horizon with three scenarios: mild, severe, very severe.
  - Scenarios: increasingly large outflows from liability categories, higher discounts on non-liquid assets; assume no CBR or interbank financing.
  - Assess whether liquidity outflows exceed available liquid assets.

### TOP-DOWN SOLVENCY STRESS TESTS BY STAFF
- Methodology
  - Panel regression techniques similar to advanced economy FSAPs, adjusted for higher volatility in exchange rate, inflation, and interest rates.
  - Model focused on loan impairment charges to net income to incorporate changes in probabilities of default and loss given default and to mitigate differences in loan categorization and loss accounting across datasets.
  - Loans classified into five categories; categories IV and V defined as NPLs; provisions and charges also necessary for categories I–III.

- Asset quality adjustments by staff
  - Restructured loans ≈ 30 percent of large loans; staff estimated one third of restructured loans to be of weaker quality at end-2015.
  - Migration of weaker restructured loans increased loan impairment charges by 1 percentage point of total loans.
  - Staff accounted for under-provisioning in each loan category, increasing loan impairment charges by an additional 2.2 percentage points.
  - Overall effect of restructuring and under-provisioning assessed to be about 3.3 percentage points.

- Scenarios and horizons
  - Staff and CBR TD tests calibrated over five years to capture full shock-to-recovery cycles.
  - BU tests calibrated over one year.
  - Three macroeconomic scenarios used for bank-by-bank assessment: baseline, V-shaped, and L-shaped.

### KEY FINDINGS, RESULTS, AND POLICY-RELEVANT CONCLUSIONS
- Capital needs and fiscal capacity
  - Stress tests indicate the banking system is likely to need additional capital; even in the baseline certain banks require new capital owing to low profitability and increasing credit losses.
  - Required resources are higher in stress scenarios but remain manageable.
  - If public funds are needed, there is sufficient fiscal space provided fiscal policy remains prudent.

- Need for granular asset review
  - Stress tests should be supplemented by a granular and comprehensive review (Asset Quality Review) of banks’ asset portfolios to gauge capitalization needs.

- Uncertainty and supervisory actions
  - Considerable uncertainty about the strength of loan portfolios due to supervisory-identified violations and possible underreporting.
  - Stricter related-party and beneficial ownership rules (published 2014) require implementation to better link exposures.

- Liquidity vulnerabilities
  - Small and medium-sized banks are particularly vulnerable to liquidity dry-ups; access to CBR and secured markets is a key vulnerability.

- Recommendations implicit in the analysis (derived from staff emphasis on actions)
  - Supplement top-down stress testing with granular asset portfolio reviews to more accurately quantify capitalization needs.
  - Strengthen implementation of related-party and beneficial-owner identification rules (implementation postponed until January 2017).
  - Monitor and address liquidity vulnerabilities of lower-rated small and medium-sized banks, given their reliance on higher-cost funding and secured markets.

*Source: EXECUTIVE SUMMARY (IMF staff note prepared as part of the 2016 Financial Sector Assessment Program for the Russian Federation).*

### 18.      The macroeconomic stress scenarios quantified the impact of negative oil price shocks

### 18.      The macroeconomic stress scenarios quantified the impact of negative oil price shocks

### Macroeconomic stress scenarios
- Three macroeconomic scenarios agreed by the authorities and the FSAP team.
- V-shaped scenario:
  - Oil price falls to US$19 per barrel.
  - Brent recovers to US$40 in 2020 in the V-shaped stress scenario.
- L-shaped scenario:
  - Oil price falls to US$25 per barrel but recovers more slowly.
  - Brent recovers to US$37 in 2020 in the L-shaped stress scenario.
- Medium-term assumptions:
  - Annual real GDP growth converges to one percent in both scenarios, half a percentage point below 2020 baseline growth.
- Key transmission channels:
  - Tail oil-price shocks calibrated to tail events in the oil price probability distribution.
  - Strong adverse effects on GDP and unemployment.
  - Large exchange rate depreciation via the relationship between oil prices and the RUB/US$ rate.
  - Inflation and short-term interest rates respond strongly to exchange rate depreciation.

### Bank profitability and short-term dynamics under stress
- Profitability effects:
  - Banks’ profitability suffers due to deterioration in asset quality and lower interest margins.
  - Ruble-denominated consumer and corporate loans are hit hardest.
- FX portfolios:
  - Exchange rate depreciation leads to weaker performance of portfolios in foreign currency with very low provisioning.
- Interest margins and funding:
  - Exchange rate depreciation triggers higher domestic inflation and higher short-term interest rates.
  - Higher funding costs and inability to fully adjust asset-side interest rates owing to maturity mismatch cause net interest margin to shrink in the short term.
  - By the end of the five-year horizon, net interest margins recover, while loan loss provisions moderate.

