## The Global Bank Stress Test (GST)

## Source details

**Canonical URL:** [The Global Bank Stress Test (GST)](https://www.imf.org/-/media/files/publications/dp/2022/english/gbstea.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/dp/2022/english/gbstea.pdf.md)
- [Structured JSON version](/-/media/files/publications/dp/2022/english/gbstea.pdf.json)

---

### Executive Summary — Purpose, framework, and headline findings
- Purpose and scope:
  - Assess systemic risk from global shocks to banking systems using a consistent cross-country framework.
  - Inform IMF multilateral surveillance (for example, through the GFSR) and bilateral surveillance, and benchmark national authorities’ stress tests.
  - GST launched in the October 2020 GFSR as the first cross-country, globally consistent, macro scenario–based stress testing exercise using publicly available data.
- Framework and approach:
  - Uses publicly available, aggregated data and a common methodology to assess the impact of global macro-financial scenarios on bank capital across major advanced economies (AEs) and emerging markets (EMs).
  - Captures cross-border spillovers via financial and real linkages in the scenarios and implicit spillbacks via multilaterally consistent scenarios.
  - Produces baseline and adverse scenario simulations (the update illustrated here uses scenarios for 2022–24 based on the October 2021 World Economic Outlook (WEO)).
  - A user-friendly tool based on the GST methodology has been developed for assessing solvency risks in AEs and EMs.
  - GST is complementary to, not a substitute for, more granular FSAPs or authorities’ stress tests.
- Key headline findings from the 2021/22 update:
  - Banks in AEs raised Capital Tier 1 (CET1) ratios by about 0.8 percentage points in 2020 and are generally resilient to continued pandemic shocks under the scenarios considered.
  - Banks in EMs exhibit relatively lower resilience compared with AEs:
    - EM banks face larger downside macro-financial risks in the adverse scenario.
    - EM banks have higher sensitivity to shocks.
    - EM banks built somewhat less capital during the peak of the pandemic in 2020.
  - Under the October 2020 GST application (illustrative historical result):
    - In the adverse scenario, a weak tail of banks corresponding to 8.3 percent of banking system assets would fail to meet minimum regulatory requirements, and the capital shortfall relative to broad statutory regulatory thresholds would reach $220 billion.
    - If COVID-19-related mitigation policies were not implemented, the weak tail of banks would reach 14 percent of banking system assets, and the global capital shortfall would be $420 billion.
  - A severe adverse (tail) scenario shows larger effects on bank capital, with EMs more exposed, underscoring material downside risks in tail events.
- Policy implications:
  - GST provides a high-level global assessment of potential pressure points that could affect domestic banking sector stability and banks’ ability to supply credit to the real economy.
  - Serves as a benchmarking tool for national authorities to compare resilience across countries and to inform capitalization policy decisions.
  - Highlights the importance of continued close monitoring of banking systems, especially in EMs with more restrained macro-financial policy space.
- Limitations:
  - Aggregate, publicly available data imply a simpler methodology than supervisory or FSAP-level stress tests; GST does not fully capture borrower-specific support policies extensively used during the pandemic.
  - The aggregate effects of borrower-support policies should be captured in the macro-financial scenarios that feed the GST.
  - Framework can be extended to incorporate direct effects of borrower-support policies; staff discuss approaches to account for such policies in the GST.

