## _stfs

## Source details

**Canonical URL:** [_stfs](https://www.imf.org/-/media/websites/imf/imported-publications/external/pubs/ft/stfs/_stfs.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-publications/external/pubs/ft/stfs/_stfs.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-publications/external/pubs/ft/stfs/_stfs.pdf.json)

---

### What is a stress test — definition, scope, and value added
- Definition and purpose
  - A stress test is a way of revaluing a portfolio using a different set of assumptions to show sensitivity to a particular shock.
  - Complements traditional models by estimating portfolio value changes under exceptional but plausible changes in underlying risk factors.
- Scope of application
  - Widely used at the individual institution level to assess market risk, credit risk, and liquidity risks.
  - Extended to “system-focused” stress tests to measure sensitivity of groups of institutions or the financial system as a whole.
- Distinctions: portfolio-level vs. system-focused
  - System-focused tests aim to identify common vulnerabilities across institutions and are more macroeconomic in nature.
  - System-focused tests require aggregation and comparison across heterogeneous entities.
- Value added of system-focused stress tests
  - Integrates a forward-looking macro perspective, system-wide focus, and uniform assessment across institutions.
  - Complements institution-level stress tests, cross-checks other analyses, identifies data weaknesses, and builds supervisory expertise.
- Role relative to institution-level testing
  - Not a replacement for regular institutional stress testing; designed to complement and leverage existing expertise.
  - Applying a common scenario across institutions helps assess overall impact, distribution of shocks, contagion potential, and confidence effects.
  - If data permit, conduct tests on both aggregated portfolios and individual portfolios.

### The stress-testing process and identifying vulnerabilities
- Process overview (iterative)
  - Identify vulnerabilities → construct scenario in a consistent macroeconomic framework → perform numerical analysis → consider second-round effects → summarize and interpret results.
  - Stages are iterative and may require forming a working group of experts.
- Focus and constraints in identifying vulnerabilities
  - Narrow the exercise to main vulnerabilities; do not attempt to stress every possible risk factor.
- Indicators and information sources
  - Macroeconomic indicators: growth, utilization rates, inflationary pressures, indebtedness, leverage, debt-servicing ability.
  - Government sector: deficit, debt stock, fiscal impulse, financing of the government budget.
  - External sector: current account deficit magnitude, official reserves, financing of the deficit, size/maturity/currency composition of external debt, exchange rate misalignment and pressures.
  - Structural financial system indicators: ownership, market shares, balance sheet structures, flow-of-funds accounts.
  - Financial Soundness Indicators (FSIs): capital adequacy, asset quality, profitability, liquidity, exposure to market risks.
    - IMF developed a core set of FSIs for the banking sector.
    - An encouraged set covers additional banking FSIs and FSIs for key nonfinancial sectors.

### Constructing scenarios — model-based and alternatives
- Model-based scenarios
  - Ideally use a macroeconomic or simulation model to provide a forward-looking, internally consistent framework linking the financial system and real economy.
  - Identify key macroeconomic and financial variables that are volatile, misaligned, or likely to have greatest impact.
- Illustrative scenario (example)
  - Rapid housing price rises driven by employment growth, rising household disposable incomes, and low interest rates → mortgage-lending boom.
  - Scenario shock: rise in unemployment, fall in disposable incomes, and a sharp rise in interest rates → affects household debt-servicing capacity.
  - Macroeconomic model outputs (employment, real incomes, prices, interest rates) inform specific bank balance sheet stress tests.
- Alternatives when models are unavailable
  - Use textbook models and existing empirical research to frame an internally consistent, forward-looking macro scenario.
  - Objective: retain discipline and focus on macro–vulnerability linkages even without a full macro model.
- Historical vs. hypothetical scenarios
  - Historical scenarios: use largest observed changes or extreme values over a specified period; more intuitive because observed.
  - Hypothetical scenarios: involve large plausible movements; may be more realistic if financial structure has changed.
  - Experiences from other countries can guide scenario design.