### Balance sheet and income statement growth projections
- Projection approach:
  - Bank finances projected conditional on macro scenarios over a five-year stress testing horizon.
  - Projections based on statistical relationships with macro indicators, bank-specific characteristics, and other assumptions.
  - These projections used to calculate bank-specific capital dynamics and capital ratios.
- Statistical estimation:
  - Relationships estimated for each balance sheet and income statement item necessary to project bank capital ratios.
  - Estimation sample period: 2005Q1 to 2015Q3.
  - Various specifications considered; best models chosen by economic interpretation, significance, higher R2, and parsimony.
- Balance sheet growth:
  - Growth rate of gross loans estimated using a panel data regression with bank fixed effects; lagged dependent variable included.
  - Assets growth assumed affected by gross loans growth.
- Net charge offs (NCOs):
  - Estimated as a share of gross loans.
  - NCOs negatively related to real output, house prices, and loan growth.
  - NCOs positively associated with higher interest rates and the term premium.
- Income statement decomposition:
  - Net income disaggregated into six components: net interest income, trading income, non-interest expenses, net fees and commissions, other net income and provisions.
  - Net interest income depends on interest rates and growth in loans and deposits. (Loans and deposits assumed to grow at the same rate.)
  - Loan provision growth negatively associated with real output growth and the term premium; higher unemployment amplifies provision growth.
  - Risks from FX lending and depreciation captured by the term premium; depreciation lowers the term premium and increases loan impairment charges.
  - Trading income increases during crises but fitted relationship is weak; trading income assumed constant over the stress horizon given poor fit.
  - Non-interest expenses and net fees and commissions modeled in relation to total asset growth.
  - Other net income (percent of total assets) negatively driven by the term premium and real estate prices (y-o-y).

### Risk-weighted assets and asset quality adjustments
- Risk weights:
  - Assumed constant at two levels: 90 percent and 100 percent.
  - 90 percent aligns with average risk weight observed at end-2015.
  - 100 percent used to reflect increase in risk weights during crises.
- Taxes and dividends:
  - Dividend payouts (if profits positive) and taxes fixed at 20 percent over the stress testing horizon.
- Adjustments for restructured bad loans and under-provisioning:
  - Restructured loans about 31 percent of large loans; one-third of restructured loans estimated to be of weaker loan quality at end-2015.
  - Migration of such loans increased provisions by about 1 percentage point in total.
  - Under-provisioning accounted for an additional increase in provisions of 2.2 percentage points.
  - Total effect of restructuring and under-provisioning on provisions assessed to be about 3.3 percentage points.

### Stress test results — headline figures
- Two sets of results:
  - First set: reported bank capital ratios at end-Q3 2015, no adjustments for regulatory forbearance, risk weights averaging 90 percent.
  - Second (more conservative) set: adjusts initial capital ratios for under-provisioning and restructuring, risk weights average 100 percent.
- First set (no asset-quality adjustment):
  - Estimated capital deficit of about ½ and 2 ½ percent of GDP in the baseline and stress scenarios respectively.
  - CBR and staff estimated aggregate capital deficit (peak during five-year horizon) of 0.3 and 0.4 percent of GDP, respectively, in the baseline scenario.
  - Deficit increases in stress scenarios compared to baseline by about 2 percentage points.
  - CBR estimated deficits of 2.5 and 2.1 percent of GDP in the V-shape and L-shape scenarios, respectively.
  - Staff estimated deficits of 2.4 and 2.7 percent of GDP in the V-shape and L-shape scenarios.
- Second set (with asset-quality adjustment):
  - Higher capital deficits of about 1 percent and 4 ½ percent of GDP in the baseline and in the stress scenarios, respectively.
  - Including restructuring and under-provisioning increases the capital deficit by 0.6 percentage points in the baseline and by about 2 percentage points in the stress scenarios.
  - Staff estimated capital deficits of 4.6 and 4.4 percent of GDP in the V-shape and L-shape scenarios respectively.
  - Peak deficits occur in different years: 2017 in the V-shape scenario, and 2019 in the L-shape scenario.

### Detailed system and bank-level outcomes
- V-shape scenario (detailed):
  - With asset quality adjustment, banks with capital ratio breaches account for 87 percent of banking sector assets in 2017.
  - Median CET1 capital and total capital ratios fall to 0.8 percentage point and 3.5 percentage points, respectively, in 2017.
  - Results without asset quality adjustment are less severe.
- L-shape scenario (detailed):
  - With asset quality adjustment, cumulative capital deficit accounts for 4.4 percent of GDP.
  - Median total capital ratio bottoms out at 7.8 percent of RWA in 2018.
- Baseline with asset quality adjustment:
  - Minimum regulatory capital requirements breached by 15 out of 37 banks in 2017, accounting for about 37 percent of banking sector assets.
  - Median CET1 capital and total capital ratios are 5.5 and 9.9 percent, respectively, in 2017, above regulatory minima.
- Medium-term profitability and recovery:
  - Some weak banks accounting for about 40 percent of banking sector assets have difficulties returning to profitability owing to low capital levels and profitability at end-2015, and moderate medium-term GDP growth projections.
  - CBR indicates weak large banks recover in the medium term; some medium- and small-sized banks (about 5 percent of banking system assets and not in staff’s sample) accumulate large capital deficits by 2020.