### How the GST works — coverage, data, methodology, and assumptions
- Coverage and dataset:
  - Sample coverage:
    - 29 economies: 24 AEs and 5 EMs.
    - The banking sector assets of these 29 economies account for 70 percent of global banking assets.
    - In each economy, banks covering at least 80 percent of the individual banking system’s total assets are included.
    - Sample composition: 53 banks in EMs and 204 banks in AEs.
  - Data sources and frequency:
    - Comparable data for bank income statements, balance sheets, and several risk metrics were sourced primarily from Fitch Connect and banks’ financial reports.
    - Historical bank dataset has an annual frequency spanning a 25-year period from 1995 to 2020.
    - Cutoff date for bank data in the 2022–24 update: end-2020.
- Methodological approach and modules:
  - Core structure:
    - Two-part methodology: (1) econometric models for P&L and OCI components; (2) a balance sheet projection module that maps P&L, RWAs, and OCI into CET1 capital ratios and capital shortfalls over the stress horizon.
    - GST projects banks’ income statement components, risk weights, and the implied capital ratios over a three-year horizon. Capital shortfall estimates are computed at the bank-level.
  - Econometric models:
    - Cross-bank-country panel regression models—bank fixed effects (FE).
    - P&L components modeled separately and aggregated via accounting identities: (1) net loan losses (with structural PD/LGD elements); (2) net interest income (NII); (3) net trading income (NTI); (4) net fee and commission income (NFCI); (5) other income/expense (residual).
  - Loan-losses and PD/LGD inference:
    - Loan-loss model coupled with a structural element decomposing loss rates into PDs and LGDs using a variant of the Frye-Jacobs (2012) methodology.
    - PDs and LGDs are inferred from projections of loan losses to derive dynamics of performing and nonperforming loan stocks. Migration of nonperforming loans back to performing status (“cures”) is allowed implicitly but not explicitly modeled.
  - NTI projection:
    - NTI ratio = average NTI ratio over the last five years minus (scalar × standard deviation of NTI ratio over the last five years).
    - Scalar is common for all banks; in specific severe scenarios a higher scalar may be used.
  - Balance sheet module and policy assumptions:
    - Bank-specific balance sheet information as of 2020 used as starting point.
    - User-defined assumptions for dividend distribution and taxes:
      - Tax rates and dividends set to zero if projected net income before taxes is negative.
      - Otherwise, tax rates = individual banks’ effective tax rates in the last year of the sample with a cap of 30 percent.
      - Dividend payout ratios = 50 percent of net income before tax if positive, and zero otherwise.
    - No deferred tax asset accumulation is considered.
- Risk-weighting and regulatory treatment:
  - Credit RWAs:
    - Total credit exposures split into standardized (STA) and internal rating based (IRB) approaches approximated from public disclosures.
    - STA risk-weighted densities are assumed constant over the stress horizon.
    - IRB densities adjusted using the Basel formulae (BCBS 2019).
    - Through-the-cycle PDs are estimated from scenario-dependent point-in-time PDs and a “smoothness” parameter set at 15 percent.
    - Downturn LGDs are held constant over the stress testing horizon.
  - Other RWAs (market, operational, residual) assumed constant.
  - Note: these assumptions likely underestimate the impact of adverse scenarios on bank capital because under stress downturn LGDs and other RWAs would be expected to increase.
- Innovations:
  - Internally and globally consistent scenarios anchored on WEO and developed using FSGM.
  - Captures cross-border spillovers and implicit spillbacks; cross-border spillovers via subsidiaries’ capital shortfalls on parent banks (inclusion/exclusion depends on exercise objective).
  - Unique global bank dataset spanning 1995–2020 sourced from public data.
  - Accounting for model uncertainty via Bayesian Model Averaging (BMA): BMA estimates about 5,000 models for each dependent variable and assigns weights based on predictive power.
  - Inferring PDs and LGDs from loan losses and using Basel formulae to project RWAs due to lack of granular supervisory data.

### Scenario design — 2022–24 update and adverse/severe adverse scenarios
- Purpose and anchoring:
  - GST updated for 2022–24 to assess risks related to COVID-19 aftermath and emerging uncertainties.
  - Exercise anchored on the October 2021 WEO baseline and informed by adverse scenarios developed for the April 2021 WEO using FSGM.
- Adverse scenario design:
  - GST adverse scenario is a linearly scaled extension of the WEO downside scenario illustrating consequences of a 2½ standard deviation shock to global growth rates, relative to the baseline, in 2022–24.
  - A second severe adverse scenario explores outcomes further into the tail of the growth distribution for each country.
- Baseline growth projections for the GST sample (exact text preserved):
  - After a contraction of almost 5 percent in 2020 and a rebound of 5½ percent in 2021, output in the group of countries included in the current GST baseline sample was projected to grow at about 42/3 percent in 2022 and 2.3 percent in 2023.
  - Growth is expected to moderate to about 2 percent over the medium term.
- Adverse scenario assumptions:
  - Considers adverse impacts from new variants, vaccine efficacy, and pace of vaccine rollout leading to slower activity and higher risk aversion.
  - Unconventional monetary policy measures assumed to prevent significant increases in sovereign rates.
  - No additional discretionary fiscal measures or policies are assumed; lack of conventional monetary policy space and shrinking fiscal space limit policy responses.
  - Weaker rebound leads to more pronounced scarring than in the baseline despite a gradual return in mobility to pre-pandemic levels.
- Calibration parameters:
  - Smoothness parameter for through-the-cycle PDs set at 15 percent.
  - NTI scalar common across countries; higher scalar used in scenarios/severe years with severely constrained market volumes or agency business.
  - Dividend payout ratio set at 50 percent of net income before tax if positive; zero otherwise.
  - Tax rate cap at 30 percent.
- GST adverse scenario macroeconomic shocks (summary):
  - In the adverse scenario, global GDP is about 5 percentage points and 2½ percentage points lower than in the baseline in 2022 and 2023, respectively.
  - By 2024, the level of GDP is:
    - about 6.5 percentage points lower than in the baseline in Advanced Economies (AEs); and
    - about 9.4 percentage points lower than in the baseline in Emerging Markets (EMs).
  - The larger GDP shock in EMs is reflected in more severe shocks to unemployment, credit spreads, and stock prices.

