After completing this reading, you should be able to:
- Describe the rationale for the use of stress testing as a risk management tool.
- Describe the weaknesses identified and recommendations for improvement in:
- The use of stress testing and integration in risk governance
- Stress testing methodologies
- Stress testing scenarios
- Stress testing handling of specific risks and products
- Describe stress testing principles for banks regarding the use of stress testing and integration in risk governance, stress testing methodology and scenario selection, and principles for supervisors.
Why Use Stress Testing As A Risk Management Tool?
- Stress testing serves to warn a firm’s management of potential adverse events arising from the firm’s risk exposure and goes further to give estimates of the amount of capital needed to absorb losses that may result from such events.
- Stress tests help to avoid any form of complacency that may creep in after an extended period of stability and profitability. It serves to remind management that losses could still occur, and adequate plans have to be put in place in readiness for every eventuality. This way, a firm is able to avoid issues like underpricing of products, something that could prove financially fatal.
- Stress testing is a key risk management tool during periods of expansion when a firm introduces new products into the market. For such products, there may be very limited loss data or none at all, and hypothetical stress testing helps to come up with reliable loss estimates.
- Under pillar 1 of Basel II, stress testing is a requirement for all banks using the Internal Models Approach (IMA) to model market risk and the internal ratings-based approach to model credit risk. These banks have to employ stress testing to determine the level of capital they are required to have.
- Stress testing supplements other risk management tools, helping banks to mitigate risks through measures such as hedging and insurance. By itself, stress testing cannot address all risk management weaknesses, nor can it provide a one-stop solution.
Weaknesses and Recommendations for Improvement in Stress Testing and Governance
The following weaknesses and recommendations are based on recent financial crises, especially the 2007/2009 financial crisis.
- Lack of involvement of the board and senior management: There’s need for the board and senior management to take a leading role in stress testing programs. As revealed during post analysis of the 2007/2009 financial crisis, firms whose management (and the board)was well integrated into stress testing were able to deal with financial trouble relatively well. Those that viewed stress testing as a non-core risk management tool with little need for management input were hit harder.
- Inadequate response to crises: Although most firms did try to employ stress testing methods as the 2007/2009 crisis developed, the measures taken were not flexible or effective enough to bring about a swift and comprehensive response to the changing conditions.
- A lack of overall organizational view: In the years leading up to the recent crisis, stress testing was silo-based. Business lines were expected to conduct stress tests on their exposures, which means that there was very little assessment of enterprise-wide exposure and the correlation between various risks. Stress testing was largely viewed as a routine procedure with little significance.
- Undeveloped stress testing programs: Although stress testing for market risk has been developed extensively over the years, stress tests for other types of risk – including credit risk and operational risk – are relatively nascent.
- Making stress testing a critical part of governance: The board of directors together with senior management should take an active role in stress testing. The ultimate responsibility for stress testing rests with the board. As such, the boar should be fully engaged in stress testing programs and foster internal debate about all key issues, including stress testing methodologies, scenario analysis, and assumptions made.
- Stress testing should be a comprehensive, enterprise-wide program: Stress testing should be undertaken, not just at the business line level, but also at the level of the firm as a whole. This way, it’s possible to identify the interrelationships between various business lines and the correlations between various risk exposures. Stress tests should complement other risk assessment tools such as the VaR.
- Stress testing should be an all-conclusive, well-structured program: Stress tests should be well phased with clear steps such as risk identification, application of modeling techniques, implementation, and use of results. At each of these phases, there should be extensive consultations, brainstorming, and involvement of senior experts within the bank.
- Good documentation: Procedures and policies should be well documented such that a deviation from the norm is quickly identified and dealt with or accounted for. Other issues that need good documentation include assumptions, test methodologies, and underlying assumptions.
- Sound infrastructure: Behind every stress testing program, there should be an adequate, flexible infrastructure for collecting and recording data. This should make risk aggregation and generation of new scenarios efficient.
- Regular review: All qualitative and quantitative aspects of the test testing program should be regularly tested and screened against laid down policies and procedures. In addition, the underlying assumptions should regularly be tested to determine their suitability in the face of changing market conditions.
Weaknesses and Recommendations for Improvement in Stress Testing Methodologies
- Inadequate infrastructure: Banks lacked robust infrastructures that would have helped them to collect data and identify risk exposures in time.
- Limited recognition of interactive effects: The stress testing models used in the recent financial crisis were not equipped to predict the interrelationships among various financial securities. For example, initial mortgage defaults may have been modeled correctly, but there had been very few studies on how such defaults would affect mortgage-backed securities (securitized products).
- Inadequate risk assessment approaches: The assumption that risk is a function of known, non-stochastic processes was exposed as a major fallacy. Risk assessment approaches built upon such an assumption were found to be inadequate.
- Inadequate firm-wide perspective: In the years leading up to the recent crisis, stress testing was silo-based. Business lines were expected to conduct stress tests on their exposures, which means that there was very little assessment of enterprise-wide exposure and the correlation between various risks. For example, if there had been adequate firm-wide analysis, the effect of mortgage defaults on securitized assets would have been modeled in a much better way.