### Comparison of CBR and staff top-down results
- Overall similarity:
  - Corresponding CBR and staff top-down stress tests show similar capital deficits.
  - Panel data regression techniques broadly similar.
- Modeling differences:
  - Staff projects assets and equity (liabilities are residual).
  - CBR projects liabilities and equity (assets are residual).
  - Credit losses peak in first year in staff’s tests; CBR shows peak in second year owing to lagged effect.
  - Dynamics of capital ratios and capital deficits are very similar with bottoming in 2017 and recovery thereafter.

### Bottom-up versus top-down
- CBR top-down more conservative than banks’ bottom-up assessments:
  - Baseline: bottom-up shows one bank with capital deficit of RUB 3 billion; top-down shows one bank with capital deficit of RUB 65 billion.
  - V-shaped: bottom-up shows six banks with capital deficits of RUB 199 billion; top-down shows seven banks with capital deficits of RUB 522 billion (2.6 times higher).
  - L-shaped: bottom-up shows two banks with capital deficits of RUB 70 billion; top-down shows six banks with capital deficits totaling RUB 426 billion (6 times higher).
- Market risk appears to be underestimated by banks, alongside credit risk.

### Single-factor stress tests (CBR)
- Credit risk scenario:
  - Default of each bank’s top five (non-bank) borrowers.
  - 338 banks with 82.5 percent of banking system assets experience an aggregate capital deficit of RUB 1,519 billion.
- Market risk scenarios:
  - Exchange rate depreciation of 20 percent: five banks comprising 0.1 percent of system assets would have a capital deficit of RUB 0.2 billion.
  - 30 percent decline in value of equity holdings: seven banks comprising 0.5 percent of system assets would have a capital deficit of RUB 0.9 billion.
  - Interest rate shocks to trading portfolios:
    - Increase of 1,000 basis points in corporate sector spreads and increase of 400 basis points in government bond spreads: 13 banks comprising 6.7 percent of system assets would have an aggregate capital deficit of RUB 20 billion.
    - If only corporate sector spreads increase by 1,000 basis points (no increase in government bond spreads): four banks comprising 1 percent of system assets would have an aggregate capital deficit of RUB 6.7 billion.
- Interpretation:
  - Single-factor credit tests point to significant concentration risk in lending.
  - Market risk tests show relatively minor effects due to small size of securities portfolios.

### Liquidity stress tests (CBR)
- Assumptions:
  - Banks would not have access to CBR financing or interbank inflows.
- Scenarios:
  - Mild, severe, and very severe calibrated to historical outflow rates and international severe stress patterns.
- Results:
  - Mild scenario: 36 banks (1.7 percent of system assets) show liquidity deficits; total shortage less than 0.1 percent of GDP.
  - Severe scenario: 111 banks (11 percent of system assets) with shortage of 0.4 percent of GDP.
  - Very severe scenario: 181 banks (41.8 percent of system assets) with shortage of 2.9 percent of GDP.
- Role of large state-owned banks:
  - Three largest state-owned banks hold special liquidity positions: significant players in corporate (state-related) and retail deposit markets, hold large stocks of high quality assets, and have access to unsecured funding.
  - Sanctions restrict large banks' access to term external funding.
  - Lower-rated small and medium banks rely on higher-yielding collateral, short-term secured markets, and CBR.

### Stress testing recommendations
- Key priorities for CBR to improve stress-testing practices:
  - Implement modern, comprehensive, and integrated TD and BU methodologies covering banks and affiliated entities.
  - Develop an “expected loss approach” for credit risk using probability of default, loss given default, and exposure at default.
  - Develop a credit registry to provide additional data.
  - Expand use of TD bank-by-bank stress test results in supervision, including capital planning and dividend policy.
  - Consider possible publication of aggregate TD stress test results in the Financial Stability Review.
  - Enhance system-wide stress tests for all banks, beginning with large and medium-sized banks, using macro scenarios and key bank stress test parameter assumptions over a three-year horizon.
- Additional recommendations:
  - Conduct stress tests on a consolidated basis to capture downstreaming of problematic assets by D-SIBs to unconsolidated subsidiaries.
  - Conduct liquidity stress tests by currency to identify shortages in particular currencies that aggregate tests may overlook.

*Source: _cr16306 - 18.      The macroeconomic stress scenarios quantified the impact of negative oil price shocks*

### 45.      An asset quality review is recommended to assess risks stemming from uncertainty

### 45.      An asset quality review is recommended to assess risks stemming from uncertainty

### Recommendation
- An asset quality review is recommended to assess risks stemming from uncertainty about asset quality, concentrated lending, and related party lending.
- Stress tests should be supplemented by a granular and comprehensive review of banks’ asset portfolios to gauge capitalization needs.

### Implementation note from source
- "The Bank of Russia has already begun to implement the recommendations of FSAP mission: as part of banking supervision improvement the Risk Analysis Service has been organized. The Service is expected to assess assets and transactions of credit institutions on ongoing basis."