### GST results, heterogeneity, and severe tail outcomes
- Baseline and adverse CET1 dynamics:
  - Baseline scenario:
    - Banks, on average, would see a strong upward trend in their CET1 ratio, reaching 17.2 percent in 2024, 3.2 percentage points above the end-2021 level.
  - Adverse scenario:
    - The average CET1 ratio would fall moderately in the first year—a drop of 1.4 percentage points below the 2021 level and 2.7 percentage points below the baseline scenario.
  - Decline in CET1 ratios over the stress testing horizon stems mainly from an increase in loan loss provisions.
  - In the baseline scenario, higher loan loss provision expenses contribute to a 2 percentage-point decline in the global CET1 ratio, 2.2 percentage points less than in the adverse scenario.
  - For all banks, the contribution of other risk factors to the difference in CET1 ratios between baseline and adverse in 2024 ranges from 0.4 percentage points for risk-weighted assets (RWA) to 0.8 percentage points for net interest income (NII).
- AE versus EM heterogeneity:
  - AE banks increased CET1 ratios by about 0.8 percentage points in 2020.
  - In EMs, the decline in CET1 ratios in the adverse scenario reaches 2.5 percentage points below the starting point and 5.5 percentage points below the baseline in 2024.
  - Cross-country and cross-bank heterogeneity: changes in CET1 ratios (difference between 2021 and trough) reach up to 7 percentage points in some cases.
  - Capital shortfalls, calculated relative to a 4½ percent minimum capital level, range from zero to 0.6 percent of GDP in the adverse scenario across economies.
  - The share of banks by assets experiencing shortfalls is higher in emerging markets: 14 percent of total bank assets in 2024.
- Severe adverse scenario design and implications:
  - Severe adverse scenario calibrated by assuming GDP growth in 2022 is equal to half of the GDP growth in 2020 for each country; corresponds to a four standard deviation shock from the baseline and equivalent to a 2½ standard deviations shock from average GDP growth spanning the years preceding the pandemic.
  - Severe adverse scenario points to continued downside risks to the global banking system in such a tail event and illustrates greater exposure to downside shocks, especially in emerging markets.

### Accounting for mitigating policies and micro–macro extensions
- Limitations of public data:
  - Publicly available data structure precludes full analysis of the direct impact of borrower-support policies on bank capital (for example, cannot break down loan losses by corporations versus households).
  - Aggregate analysis could understate the positive effect on bank capital of financial policies to support household and corporate borrowers.
- GST methodological extensions to account for policies:
  - Extension decomposes aggregate loan loss provision forecasts into contributions from corporate and retail loan portfolios using direct step regression specifications (local projection method).
  - Two micro data-based modeling approaches can be “plugged into” the GST to account for policies that affect PDs and LGDs through payment moratoria and government guarantees.
  - Mechanisms: moratoria primarily affect PDs; guarantees primarily affect LGDs but may also reduce PDs via lower lending rates and reduced debt service burden.
- Corporate stress testing (Box 3 summary):
  - Tressel and Ding (2021) tool stress-tests publicly listed nonfinancial corporates in 24 AEs and EMs; firm-level metrics include liquidity (interest coverage ratio and cash balance), viability (interest coverage ratio and price-to-book ratio), and solvency (book value of equity).
  - Firm outcomes aggregated to country level and linked to bank loan and bond markets’ PDs to assess systemic risk and spillovers.
- Household micro–macro simulation (Box 4: IDHBS model):
  - Integrated Dynamic Household Balance Sheet (IDHBS) model obtains scenario- and policy-counterfactual estimates of PDs and LGDs for the household sector using large-scale micro databases.
  - Four-layer structure: macro-financial scenarios → employment status simulation → household-level PDs and LGDs → household balance sheet simulation → bank-level and country-level impacts.
  - Key drivers: unemployment, interest rates, house prices.
  - Uses household microdata to simulate P&L and balance sheets, translate to PDs/LGDs, and compute bank loan losses and RWAs.
  - Policy counterfactuals for debt moratoria suggest household-sector policies can have notable mitigating effects and can significantly shield banks from incurring material capital losses.

### Annex figure summary — GST macroeconomic scenarios for Emerging Markets (2021–2024)
- Figure presents Baseline and Severe adverse scenario paths for EMs across five macro variables for 2021–2024:
  - GDP Scenarios (Level, 2021=100): axis tick/label values shown: 90, 100, 110, 95, 105, 115, 120; time horizon labels: 2021, 2022, 2023, 2024.
  - Unemployment Rate Scenarios (Percent): axis tick/label values shown: 0, 15, 5, 10; time horizon labels: 2021, 2022, 2023, 2024.
  - Stock Price Scenarios (Level, 2021=100): axis tick/label values shown: 0, 150, 200, 50, 100; time horizon labels: 2021, 2022, 2023, 2024.
  - Corporate Credit Spread Scenarios (Percentage points): axis tick/label values shown: 0, 5, 1, 2, 3, 4; time horizon labels: 2021, 2022, 2023, 2024.
  - Term Premium (Percent): axis tick/label values shown: 0, 6, 1, 2, 3, 4, 5; time horizon labels: 2021, 2022, 2023, 2024.
- Sources for scenarios and estimates: IMF, October 2021 World Economic Outlook; Bloomberg; IMF staff estimates.