### Balance sheet implementation — top-down and bottom-up approaches
- Two main translation approaches
  - Bottom-up approach
    - Estimates responses at portfolio level using disaggregated data from individual financial institutions.
    - Advantages: better use of portfolio data; provides sensitivity of individual institutions and risk concentrations.
    - Disadvantages: potential inconsistencies in how institutions apply scenarios and produce estimates.
  - Top-down approach
    - Uses aggregated or macro-level data to estimate group responsiveness (e.g., regressions of aggregated information on macro variables).
    - Advantages: easier to implement with aggregated data; consistent and uniform method.
    - Disadvantages: based on aggregate historical relationships that may not hold in the future.
- Institution involvement and practical considerations
  - Involve individual institutions where possible; they have best access to data and portfolio knowledge.
  - For countries with rudimentary systems, central bank or supervisory agency may need to provide guidance or undertake empirical analysis.
  - Combining top-down and bottom-up approaches is ideal, but data limitations may prevent both.
- Coverage and feasibility
  - Coverage should represent a meaningful critical mass while keeping the number of institutions feasible (e.g., fewer than 20).
  - Include banks and significant nonbank financial institutions; may include nonfinancial companies to assess corporate sector sensitivity.
- Data quality and confidentiality
  - Data availability and quality constrain stress-test scope.
  - Difficulties: lack of balance sheet exposure data, complex institutions, derivatives exposures, or missing risk measures (e.g., duration or default measures).
  - Possible remedies: work with sophisticated institutions to obtain or calibrate data; have institutions run tests under agreed assumptions and share non-confidential aggregated results.
- Scenario severity
  - Scenarios should be beyond the normal range of experience—exceptional but plausible.
  - Historical scenarios are intuitive; hypothetical scenarios can better reflect structural changes.
- Use of top-down checks
  - Top-down results provide a useful check on bottom-up results and can be used when institutions lack capacity (e.g., regression model of loan loss rates for the banking system to estimate credit-quality impact).

### Second-round effects and contagion
- Limits of the no-behavioral-response assumption
  - Assumption of no behavioral response becomes harder to justify once the time horizon of a scenario extends beyond a year or more.
  - For systemically important institutions or for systems as a whole, the assumption of no feedback effects may be an oversimplification.
  - Policy environment may change over a longer horizon as monetary or supervisory authorities react to shocks.
  - Financial sector safety nets may provide a buffer over longer horizons.
- Typical second-round effects
  - Direct credit losses from counterparty failures.
  - Increased funding costs for weakened banks.
  - Portfolio adjustments (such as a tightening of lending criteria) affecting aggregate demand.
- Contagion modeling approach
  - Typical two-stage exercise: stress test individual balance sheets and income statements; then examine counterparty exposures to the institutions made most vulnerable (interbank loans, cross-shareholdings, deposits, or other exposures).

### Interpretation, presentation, and limitations of stress tests
- Uses and caveats
  - Stress tests are useful for identifying latent risk exposures in a systematic and intuitive manner.
  - Regularly conducted stress tests provide information about changes in the risk profile of the system over time.
  - Care should be taken not to portray stress-test results as providing a precise measure of the magnitude of losses.
- Limitations
  - Unlikely to capture the full range and interaction of risk exposures (such as operational risk and legal risk).
  - May give only a partial picture of the true nature of risk-taking by participating institutions.
  - Typically consider only part of a bank’s income-generating operations; banks may have significant income flows unaffected by the specific stress test scenarios analyzed.
- Complementary analyses and validation
  - Compare stress-test results with other measures of risk exposure, such as financial soundness indicators.
  - Clear presentation of output facilitates analysis and discussion.
  - Bottom-up approaches benefit from descriptive statistics (mean, median, standard deviation, minimum, maximum, and number of institutions in each decile) and peer-group analysis to show distribution of aggregate impacts across institutions.
  - Supervisory agency-conducted stress tests on reported balance sheet data using a common framework can act as an accuracy and consistency check on institution-provided results.