### Rationale and context (capital and liquidity)
- "Capital ratios remained stable due to capital injection ... and the use of regulatory forbearance to shield risk weighted assets from the impact of the depreciation."
- Liquidity developments highlighted:
  - "Liquidity dried up following a strong outflow of retail deposits in December 2014..."
  - "...which was compensated by CBR stepping in to support liquidity of the banking system."
  - "Since then, confidence returned and the liquidity situation has improved ...... ... amid declining loan-to-deposit ratios."
- Charts and series presented (labels preserved):
  - Capital Adequacy (In percent): Capital to risk-weighted assets; Core capital to risk-weighted assets; CAR Min. Req.; CoreCapital Min. Req. (time points include Jan-11 Apr-12 Jul-13 Oct-14 Feb-16)
  - Equity to Total Assets of Russian Banks (In percent) (time points include Apr-08 Jul-09 Oct-10 Jan-12 Apr-13 Jul-14 Oct-15)
  - Retail Deposits (Percent change, M-o-m, SA) (time points include Jan-08 Aug-09 Mar-11 Oct-12 May-14 Jan-16)
  - Main Funding Sources (In percent of total liabilities): Saving accounts; Funds received from CBR, rhs
  - Liquidity Assets to Total Assets (In percent)
  - Loan-to-Deposit Ratio (In percent) — Total; Ruble; FX

### Asset quality findings
- "NPLs remain lower than their 2008 peak ... although provisions are higher once FX loans, loans to FIs and governmental institutions are excluded."
- Sectoral and currency patterns:
  - "Overdue loans in rubles have increased primarily in the construction and retail trade sectors..."
  - "... while overdue loans in FX have risen primarily in the real estate, agriculture, and mining sectors."
  - "Nearly 25 percent of the construction portfolio is overdue..."
  - "Loans in FX show weak performance in agriculture, mining, and machinery and equipment."
- Charts and series presented (labels preserved):
  - Overdue Loans and NPLs (In percent of total loans) — Households; Corporate; NPLs (time points include Jan-08 Aug-09 Mar-11 Nov-12 Jun-14 Feb-16)
  - Provisions (In Russian ruble denominated loans; in percent of total loans) — Loans to corporates; Mortgage loans; Consumer loans (Apr-09 to Apr-15)
  - Ruble Overdue Loans to Total Loans by Sector (Mar-09 to Dec-15): Other; Real estate; Transport and communications; Agriculture, hunting and forestry; Electricity, gas and water; Manufacturing; Mining and quarrying; Wholesale and retail trade; Construction
  - Foreign Exchange Overdue Loans to Total Loans (In percent, by sector) (Mar-09 to Dec-15): Other; Wholesale and retail trade; Transport and communications; Construction; Electricity, gas and water; Manufacturing; Agriculture, hunting and forestry; Mining and quarrying; Real estate
  - Ruble Overdue Loans by Portfolio and Foreign Exchange Overdue Loans by Portfolio (Mar-09 to Dec-15) — portfolios include Wood products; Paper products; Construction; Mineral products; Machinery and equipment; Agriculture, hunting and forestry; Average

### Stress testing framework and methodology
- "Staff Stress Testing Framework" components (as presented):
  - Stress Testing Methodology
  - Macroeconomic Scenarios: Baseline, V-shape adverse and L-shape adverse
  - PROJECTIONS / ESTIMATION / DATA / MODELS/ASSUMPTIONS
  - Data: "Historical data on macroeconomic, financial variables, individual bank data on Balance Sheet and Income statement Items."
  - Models/assumptions: "Panel regression models; Constant shares; Estimate relashionships between Income Statement and Balance sheet items (dependent variables) and macroeconomic, financial variables and bank specific variables."
  - Other assumptions: "Risk weights, Asset Quality Adjustment, Dividend payout."
  - Projections: "Project Balance Sheet (or use assumptions of balance sheet dynamics). Project Income Statements. Common Equity Capital ratio."
  - Data sources and items referenced: "IMF; Bankscope"
- Income statement construction notes (verbatim labels and formulas preserved):
  - "Net interest income = Interest Income – Interest Expense"
  - "Securities Income (Return on Securities ) = Sum of 1. Net Gains (Losses) on Trading and Derivatives 2. Net Gains (Losses) on Other Securities 3. Net Gains (Losses) on Assets at FV through Income Statement 4. Change in Value of AFS Investments (part of Fitch Comprehensive Income)"
  - "Return on Securities = Securities Income/Securities (Balance Sheet)"
  - "Net Expenses = Expenses – Fee Income"
  - "Expenses = Sum of 1. Personnel Expenses 2. Other operating Expenses"
  - "Fee Income = Sum of 1. Net Fees and Commission"