*Source: Excerpts from "The Global Bank Stress Test" — Executive Summary; "2. How Does the GST Work?"; Box 1 "Designing Adverse Scenario in FSAPs"; Box 4 "Household Debt: Assessing Mitigating Policies Using a Micro–Macro Simulation Model"; Annex Figure 3.2 (IMF staff).*

### Executive Summary  ................................................................................................v

### Executive Summary

### Purpose and Scope
- Assess systemic risk from global shocks to banking systems using a consistent cross-country framework.
- Provide insights to inform IMF multilateral surveillance (for example, through the GFSR) and bilateral surveillance, and to benchmark national authorities’ stress tests.
- GST launched in the October 2020 GFSR as the first cross-country, globally consistent, macro scenario–based stress testing exercise using publicly available data.

### GST Framework and Approach
- Uses publicly available, aggregated data and a common methodology to assess the impact of global macro-financial scenarios on bank capital across major advanced economies (AEs) and emerging markets (EMs).
- Captures cross-border spillovers via financial and real linkages in the scenarios and implicit spillbacks via multilaterally consistent scenarios.
- Produces baseline and adverse scenario simulations (the update illustrated here uses scenarios for 2022–24 based on the October 2021 World Economic Outlook (WEO)).
- A user-friendly tool based on the GST methodology has been developed for assessing solvency risks in AEs and EMs.
- The GST is complementary to, not a substitute for, more granular FSAPs or authorities’ stress tests because publicly available data constrain granularity and the ability to model borrower-specific support policies.

### Key Findings from the 2021/22 Update
- Banks in AEs raised Capital Tier 1 (CET1) ratios by about 0.8 percentage points in 2020 and are generally resilient to continued pandemic shocks under the scenarios considered.
- Banks in EMs exhibit relatively lower resilience compared with AEs:
  - EM banks face larger downside macro-financial risks in the adverse scenario.
  - EM banks have higher sensitivity to shocks.
  - EM banks built somewhat less capital during the peak of the pandemic in 2020.
- Under the October 2020 GST application (illustrative historical result cited):
  - In the adverse scenario, a weak tail of banks corresponding to 8.3 percent of banking system assets would fail to meet minimum regulatory requirements, and the capital shortfall relative to broad statutory regulatory thresholds would reach $220 billion.
  - If COVID-19-related mitigation policies were not implemented, the weak tail of banks would reach 14 percent of banking system assets, and the global capital shortfall would be $420 billion.
- A severe adverse (tail) scenario shows larger effects on bank capital, with EMs more exposed, underscoring material downside risks in tail events.

### Policy Implications and Uses
- GST provides a high-level global assessment of potential pressure points that could affect domestic banking sector stability and banks’ ability to supply credit to the real economy.
- Serves as a benchmarking tool for national authorities to compare resilience across countries and to inform capitalization policy decisions.
- Highlights the importance of continued close monitoring of banking systems, especially in EMs with more restrained macro-financial policy space.

### Limitations and Extensions
- Aggregate, publicly available data imply a simpler methodology than supervisory or FSAP-level stress tests; GST does not fully capture borrower-specific support policies extensively used during the pandemic.
- The aggregate effects of borrower-support policies should be captured in the macro-financial scenarios that feed the GST.
- The framework can be extended to incorporate direct effects of borrower-support policies; staff discuss approaches to account for such policies in the GST.

*Source: Executive Summary, "The Global Bank Stress Test" (IMF Departmental Paper).*

### 2. How Does the GST Work?

### 2. How Does the GST Work?

### Overview
- The GST provides a framework for conducting a stress test on bank capital in response to global shocks for 29 major banking systems.
- It uses a solvency stress testing methodology based on country-specific panel data and econometric and structural models to link dynamics of components of individual bank income and expense to the evolution of macro and financial variables.
- The GST projects banks’ income statement components, risk weights, and the implied capital ratios over a three-year horizon.
- Capital shortfall estimates are computed at the bank-level.

### Coverage and dataset
- Sample coverage:
  - 29 economies: 24 AEs and 5 EMs.
  - The banking sector assets of these 29 economies account for 70 percent of global banking assets.
  - In each economy, banks covering at least 80 percent of the individual banking system’s total assets are included.
  - Sample composition: 53 banks in EMs and 204 banks in AEs.
- Data sources and frequency:
  - Comparable data for bank income statements, balance sheets, and several risk metrics were sourced primarily from Fitch Connect and banks’ financial reports.
  - Historical bank dataset has an annual frequency spanning a 25-year period from 1995 to 2020.
- Cutoff date for bank data in the 2022–24 update: end-2020.