### Publication of results — confidentiality and disclosure
- Challenges to public dissemination
  - Participating institutions may resist disclosure that could identify specific firms for fear of negative market interpretation or competitive disadvantage.
  - Analysts may misinterpret chosen scenarios as reflecting an official view of the most likely or most problematic scenario.
- Experience and practical disclosures
  - Experience shows difficulties can usually be overcome; many countries publish summary or aggregated information.
  - Useful disclosures without revealing identities:
    - Summary information such as the mean and the range.
    - Disclosure of scenarios undertaken to raise awareness of risks and inform institution-level stress-testing programs.

### FSAP experience, capacity building, and recent trends
- Role in FSAPs
  - Stress tests are integral to Financial Sector Assessment Programs (FSAPs) conducted by the IMF and World Bank, providing quantitative measures of financial system vulnerability to different shocks.
  - Stress tests complement qualitative assessments of legal, institutional, regulatory, and supervisory frameworks and empirical analysis of financial soundness indicators.
- Practical constraints in FSAPs
  - Data availability is key in determining approach and sophistication of FSAP stress tests.
  - Most FSAP analyses are performed on a bank-by-bank basis, based on single factor and scenario approaches.
  - Contagion risks and second-round effects have typically not been addressed in many FSAPs, although some incorporate elements of interbank contagion.
  - Involvement of authorities varies according to expertise and ability or willingness to provide data; some country authorities are precluded from providing data on individual institutions by bank secrecy laws or conventions.
- Trends in FSAP practice
  - Greater role for country authorities and individual financial institutions in design and implementation.
  - Increased reliance on internal models of banks to evaluate shock impacts.
  - Increased use of macroeconomic models to calibrate macro scenarios.
  - Greater coverage of nonbank financial institutions.
- IMF capacity-building response
  - IMF offers technical assistance and cooperation on implementation of stress-testing programs to central banks and supervisors.
  - Assistance includes bilateral support and cooperation through workshops and seminars.

### Concluding remarks — process summary and best practices
- Stress testing is best seen as a process of identifying vulnerabilities and estimating sensitivity of balance sheets to a variety of shocks.
- System-focused stress tests link a forward-looking macroeconomic perspective with an assessment of sensitivity of a group of institutions to major changes in the economic and financial environment.
- Recommended process steps
  - Identify specific vulnerabilities or areas of concern.
  - Construct scenarios within a consistent macroeconomic framework.
  - Iteratively isolate key vulnerabilities using qualitative and quantitative elements.
  - Use a range of numerical indicators, including macro-level indicators, broad structural indicators, and micro-level indicators.
  - Ideally, an econometric or simulation model should form the basis of scenarios.
  - A working group of selected experts may facilitate the process.
- Translating scenarios into institutions’ balance sheets and income statements
  - Bottom-up approach: impact estimated using data on individual portfolios.
  - Top-down approach: impact estimated using aggregated data.
- Public dissemination challenges are surmountable as evidenced by wide publication by countries.
- FSAP experience demonstrates stress tests provide quantitative measures of vulnerability useful alongside other analyses to assess overall stability of a financial system.
- Recent trends: greater integration of macro perspective, more involvement by country authorities and institutions, and greater coverage of the financial sector.

*Paul Hilbers and Matthew T. Jones*

### Section 1

### _stfs - Section 1

### Preface and publication details
- ©2004 International Monetary Fund
- Editor: Sean M. Culhane, IMF External Relations Department
- Cover design and composition: Laura J. Hurst; Grammarians, Inc.
- ISBN 1-58906-402-X
- Published September 2004
- Pamphlet based on IMF Working Paper WP/04/127, “Stress Testing Financial Systems: What to Do When the Governor Calls,” by Matthew T. Jones, Paul Hilbers, and Graham Slack, July 2004.
- IMF Publication Services price for the original paper: US$15.00