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16306.pdf*

### 2.       Other Operating Income

### 2.       Other Operating Income

### Macroeconomic Scenarios and Key Projections
- Staff designed three scenarios (Baseline, V-Shape, L-Shape) based on option implied oil price scenarios.  
- Output responds according to oil price changes; decline in GDP would be similar to previous recessions; the ruble depreciates when oil prices fall and unemployment rises.  
- Table: Macroeconomic Scenarios: Projections 2015–20 (selected series)
  - A. Baseline Scenario (2015 actual; 2016–20 projections)
    - Real GDP (%): 2015 = -3.7; 2016 = -1.0; 2017 = 1.0; 2018 = 1.5; 2019 = 1.5; 2020 = 1.5
    - Brent price (USD/bbl, avg.): 2015 = 53; 2016 = 43; 2017 = 50; 2018 = 54; 2019 = 57; 2020 = 60
    - Exchange Rate (RUB/USD, avg.): 2015 = 61; 2016 = 73; 2017 = 68; 2018 = 64; 2019 = 62; 2020 = 61
    - Inflation (%, avg.): 2015 = 15.5; 2016 = 8.3; 2017 = 7.0; 2018 = 5.0; 2019 = 4.0; 2020 = 4.0
    - Short-term Interest Rates (%, avg.): 2015 = 11.0; 2016 = 8.5; 2017 = 5.5; 2018 = 5.5; 2019 = 5.5; 2020 = 5.5
    - Unemployment (%, avg.): 2015 = 5.6; 2016 = 6.4; 2017 = 6.1; 2018 = 6.0; 2019 = 6.0; 2020 = 6.0
  - B. V-Shape Scenario
    - Real GDP (%): 2015 = -3.7; 2016 = -7.8; 2017 = 2.6; 2018 = 2.9; 2019 = 1.1; 2020 = 1.0
    - Brent price (USD/bbl, avg.): 2015 = 53; 2016 = 19; 2017 = 24; 2018 = 32; 2019 = 36; 2020 = 40
    - Exchange Rate (RUB/USD, avg.): 2015 = 61; 2016 = 100; 2017 = 88; 2018 = 75; 2019 = 71; 2020 = 67
    - Inflation (%, avg.): 2015 = 15.5; 2016 = 15.9; 2017 = 6.1; 2018 = 4.0; 2019 = 4.2; 2020 = 4.2
    - Short-term Interest Rates (%, avg.): 2015 = 11.0; 2016 = 17.7; 2017 = 7.6; 2018 = 5.4; 2019 = 5.7; 2020 = 5.7
    - Unemployment (%, avg.): 2015 = 5.6; 2016 = 9.5; 2017 = 9.2; 2018 = 8.0; 2019 = 7.4; 2020 = 6.9
  - C. L-Shape Scenario
    - Real GDP (%): 2015 = -3.7; 2016 = -5.7; 2017 = -0.3; 2018 = 1.7; 2019 = 1.1; 2020 = 1.0
    - Brent price (USD/bbl, avg.): 2015 = 53; 2016 = 25; 2017 = 25; 2018 = 29; 2019 = 33; 2020 = 37
    - Exchange Rate (RUB/USD, avg.): 2015 = 61; 2016 = 88; 2017 = 91; 2018 = 83; 2019 = 78; 2020 = 74
    - Inflation (%, avg.): 2015 = 15.5; 2016 = 13.3; 2017 = 5.4; 2018 = 4.1; 2019 = 4.2; 2020 = 4.2
    - Short-term Interest Rates (%, avg.): 2015 = 11.0; 2016 = 15.2; 2017 = 6.9; 2018 = 5.5; 2019 = 5.7; 2020 = 5.7
    - Unemployment (%, avg.): 2015 = 5.6; 2016 = 8.9; 2017 = 9.5; 2018 = 8.6; 2019 = 8.0; 2020 = 7.4
- Figure findings:
  - GDP growth is highly correlated with changes in oil prices.
  - Ruble depreciates strongly when oil prices fall, especially after 2008.
  - Inflation pass-through is high; higher inflation is accommodated by higher interest rates in line with CBR’s objective.

### Asset Quality Adjustments and Loan Impairment
- Asset quality adjustments were based on the aggregate loan portfolio; discussions led to an estimate of weak restructured loans at 30 percent of restructured loans.
- Combined effect of loan migration and higher provisioning results in loan impairment charge of 3.3 percentage points.
- Loan Quality of Russian Banking System (01/01/2016) — selected entries (absolute amounts and ratios as provided)
  - Current LP (levels): Standard = 26,254; Substandard = 22,237; Doubtful = 4,769; Problem = 1,409; Loss = 3,442; Total = 58,111
  - Current Prov. (levels): Standard = [not listed]; Substandard = 400; Doubtful = 858; Problem = 579; Loss = 2,654; Total Prov. = 4,492
  - Restructured Loans: 30.7%
  - Bad Loans: 30.0%
- Provisioning Adjustment (Adjusted Prov.):
  - Standard: Adjusted Prov. = 758 (Adj. Prov. Ratio = 3%)
  - Substandard: Adjusted Prov. = 1,022 (Adj. Prov. Ratio = 21%)
  - Doubtful: Adjusted Prov. = 696 (Adj. Prov. Ratio = 49%)
  - Loss: Adjusted Prov. = 3,279 (Adj. Prov. Ratio = 95%)
  - Total Adjusted Prov. = 5,755; resulting Loan Imp. Charge = 2.2% (from provisioning adjustment alone)
- Migration of Restructured Loans and Provisioning Adjustment yields final Loan Imp. Charge = 3.3%