### Methodological approach and modules
- Core structure:
  - Two-part methodology: (1) econometric models for P&L and OCI components; (2) a balance sheet projection module that maps P&L, RWAs, and OCI into CET1 capital ratios and capital shortfalls over the stress horizon.
- Econometric models:
  - All models are cross-bank-country panel regression models—bank fixed effects models (FE).
  - Components modeled separately and then aggregated to net income via accounting identities.
  - P&L components modeled include:
    - (1) net loan losses supplemented with structural model elements for PDs and LGDs;
    - (2) net interest income (NII);
    - (3) net trading income (NTI);
    - (4) net fee and commission income (NFCI);
    - (5) other income/expense (residual term equal to pre-tax net income minus the four main components).
- Loan-losses and PD/LGD inference:
  - Loan-loss model coupled with a structural element decomposing loss rates into PDs and LGDs using a variant of the Frye-Jacobs (2012) methodology.
  - PDs and LGDs are inferred from projections of loan losses to derive dynamics of performing and nonperforming loan stocks.
  - Migration of nonperforming loans back to performing status (“cures”) is allowed implicitly but not explicitly modeled.
- NTI projection:
  - NTI projected using a historical standard deviation-based approach because in-sample panel regression predictive power for NTI ratios is generally low.
  - For each bank, NTI ratio = average NTI ratio over the last five years minus (scalar × standard deviation of NTI ratio over the last five years).
  - The scalar is common for all banks; in specific severe scenarios a higher scalar may be used.
- Balance sheet module and policy assumptions:
  - Bank-specific balance sheet information as of 2020 used as starting point.
  - User-defined assumptions for dividend distribution and taxes:
    - Tax rates and dividends set to zero if projected net income before taxes is negative.
    - Otherwise, tax rates = individual banks’ effective tax rates in the last year of the sample with a cap of 30 percent.
    - Dividend payout ratios = 50 percent of net income before tax if positive, and zero otherwise.
  - No deferred tax asset accumulation is considered.

### Risk-weighting and regulatory treatment
- Credit RWAs:
  - Total credit exposures split into standardized (STA) and internal rating based (IRB) approaches approximated from public disclosures.
  - STA risk-weighted densities are assumed constant over the stress horizon.
  - IRB densities adjusted using the Basel formulae (BCBS 2019).
  - Through-the-cycle PDs are estimated from scenario-dependent point-in-time PDs and a “smoothness” parameter set at 15 percent.
  - Downturn LGDs are held constant over the stress testing horizon.
- Other RWAs (market, operational, residual) assumed constant.
- Note: these assumptions likely underestimate the impact of adverse scenarios on bank capital because under stress downturn LGDs and other RWAs would be expected to increase.

### Innovations introduced by the GST
- Consistent stress testing methodology ensuring comparable analysis of main drivers of bank financial statements across countries.
- Internally and globally consistent scenarios:
  - Baseline scenarios reflect globally consistent forecasts from the WEO developed by IMF desk economists.
  - Adverse scenarios designed using the IMF Flexible System of Global Models (FSGM) to ensure internal consistency.
  - Each scenario characterized by: real GDP growth, unemployment rate, short-term interest rates, term spreads, two measures of risk premiums (VIX and corporate bond spreads), country-specific stock returns, and global oil price growth.
- Spillovers:
  - Captures cross-border spillovers and implicit spillbacks via the multilaterally consistent global models used in scenario design.
  - Cross-border spillovers also captured via impact of subsidiaries’ capital shortfalls on parent banks.
  - Inclusion/exclusion of subsidiaries depends on exercise objective (multilateral surveillance excludes subsidiaries of parents in-sample to avoid double-counting; bilateral surveillance includes subsidiaries).
- A unique global bank dataset spanning 1995–2020 sourced from public data for cross-country comparability.
- Accounting for model uncertainty via Bayesian Model Averaging (BMA):
  - BMA estimates about 5,000 models for each dependent variable and assigns weights based on predictive power.
- Inferring PDs and LGDs from loan losses due to lack of granular supervisory data, then using Bayesian/structural methods and Basel formulae to project RWAs.

### Limitations and data challenges
- Use of public data imposes constraints:
  - Less granularity and coverage than supervisory data used in FSAP stress tests or authority-conducted stress tests.
  - Methodology is simpler and more aggregated; captures high-level balance sheet dynamics only.
  - Results should be interpreted with caution relative to exercises using granular supervisory data.
- Specific challenges:
  - For a few large emerging markets, many components of banks’ published income statements appeared insensitive to macro variables; those countries were excluded from the GST.
  - Public data structure precludes analysis of borrower-support policy impacts on bank capital (e.g., cannot break down loan losses by enterprises versus households).
  - Aggregate stress tests may understate the positive effects on bank capital of targeted borrower support policies; ancillary analysis is required to assess such policies’ effects.