### What Is a Stress Test?
- Definition and purpose
  - A stress test is a way of revaluing a portfolio using a different set of assumptions to show sensitivity to a particular shock.
  - Complements traditional models by estimating portfolio value changes under exceptional but plausible changes in underlying risk factors.
- Scope of application
  - Widely used at the individual institution level to assess market risk, credit risk, and liquidity risks.
  - Extended to “system-focused” stress tests to measure sensitivity of groups of institutions or the financial system as a whole.
- Distinctions between portfolio-level and system-focused stress tests
  - System-focused tests aim to identify common vulnerabilities across institutions and are more macroeconomic in nature.
  - System-focused tests often involve greater aggregation and heterogeneity, requiring comparison or addition of different entities (“apples” and “oranges”).
- Value added of system-focused stress tests
  - Integrates forward-looking macro perspective, system-wide focus, and uniform assessment across institutions.
  - Complements institution-level stress tests, cross-checks other analyses, identifies weaknesses in data collection/reporting, and builds supervisory and institutional expertise.
- Role relative to institution-level stress testing
  - Not a replacement for regular institutional stress testing; designed to complement and leverage existing expertise.
  - Applying a common scenario to multiple institutions helps assess overall impact and distribution of shocks, including potential for contagion and confidence effects.
  - If data permit, stress tests should be conducted on both aggregated portfolios and individual portfolios.

### The Process of Stress Testing a System
- Overview
  - Best conceived as a process: identify vulnerabilities → construct scenario in a consistent macroeconomic framework → perform numerical analysis → consider second-round effects → summarize and interpret results.
  - Stages are iterative and may require forming a working group to draw on diverse expertise.

### Identifying Vulnerabilities
- Focus and constraints
  - Narrow the exercise to main vulnerabilities; do not attempt to stress every possible risk factor.
- Indicators and information sources
  - Macroeconomic indicators: growth, utilization rates, inflationary pressures, indebtedness, leverage, debt-servicing ability.
  - Government sector: deficit, debt stock, fiscal impulse, financing of the government budget.
  - External sector: current account deficit magnitude, official reserves, financing of the deficit, size/maturity/currency composition of external debt, exchange rate misalignment and pressures.
  - Structural financial system indicators: ownership, market shares, balance sheet structures, flow-of-funds accounts.
  - Financial Soundness Indicators (FSIs): capital adequacy, asset quality, profitability, liquidity, exposure to market risks.
    - IMF developed a core set of FSIs for the banking sector.
    - An encouraged set covers additional banking FSIs and FSIs for key nonfinancial sectors.

### Constructing Scenarios
- Model-based scenarios
  - Ideally use a macroeconomic or simulation model to provide a forward-looking, internally consistent framework linking the financial system and real economy.
  - Identify key macroeconomic and financial variables that are volatile, misaligned, or likely to have greatest impact.
- Example scenario (illustrative)
  - Rapid housing price rises driven by employment growth, rising household disposable incomes, and low interest rates → mortgage-lending boom.
  - Scenario shock: rise in unemployment, fall in disposable incomes, and a sharp rise in interest rates → affects household debt-servicing capacity.
  - Macroeconomic model outputs (employment, real incomes, prices, interest rates) inform specific bank balance sheet stress tests.
- Alternatives when models are unavailable
  - Use textbook models and existing empirical research to frame an internally consistent, forward-looking macro scenario.
  - Objective: retain discipline and focus on macro–vulnerability linkages even without a full macro model.
- Historical vs. hypothetical scenarios
  - Historical scenarios: use largest observed changes or extreme values over a specified period; more intuitive because observed.
  - Hypothetical scenarios: involve large plausible movements; may be more realistic if financial structure has changed.
  - Experiences from other countries can guide scenario design.