### Stress Tests: V-Shape Scenario and Results
- Credit losses peak in the first two years in the V-Shape scenario.
- Average capital ratios (CBR) bottom out in 2017 and recover by 2020.
- Median capital ratios (staff) breach total capital ratio threshold (CAR Threshold) in 2017, and recover slightly above by 2020.
- Total capital deficit is about 2½ percent of GDP (CBR and staff), with additional capital charge of about 2 percent of GDP (staff) due to asset quality adjustments.
- When adjusting for forbearance, median capital ratios (Staff) are more affected.
- Stress test quantitative highlights (V-Shape, Top-down, Staff with asset quality adjustment — selected lines)
  - Credit losses (In percent of total loans): Year 2016 = 6.2; 2017 = 1.9; 2018 = 1.7; 2019 = 1.8; 2020 = 1.8
  - Total capital deficit (In percent of GDP): 2016 = 0.5; 2017 = 0.8; 2018 = 0.9; 2019 = 1.0; 2020 = 1.0
  - No. of banks with capital deficit (Top-down, Staff with asset quality adjustment): 2016 = 15; 2017 = 15; 2018 = 15; 2019 = 14; 2020 = 12
  - Share of banks with capital deficit (percent of sample total assets): 2016 = 37.0; 2017 = 37.0; 2018 = 37.7; 2019 = 37.5; 2020 = 18.7
- Top-down (CBR) without asset quality adjustment — Total capital ratio (asset weighted mean)
  - 2016 = 12.3; 2017 = 12.5; 2018 = 12.6; 2019 = 12.7; 2020 = 12.8
- Top-down (Staff) without asset quality adjustment — Median CET1 capital ratio
  - 2016 = 9.0; 2017 = 8.4; 2018 = 10.1; 2019 = 12.4; 2020 = 14.8

### Detailed Stress Test Outcomes (selected aggregates and metrics)
- Top-down, CBR (sample of 681 banks) — Credit losses (in percent of total loans) summed across years (Baseline vs Stress Scenarios): cumulative patterns show higher losses and larger capital deficits under stress scenarios.
- Top-down, Staff (sample of largest 37 banks) — With asset quality adjustment, total capital deficit (In percent of GDP) peaks around:
  - 2016 = 3.8; 2017 = 4.6; 2018 = 4.4; 2019 = 4.3; 2020 = 4.2 (includes market and liquidity risk)
- Coverage: Staff top-down exercises cover about 80% of total banking assets.

### Banks’ Bottom-Up and CBR Top-Down Results for 12 Banks (2016)
- Bottom-Up (12 banks) — Main results (bn RUB)
  - Baseline: Total losses = 947; Losses on credit risk = 786; Financial result = 730; Aggregate capital deficit = 3; Number of banks with capital deficit = 1
  - V-Shape Scenario: Total losses = 2,882; Losses on credit risk = 1,978; Financial result = -1,159; Aggregate capital deficit = 199; Number of banks with capital deficit = 6
  - L-Shape Scenario: Total losses = 2,122; Losses on credit risk = 1,536; Financial result = -466; Aggregate capital deficit = 70; Number of banks with capital deficit = 2
  - Breakdown: Corporate portfolio losses (V-Shape) = 1,572; Retail portfolio losses (V-Shape) = 407
  - Market risk items (V-Shape): Equity risk = 76; FX risk = 54; Interest rate risk on trading portfolio = 17
- CBR Top-Down (same 12 banks) — Main results (bn RUB)
  - Baseline: Total losses = 1,687; Losses on credit risk = 1,277; Financial result = 334; Aggregate capital deficit = 65; Number of banks with capital deficit = 1
  - V-Shape Scenario: Total losses = 3,746; Losses on credit risk = 2,221; Financial result = -680; Aggregate capital deficit = 522; Number of banks with capital deficit = 7
  - L-Shape Scenario: Total losses = 3,425; Losses on credit risk = 2,031; Financial result = -505; Aggregate capital deficit = 426; Number of banks with capital deficit = 6
  - Market risk items (V-Shape): Interest rate risk on trade portfolio = 412; Equity risk = 225; FX risk = 112
  - Losses on interest rate risk on banking book (V-Shape) = 720