### Scenario design and the 2022–24 update
- Purpose of the update:
  - GST updated for 2022–24 to assess risks related to COVID-19 aftermath and emerging uncertainties.
- Scenario anchoring and design:
  - Exercise anchored on the October 2021 WEO baseline and informed by adverse scenarios developed for the April 2021 WEO using FSGM.
  - The GST adverse scenario is a linearly scaled extension of the WEO downside scenario illustrating consequences of a 2½ standard deviation shock to global growth rates, relative to the baseline, in 2022–24.
  - A second severe adverse scenario explores outcomes further into the tail of the growth distribution for each country.
- Baseline growth projections for the GST sample (exact text preserved):
  - After a contraction of almost 5 percent in 2020 and a rebound of 5½ percent in 2021, output in the group of countries included in the current GST baseline sample was projected to grow at about 42/3 percent in 2022 and 2.3 percent in 2023.
  - Growth is expected to moderate to about 2 percent over the medium term.
- Assumptions in the adverse scenario:
  - Considers adverse impacts from new variants, vaccine efficacy, and pace of vaccine rollout leading to slower activity and higher risk aversion.
  - Unconventional monetary policy measures assumed to prevent significant increases in sovereign rates.
  - No additional discretionary fiscal measures or policies are assumed; lack of conventional monetary policy space and shrinking fiscal space limit policy responses.
  - Weaker rebound leads to more pronounced scarring than in the baseline despite a gradual return in mobility to pre-pandemic levels.
- Other specified parameters and calibrations:
  - Smoothness parameter for through-the-cycle PDs set at 15 percent.
  - NTI scalar common across countries; higher scalar used in scenarios/severe years with severely constrained market volumes or agency business.
  - Dividend payout ratio set at 50 percent of net income before tax if positive; zero otherwise.
  - Tax rate cap at 30 percent.

*Source: IMF staff (excerpt from “2. How Does the GST Work?”).*

### Box 1. Designing Adverse Scenario in FSAPs

### Box 1. Designing Adverse Scenario in FSAPs

### Scenario design principles
- FSAP scenario design includes three steps: outlining risk transmission channels, choosing the severity of the shock, and mapping into full-fledged macro-financial scenarios.
- Risk transmission channels are guided by the country risk assessment matrix (RAM) underpinning the IMF’s bilateral surveillance and informed by the IMF Global Risk Assessment Matrix (G-RAM) to support multilateral consistency.
- Real GDP is the anchor variable of the scenario because a recession typically defines the worst macro-financial environment for most financial institutions.
- A full set of other macro variables consistent with the GDP shock is simulated using existing IMF dynamic stochastic general equilibrium models (such as FSGM or the Global Financial Model (GFM) (see Vitek (2018)), S-VAR type empirical models, or models used by the country authorities).

### Calibration of severity
- Severity is measured by the deviation of the GDP path compared to the baseline and should be “severe but plausible,” typically aiming at once in 20-year events.
- Usual calibration approaches include:
  - using at least two-standard-deviation as a rule of thumb; or
  - conditional shocks using growth at risk (GaR) taking into account the evolution of the output gap or the level of GDP (Adrian and others, 2020).
- IMF staff rule of thumb: shocks to GDP growth should represent a deviation from the IMF baseline projection over the first two to three years of the scenario of at least two historical standard deviations from the mean.
- The two standard deviation shock to cumulative growth in the first two years is based on the unconditional historical distribution of GDP growth.
- This rule of thumb is used in combination with GaR as a minimum severity related to cyclical vulnerabilities.

### Pandemic and cyclical adjustments to calibration
- Using standard severity approaches could result in overly benign (severe) scenarios when the economy is at the top (bottom) of economic and financial cycles.
- Some FSAPs have applied smaller/larger shocks depending on the cyclical state, targeted the size of the output gap, or set the decline of the level of GDP to those observed during benchmark crisis episodes.
- Example: the size of the pandemic shock implied that, in many countries, even the baseline scenario was more improbable than one percentile GaR estimates based on historical data; applying an additional two-standard-deviation shock to an already weak baseline could have been seen as less plausible.
- If the baseline forecast for GDP growth is significantly above the long-term average, larger shocks (for example, minimum three standard deviation shocks) are considered to ensure that the adverse scenario represents a stress event.
- More recent FSAPs aim at 2–3 percentile GaR estimates.

### Global Stress Test (GST) adverse scenario: macroeconomic shocks
- In the adverse scenario, global GDP is about 5 percentage points and 2½ percentage points lower than in the baseline in 2022 and 2023, respectively.
- A sharper tightening in financial conditions for vulnerable businesses in emerging markets and developing economies results in a larger shock for EMs.
- By 2024, the level of GDP is:
  - about 6.5 percentage points lower than in the baseline in Advanced Economies (AEs); and
  - about 9.4 percentage points lower than in the baseline in Emerging Markets (EMs).
- The larger GDP shock in EMs is reflected in more severe shocks to unemployment, credit spreads, and stock prices.

### GST results: bank capital and profitability
- Baseline scenario:
  - Banks, on average, would see a strong upward trend in their CET1 ratio, reaching 17.2 percent in 2024, 3.2 percentage points above the end-2021 level.
- Adverse scenario:
  - The average CET1 ratio would fall moderately in the first year—a drop of 1.4 percentage points below the 2021 level and 2.7 percentage points below the baseline scenario.
- The resulting decline in the CET1 ratio over the stress testing horizon stems mainly from an increase in loan loss provisions.
- In the baseline scenario, higher loan loss provision expenses contribute to a 2 percentage-point decline in the global CET1 ratio, 2.2 percentage points less than in the adverse scenario.
- For all banks, the contribution of other risk factors to the difference in CET1 ratios between baseline and adverse in 2024 ranges from:
  - 0.4 percentage points for risk-weighted assets (RWA); to
  - 0.8 percentage points for net interest income (NII).