### Balance Sheet Implementation
- Two main translation approaches
  - Bottom-up approach
    - Estimates responses at portfolio level using disaggregated data from individual financial institutions.
    - Advantages: better use of portfolio data; provides sensitivity of individual institutions and risk concentrations.
    - Disadvantages: potential inconsistencies in how institutions apply scenarios and produce estimates.
  - Top-down approach
    - Uses aggregated or macro-level data to estimate group responsiveness (e.g., regressions of aggregated information on macro variables).
    - Advantages: easier to implement with aggregated data; consistent and uniform method.
    - Disadvantages: based on aggregate historical relationships that may not hold in the future.
- Institution involvement and practical considerations
  - Involve individual institutions where possible; they have best access to data and portfolio knowledge.
  - For countries with rudimentary systems, central bank or supervisory agency may need to provide guidance or undertake empirical analysis.
  - Combining top-down and bottom-up approaches is ideal, but data limitations may prevent both.
- Coverage and feasibility
  - Coverage should represent a meaningful critical mass while keeping the number of institutions feasible (e.g., fewer than 20).
  - Include banks and significant nonbank financial institutions; may include nonfinancial companies to assess corporate sector sensitivity.
- Data quality and confidentiality
  - Data availability and quality constrain stress-test scope.
  - Difficulties: lack of balance sheet exposure data, complex institutions, derivatives exposures, or missing risk measures (e.g., duration or default measures).
  - Possible remedies: work with sophisticated institutions to obtain or calibrate data; have institutions run tests under agreed assumptions and share non-confidential aggregated results.
- Scenario severity
  - Scenarios should be beyond the normal range of experience—exceptional but plausible.
  - Historical scenarios are intuitive; hypothetical scenarios can better reflect structural changes.
- Use of top-down checks
  - Top-down results provide a useful check on bottom-up results and can be used when institutions lack capacity (e.g., regression model of loan loss rates for the banking system to estimate credit-quality impact).

### Second-Round Effects
- Typical assumptions and limitations
  - Most stress tests assume no change in portfolio behavior or structure in response to the shock.
  - Tests are applied to a balance sheet at a point in time or with a forecast over a specific horizon, valuing shocks at market prices.
- Justification for limited behavioral response
  - Limited behavioral-response assumption may be valid for short horizons or where portfolio restructuring is slow or would incur “fire-sale” losses.
  - Example time horizon: one- to three-month horizon may justify assuming limited restructuring for large loan portfolios.
  - Assumption may also be justifiable for individual institutions without large systemic impact.

*Source: _stfs - Section 1*

### Section 2

### _stfs - Section 2

### Second-round effects and contagion
- Assumption of no behavioral response becomes harder to justify once the time horizon of a scenario extends beyond a year or more.
- For systemically important institutions or for systems as a whole, the assumption of no feedback effects implicit in many stress tests may be an oversimplification.
- Policy environment may change over a longer horizon as monetary or supervisory authorities react to shocks.
- Financial sector safety nets may provide a buffer over longer horizons.
- Second-round effects materialize as:
  - direct credit losses from counterparty failures,
  - increased funding costs for weakened banks,
  - portfolio adjustments (such as a tightening of lending criteria) affecting aggregate demand.
- Contagion models are often used to consider second-round effects and linkages between institutions:
  - Typical two-stage exercise: stress test of individual balance sheets and income statements; then examine counterparty exposures to the institutions made most vulnerable (interbank loans, cross-shareholdings, deposits, or other exposures).

### Interpretation of stress tests
- Stress tests are useful for identifying latent risk exposures in a systematic and intuitive manner.
- Regularly conducted stress tests provide information about changes in the risk profile of the system over time.
- Care should be taken not to portray stress-test results as providing a precise measure of the magnitude of losses.
- Limitations of stress tests:
  - Unlikely to capture the full range and interaction of risk exposures (such as operational risk and legal risk).
  - May give only a partial picture of the true nature of risk-taking by participating institutions.
  - Typically consider only part of a bank’s income-generating operations; banks may have significant income flows unaffected by the specific stress test scenarios analyzed.
- Complementary analyses:
  - Compare stress-test results with other measures of risk exposure, such as financial soundness indicators.
- Presentation and validation:
  - Clear presentation of output facilitates analysis and discussion.
  - Bottom-up approaches benefit from descriptive statistics (mean, median, standard deviation, minimum, maximum, and number of institutions in each decile) and peer-group analysis to show distribution of aggregate impacts across institutions.
  - Supervisory agency-conducted stress tests on reported balance sheet data using a common framework can act as an accuracy and consistency check on institution-provided results.