### Single-Factor and Liquidity Tests
- Single-factor tests (end-2015) — selected stress outcomes (CBR estimates)
  - Default of top five bank borrowers: Number of banks with capital deficit = 338; Share of total banking assets = 82.5%; Aggregated capital deficit = 1,519 bn RUB
  - Exchange rate depreciation of 20 percent: Number of banks with capital deficit = 5; Share of total banking assets = 0.1%; Aggregated capital deficit = 0.2 bn RUB
  - Equity decline of 30 percent: Number of banks with capital deficit = 7; Share of total banking assets = 0.5%; Aggregated capital deficit = 0.9 bn RUB
  - Interest rate risk shocks and trading book scenarios produce smaller numbers of affected banks but nontrivial aggregated deficits (examples listed in table).
- Assumptions of Liquidity Tests (CBR estimates) — Outflows and assumptions by scenario (in percent)
  - Mild Scenario:
    - Individual deposits outflow = 20 (10 for largest bank)
    - Deposits of legal entities outflow = 10
    - Funds of legal entities in settlement and other accounts outflow = 20
    - Interbank loans from non-residents outflow = 30
    - Interbank loans from residents outflow = 0
    - Fire Sales discounts: Cash and correspondent accounts with the CBR = 0; Highly liquid assets = 5; Liquid assets = 20; Securities not included in liquid assets = 30
  - Severe Scenario:
    - Individual deposits outflow = 30 (10 for largest bank)
    - Deposits of legal entities outflow = 10
    - Funds of legal entities in settlement and other accounts outflow = 30
    - Interbank loans from non-residents outflow = 50
    - Interbank loans from residents outflow = 0
    - Fire Sales discounts: Cash and correspondent accounts with the CBR = 0; Highly liquid assets = 5; Liquid assets = 20; Securities not included in liquid assets = 40
  - Very Severe Scenario:
    - Individual deposits outflow = 30 (10 for largest bank)
    - Deposits of legal entities outflow = 30
    - Funds of legal entities in settlement and other accounts outflow = 40 (20 for largest bank)
    - Interbank loans from non-residents outflow = 50
    - Interbank loans from residents outflow = 40 (20 for largest bank)
    - Fire Sales discounts: Cash and correspondent accounts with the CBR = 0; Highly liquid assets = 5; Liquid assets = 20; Securities not included in liquid assets = 65

### Regression and Balance Sheet–Income Relationships
- Regression results (selected coefficients and diagnostics; source: IMF staff calculations)
  - A: Balance Sheet Items
    - Gross loans (yoy %) equation includes lagged gross loans L, coefficient 0.382*** (standard error 0.0310)
    - Total assets (yoy %) equation: Gross loans (yoy %) coefficient 0.853*** (0.0113); Net Charge Offs (% Gross Loans) includes a coefficient -0.0158*** (0.00215)
    - Real GDP (yoy %) coefficient in gross loans equation = -0.125*** (0.0306)
    - Observations: 1,487; R-squared = 0.361; Firm FE = YES; YEAR>2007 = YES
  - B: Income Statement Items (selected)
    - Loans * Lending Rate (yoy %) = L coefficient = 0.445*** (0.0342) in Net interest income (yoy %)
    - Deposits * Deposit Rate (yoy %) = L coefficient = -0.168*** (0.0264) in Net interest income (yoy %)
    - Real GDP (yoy %) effect on Net Fees and Commissions = -0.0290*** (0.00314)
    - Unemployment coefficient in Non-interest Expense (yoy %) = 0.101*** (0.0207)
    - Observations across income regressions: range 1,136 to 1,554; R-squared values vary (e.g., 0.354, 0.152, 0.232, 0.132, 0.412, 0.559)
    - Robust standard errors in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1

*Source: IMF staff estimates and tables in the provided chapter.*

### 2. Channels of

### 2. Channels of Risk Propagation

### Methodology (Balance-sheet / Satellite models / Stress test horizon)
- Top-Down by FSAP Team:
  - Balance-sheet model (loan impairment charge/ provisions/NPLs)
  - Satellite Models for Macrofinancial linkages: Models for credit losses, NPLs, loan impairment charge, pre-impairment income
  - Satellite model bank by bank using IFRS data
  - Stress test horizon: Five years, 2016–20
- Top-Down by Authorities:
  - Supplemental balance sheet analysis of impact of restructuring and under-provisioning
  - Models for credit losses, provisions, pre-impairment income
  - CBR macro model has second round effects related to the dependency between credit growth and growth of fixed capital investments
  - Stress test horizon: Five years, 2016–20
- Bottom-Up by Authorities:
  - Balance-sheet model (NPLs/provisions)
  - Banks’ internal models
  - Own models
  - Stress test horizon: One year

### Tail shocks / Scenario analysis / Sensitivity analysis
- Top-Down by FSAP Team:
  - Macro scenarios are shocks conditioned upon oil price, GDP, exchange rate, inflation, interest rates, unemployment
  - Baseline, adverse V-shaped scenario, adverse L-shaped scenario (magnitude of shocks derived from adverse oil price scenario plus macro model)
  - Sensitivity analysis: Not performed by staff
- Top-Down by Authorities:
  - Macro scenarios are shocks conditioned upon oil price, GDP, exchange rate, inflation, interest rates, unemployment
  - Baseline, adverse V-shaped scenario, adverse L-shaped scenario
  - Sensitivity analysis: Single-factor shocks are applied; concentration risk: default of top five borrowers; market risks: FX 20 percent depreciation, equity 30 percent decline, interest rate +1000 bps corporate bonds and+400 bps, government bonds
- Bottom-Up by Authorities:
  - Macro scenarios are shocks conditioned upon oil price, GDP, exchange rate, inflation, interest rates, unemployment.
  - Baseline, adverse V-shaped scenario, adverse L-shaped scenario
  - Sensitivity analysis: Not performed by banks