### Heterogeneity across AEs and EMs; capital shortfalls and asset shares under stress
- The GST points to relatively lower levels of resilience of banks in EMs than in AEs.
  - AE banks increased CET1 ratios by about 0.8 percentage points in 2020.
  - In EMs, the decline in CET1 ratios in the adverse scenario reaches 2.5 percentage points below the starting point and 5.5 percentage points below the baseline in 2024.
- Drivers of larger impact on EM banks:
  - larger downside macro-financial risks in plausible downside scenarios;
  - higher estimated sensitivity of loan losses to growth shocks; and
  - somewhat less capital built over the past two years.
- Cross-country and cross-bank heterogeneity:
  - Changes in CET1 ratios (difference between 2021 and trough) reach up to 7 percentage points in some cases.
  - Capital shortfalls, calculated relative to a 4½ percent minimum capital level, range from zero to 0.6 percent of GDP in the adverse scenario across economies.
  - The share of banks by assets experiencing shortfalls is higher in emerging markets: 14 percent of total bank assets in 2024.

### Severe adverse scenario
- The severe adverse scenario is simulated as a linear extension of the scenario derived from the multilateral global model but using separate econometric estimates of relationships between GDP growth and other macro and financial variables used in the GST.
- The initial GDP shock, for each country, is calibrated by assuming GDP growth in 2022 is equal to half of the GDP growth in 2020.
- This shock corresponds to a four standard deviation shock from the baseline—which builds in rapid year-over-year growth rates as output gaps close following the extremely sharp fall in 2020—but is equivalent to a 2½ standard deviations shock from average GDP growth spanning the years preceding the pandemic.
- The severe adverse scenario points to continued downside risks to the global banking system in such a tail event and illustrates greater exposure to downside shocks, especially in emerging markets.

### Accounting for mitigating policies
- Publicly available data structure precludes full analysis of the direct impact of borrower-support policies on bank capital in the GST (for example, the data do not allow a breakdown of loan losses by corporations and households).
- Aggregate analysis could understate the positive effect on bank capital of financial policies to support household and corporate borrowers.
- Modeling the effects of support policies (for example, debt payment moratoria and guarantees) requires granular sectoral data differentiating exposures to households and corporates.
- IMF staff methodological developments:
  - An extension to the GST that decomposes aggregate loan loss provision forecasts into contributions from corporate and retail loan portfolios using direct step regression specifications (local projection method) to capture exposure shares and macro-financial drivers.
  - Two micro data-based modeling approaches to analyze implications of mitigating policies on nonfinancial corporates and households (these models can be “plugged into” the GST model suite to account for policies that affect PDs and LGDs through payment moratoria and government guarantees).
- Mechanisms:
  - Moratoria primarily affect probabilities of default (PDs).
  - Guarantees primarily affect loss given default (LGDs) but may also have secondary positive effects on PDs via lower lending rates and reduced debt service burden.

### Corporate stress testing: methodology for assessing mitigating policies (Box 3 summary)
- Tressel and Ding (2021) developed a methodology for stress-testing publicly listed nonfinancial corporates in 24 AEs and EMs to assess the impact of the COVID-19 pandemic.
- The tool performs multifactor sensitivity analysis and dynamic multiyear scenario-based stress tests of individual firms.
- Three main firm-level metrics assessed:
  - Liquidity: interest coverage ratio and cash balance.
  - Viability: interest coverage ratio and price-to-book ratio.
  - Solvency: book value of equity.
- Firm-level outcomes are aggregated at the country level and linked to bank loan and bond markets’ PDs to assess potential systemic risk and spillovers to the financial system.
- Multifactor sensitivity analysis covers:
  - real shocks and behavioral choices affecting sales and/or production costs;
  - financial shocks affecting borrowing costs, roll-over risks, and borrowing capacity;
  - policy shocks such as monetary policies, fiscal policies (for example, guarantee programs, wage subsidies, grants, tax measures), and regulatory policies (moratoria on debt payment).
- The dynamic scenario-based stress test generates scenario-dependent multiyear projections of firm-level financial health indicators and allows comparison of projections under different macroeconomic scenarios with and without policies to assess general equilibrium effects.
- The general equilibrium impact of macroeconomic policies is reflected in the path of macroeconomic variables (for example, financial conditions, real GDP path); direct impacts of micro policies at the firm level (for example, a grant, wage subsidy, or debt moratoria) are not quantified in the scenario-based stress tests described.