### Publication of results
- Public dissemination presents challenges to confidentiality and interpretation:
  - Participating institutions may resist disclosure that could identify specific firms for fear of negative market interpretation or competitive disadvantage.
  - Analysts may misinterpret chosen scenarios as reflecting an official view of the most likely or most problematic scenario.
- Experience shows difficulties can usually be overcome; many countries publish summary or aggregated information.
- Useful disclosures without revealing identities:
  - Summary information such as the mean and the range.
  - Disclosure of scenarios undertaken to raise awareness of risks and inform institution-level stress-testing programs.

### FSAP experience and capacity building
- Stress tests are integral to Financial Sector Assessment Programs (FSAPs) conducted by the IMF and World Bank, providing quantitative measures of financial system vulnerability to different shocks.
- Stress tests complement qualitative assessments of legal, institutional, regulatory, and supervisory frameworks and empirical analysis of financial soundness indicators.
- Data availability is key in determining approach and sophistication of FSAP stress tests.
- Most FSAP analyses are performed on a bank-by-bank basis, based on single factor and scenario approaches.
- Contagion risks and second-round effects have typically not been addressed in many FSAPs, although some incorporate elements of interbank contagion.
- Involvement of authorities varies according to expertise and ability or willingness to provide data; some country authorities are precluded from providing data on individual institutions by bank secrecy laws or conventions.
- Among countries that have published the IMF’s summary assessment of the FSAP mission (the Financial System Stability Assessment or FSSA), most have included a summary of stress-testing results.
- Trends in FSAP practice:
  - Greater role for country authorities and individual financial institutions in design and implementation.
  - Increased reliance on internal models of banks to evaluate shock impacts.
  - Increased use of macroeconomic models to calibrate macro scenarios.
  - Greater coverage of nonbank financial institutions.
- IMF capacity-building response:
  - IMF offers technical assistance and cooperation on implementation of stress-testing programs to central banks and supervisors.
  - Assistance includes bilateral support and cooperation through workshops and seminars.

### Concluding remarks
- Stress testing is best seen as a process of identifying vulnerabilities and estimating sensitivity of balance sheets to a variety of shocks.
- System-focused stress tests link a forward-looking macroeconomic perspective with an assessment of sensitivity of a group of institutions to major changes in the economic and financial environment.
- Process steps:
  - Identify specific vulnerabilities or areas of concern.
  - Construct scenarios within a consistent macroeconomic framework.
  - Isolating key vulnerabilities is iterative, involving qualitative and quantitative elements.
  - Use a range of numerical indicators, including macro-level indicators, broad structural indicators, and micro-level indicators.
  - Ideally, an econometric or simulation model should form the basis of scenarios.
  - A working group of selected experts may facilitate the process.
- Translating scenarios into institutions’ balance sheets and income statements:
  - Bottom-up approach: impact estimated using data on individual portfolios.
  - Top-down approach: impact estimated using aggregated data.
- Public dissemination challenges are surmountable as evidenced by wide publication by countries.
- FSAP experience demonstrates stress tests provide quantitative measures of vulnerability useful with other analyses to assess overall stability of a financial system.
- Recent trends: greater integration of macro perspective, more involvement by country authorities and institutions, and greater coverage of the financial sector.

_ Paul Hilbers and Matthew T. Jones_

---


_Source: https://www.imf.org/-/media/websites/imf/imported-publications/external/pubs/ft/stfs/_stfs.pdf_