### Risks and Buffers (Risks assessed / Behavioral adjustments)
- Top-Down by FSAP Team:
  - Risks/factors assessed: Credit losses, profitability, income after tax, funding costs, exchange rate
  - Behavioral adjustments: Dividend payout can be restricted by CBR.
- Top-Down by Authorities:
  - Risks/factors assessed: Credit losses, profitability, income after tax, funding costs, market risk, fixed income holdings of banks/sovereigns, exchange rate
  - Behavioral adjustments: Dividend payout can be restricted by CBR.
- Bottom-Up by Authorities:
  - Risks/factors assessed: Same as authorities’ TD stress tests
  - Behavioral adjustments: Banks’ internal models

### Regulatory and Market-Based Standards and Parameters (Calibration / Standards)
- Calibration of risk parameters:
  - Top-Down by FSAP Team:
    - NPLs/provisions
    - RWA according to standardized approach
    - Risk weights increase from 90 percent to 100 percent in the prudent approach
  - Top-Down by Authorities:
    - NPLs/provisions
    - RWA according to standardized approach
  - Bottom-Up by Authorities:
    - Own methodologies
    - RWA similar to authorities’ stress tests
- Regulatory/Accounting and Market-Based Standards:
  - Top-Down by FSAP Team:
    - Hurdle rate: Basel III schedule, local regulatory requirements
    - CET1 of 4.5 percent, CAR of 8 percent
  - Top-Down by Authorities:
    - Hurdle rate: Basel III schedule, local regulatory requirements
    - CET1 of 4.5 percent, CAR of 8 percent
  - Bottom-Up by Authorities:
    - Hurdle rate: Basel III schedule, local regulatory requirements
    - CET1 of 4.5 percent, T1 of 6 percent, CAR of 8 percent

### Reporting Format for Results
- Top-Down by FSAP Team:
  - Output presentation: Dispersion of capital ratios: median
  - Capital deficit, system wide
  - Pass or fail; percentage of assets that fail
  - Number of banks undercapitalized
- Top-Down by Authorities:
  - Output presentation: Distributions by capital ratio: weighted average
  - Capital deficit, system wide
  - Pass or fail; percentage of assets that fail
  - Number of undercapitalized banks
- Bottom-Up by Authorities:
  - Output presentation: Capital shortfall, system wide.
  - Pass or fail; percentage of assets that fail
  - Number of banks undercapitalized

### Asset quality adjustment
- Top-Down by FSAP Team:
  - Stress testing methodology: Adjustment for weak restructured loans and under-provisioning lead to one-off increase in provisions at beginning of stress testing period
- Top-Down by Authorities:
  - Stress testing methodology: Not performed by authorities
- Bottom-Up by Authorities:
  - Stress testing methodology: Not performed by banks

---

### BANKING SECTOR: LIQUIDITY RISK

### Institutional Perimeter / Data
- Top-Down by FSAP Team:
  - Institutions included: Carried out by authorities with agreed scenarios
  - Market share: N/A
  - Data and baseline date: N/A
- Top-Down by Authorities:
  - Institutions included: 681
  - Market share: 99.7 percent
  - Data and baseline date: Institutions’ public data and other data up to 4th Q 2015
- Bottom-Up by Authorities:
  - Institutions included: Banks included in authorities’ liquidity scenarios
  - Market share: N/A
  - Data and baseline date: Stand-alone basis / N/A

### Channels of Risk Propagation (Methodology)
- Top-Down by FSAP Team: N/A
- Top-Down by Authorities: Cash outflow asset discounts and sensitivity analysis for liquidity risk
- Bottom-Up by Authorities: N/A

### Risks and Buffers (Liquidity risks / Buffers)
- Top-Down by FSAP Team: N/A
- Top-Down by Authorities:
  - Risks: Outflows from liabilities liquidity shock; Discounts on assets; Market liquidity shock
  - Buffers: Counterbalancing capacity stand-alone
- Bottom-Up by Authorities: N/A

### Tail shocks (Size of the shock)
- Top-Down by FSAP Team: N/A
- Top-Down by Authorities:
  - Bank run and dry up of wholesale funding markets, taking into account haircuts to liquid assets
  - Outflows: household deposits-30 percent (10 percent largest bank); legal entities 30 percent; corporate settlement and other accounts-40 percent (20 for largest bank); interbank loans non-resident 50 percent; interbank loans resident 40 percent (20 for largest bank)
  - Haircut: highly liquid asset, 5 percent; liquid asset 20 percent; low-liquid asset 65 percent
- Bottom-Up by Authorities: N/A

### Regulatory standards / Reporting
- Regulatory standards: N/A
- Amount of liquidity deficit and number of banks with liquidity deficit: N/A
- Reporting Format for Results: Amount of liquidity deficit and number of banks with liquidity deficit: N/A

*Source: _cr16306 - 2. Channels of*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2016/_cr16306.pdf_