*Italic: Source: Box 1. Designing Adverse Scenario in FSAPs — gbstea - Box 1. Designing Adverse Scenario in FSAPs (PDF).*

### Box 4. Household Debt: Assessing Mitigating Policies Using a Micro–Macro

### Box 4. Household Debt: Assessing Mitigating Policies Using a Micro–Macro Simulation Model

### Framework and purpose
- Presents a structural micro–macro simulation model called the Integrated Dynamic Household Balance Sheet (IDHBS) model.
- Purpose: obtain macro scenario- and policy-counterfactual estimates of PDs and LGDs for the household sector where time series of household credit risk metrics are usually not publicly available.
- Rationale: a structural micro simulation-based approach enables rich policy counterfactual analysis that time series models would not allow.
- Citations in source: Gross and others (2021); Gross and Población (2017). Footnotes: 1, 2.

### Model structure (four layers) — key components and information flow
- Macro-financial scenarios
  - Adverse and policy counterfactual scenario assumptions.
  - Most important variables: unemployment, interest rates, house prices.
- Employment status simulation
  - Country-specific logistic models: relate employment status to household member characteristics.
  - Simulate individuals’ employment status to match aggregate unemployment in a scenario.
- Household member-level → Household-level
  - Household members are mapped to their households.
  - Obtain PDs for mortgages and consumer credit.
  - Obtain LGDs for mortgages.
- Household balance sheet simulation (PDs, LGDs) → Bank-level → Country-level
  - Impact through loan losses and risk-weighted assets.
  - Either for individual banks or banking system per country.
- Note: Step D (bank-level) could be linked to the GST model, where loan losses and risk weights are aligned with the outcome of the micro–macro simulation model.

### Data inputs and behavioral channels
- Uses large-scale micro databases to simulate households’ P&L and balance sheets.
- Employment status determines wage income (or unemployment benefit), ability to service debt, and thus PDs.
- Interest rate changes influence debt service burden for households with variable rate loans.
- House prices drive value of collateral and hence LGDs.
- Example household databases referenced: ECB’s Household Finance and Consumption Survey; Panel Study of Income Dynamics for the United States (see Gross and others (2021) for details). Footnote: 2.

### Uses and applications
- Employed for policy counterfactual analyses with a focus on debt moratoria (Gross and others (forthcoming)).
- Used at the European Central Bank (ECB) and by various national central banks in Europe to inform ex ante impact of borrower-based measures such as loan-to-value and debt service to income (DSTI) caps. Footnote: 1.
- Model useful for assessing:
  - Borrower-based macroprudential policies.
  - Monetary policy (for example, interest rate-based policy).
  - Fiscal policy (for example, design of unemployment benefits).
  - Policies and banks’ own behavioral responses in the pandemic.

### Key findings from policy counterfactuals (debt moratoria)
- Estimated mortgage PDs, LGDs, and loss rates suggest household sector–oriented policies can have notable mitigating effects.
- The implied counterfactual impact of household debt moratoria on banks is significant for many countries, shielding the banks from incurring material capital losses.

*Source: Box 4. Household Debt: Assessing Mitigating Policies Using a Micro–Macro Simulation Model, IMF staff (from the provided content).*

### Annex Figure 3.2. Global Bank Stress Test (GST) Macroeconomic Scenarios, Emerging Markets

### Annex Figure 3.2. Global Bank Stress Test (GST) Macroeconomic Scenarios, Emerging Markets

### Overview
- The figure presents GST macroeconomic scenarios for Emerging Markets (EMs) across five macro variables for 2021–2024.
- Two scenario paths are shown for each variable: Baseline scenario and Severe adverse scenario.
- Sources: IMF, October 2021 World Economic Outlook; Bloomberg; IMF staff estimates.

### 1. GDP Scenarios (Level, 2021=100)
- Axis tick/label values shown: 90, 100, 110, 95, 105, 115, 120.
- Time horizon: 2021, 2022, 2023, 2024 (labels shown as 2021222324).
- Scenarios: Baseline scenario; Severe adverse scenario.

### 2. Unemployment Rate Scenarios (Percent)
- Axis tick/label values shown: 0, 15, 5, 10.
- Time horizon: 2021, 2022, 2023, 2024 (labels shown as 2021222324).
- Scenarios: Baseline scenario; Severe adverse scenario.

### 3. Stock Price Scenarios (Level, 2021=100)
- Axis tick/label values shown: 0, 150, 200, 50, 100.
- Time horizon: 2021, 2022, 2023, 2024 (labels shown as 2021222324).
- Scenarios: Baseline scenario; Severe adverse scenario.

### 4. Corporate Credit Spread Scenarios (Percentage points)
- Axis tick/label values shown: 0, 5, 1, 2, 3, 4.
- Time horizon: 2021, 2022, 2023, 2024 (labels shown as 2021222324).
- Scenarios: Baseline scenario; Severe adverse scenario.

### 5. Term Premium (Percent)
- Axis tick/label values shown: 0, 6, 1, 2, 3, 4, 5.
- Time horizon: 2021, 2022, 2023, 2024 (labels shown as 2021222324).
- Scenarios: Baseline scenario; Severe adverse scenario.

*Sources: IMF, October 2021 World Economic Outlook; Bloomberg; IMF staff estimates.*

---


_Source: https://www.imf.org/-/media/files/publications/dp/2022/english/gbstea.pdf_