- There’s need for comprehensive, firm-wide stress testing: Stress testing should be a comprehensive endeavor that involves everyone in an organization, from the board all the way down. There should be efforts to establish the spillover effects of stress events.
- Multiple measures: Firms should make efforts to establish how stress events would affect multiple measures. These include asset/liability values, net income, and capital requirements.
- Risk concentrations: A bank may subdivide risks according to aspects such as region, industry or name, and then conduct stress tests that reveal the loss exposure in each concentration.
Weaknesses and Recommendations for Improvement in Stress Testing Scenarios
- Stress tests were not sufficiently severe: In the years leading up to the recent financial crisis, stress test scenarios lacked depth and width, so that they didn’t assess the impact of severe, totally unprecedented and financial crippling events. Banks would conduct mild tests of short duration. There were no attempts to establish the spillover effects of various shock events.
- A lack of forward-looking scenarios: Most of the scenarios upon which stress tests were based only took into account backward-looking, historical events. Shock events that had not been experienced in history despite having positive probabilities of occurrence would be ruled out. It was wrongly assumed that such events would never occur.
- An overreliance on sensitivity analysis: Prior to the recent crisis, banks used to rely on sensitivity analysis to generate scenarios. A critical assumption of sensitivity analysis is that it focuses on the impact of a shock on a single factor while holding other factors constant. By so doing, it fails to take into consideration the spillover effect and feedback effects arising from correlations arising from various risk factors.
- A combination of historical and forward-looking scenarios: Scenarios should not be based purely on historical events. Instead, they should be based on potential future events, emerging risks, and the new products getting introduced to the market.
- Time horizon: Stress scenarios should encompass a range of time horizons, not just the traditional 1-year maximum. As shown by the recent crisis, conditions can deteriorate quickly,and recessions can continue for longer than initially anticipated.
- Stress scenarios should be in-depth and severe: Stress scenarios should consist of both mild events and sufficiently severe events that have the potential to threaten a firm’s very existence.
- Reverse stress testing: Under reverse stress testing, a firm should start from a predetermined outcome, followed by an identification of events that could lead to the outcome. Finally, the firm should evaluate the effectiveness of its hedging/risk mitigation measures to deal with the outcome. A good example would be panic withdrawals in a bank.
Weaknesses and Recommendations for Improvement in Stress Test Handling
- Risks arising from complex, structured products: Prior to the recent financial crisis, the risks attached to complex structured products were not adequately understood. Structured products such as mortgage-backed securities and collateralized debt obligations rely heavily on the underlying pool of securities. Stress testing of structured products largely avoided severe scenarios, resulting in a gross underestimation of potential losses.Some of the recommendations put forth to improve stress tests conducted on structured products include:
- The stress test should take into account all relevant information about the underlying asset(s). For instance, if the underlying asset is a portfolio of corporate bonds, the stress test should look into the management, creditworthiness, financial statements, as well as contingent funding needs of the corporate entities involved.
- The stress test should consider the subordination level of the product. Tranches that receive payment only after other tranches have been paid – e.g., the equity tranche – should carry more risk.
- The test should take into consideration the contractual obligations of the underlying. For example, the principal amount of a loan may fall due immediately a scheduled interest payment is missed.
- Basis risk: Basis is the difference between prices or interest rates between the cash market and the futures markets. The possibility of a change in basis between the time a futures contract is opened and the time it’s closed is called basis risk. This risk can reduce the effectiveness of a hedging position. In the lead up to the recent financial crisis, basis risk was largely ignored by banks. That means some of their hedged positions were actually not hedged at all. At best, some positions were only partially hedged.
- Counterparty credit risk: In the recent financial crisis, entities expected to provide protection sustained severe losses, leaving protection buyers badly exposed.
- Pipeline risk: Pipeline risk refers to the risk that a bank may fail to complete the securitization process as originally planned – possibly due to a lack of access to the securitization market – leaving it with no choice but to “warehouse” the underlying assets. In the lead up to the recent financial crisis, banks wrongly assumed that warehouse risk was close to zero, implying that the securitization market would remain liquid and easy to operate in throughout.
- Funding liquidity risk: In the lead up to the recent crisis, the inability of banks to meet their short-term day-to-day funding needs was assumed to be minimal.
- Banks should improve stress test handling by including basis risk in their assessment of overall risk.
- Banks should include the potential risks arising from the financial conditions of counterparties. Risk cannot entirely be eliminated by taking hedging positions.
- Pipeline risk should be included in a bank’s scenario analysis tests.
- Funding liquidity risk should form an integral part of stress test going forward.
Recommendations to Supervisors
To improve stress testing programs, supervisors have a starring role. They should:
- Frequently assess a bank’s stress testing procedures. That includes testing compliance with policies and other documented rules. They are also expected to share their views on the direction of financial markets and advise the bank on the best way to prepare for all possible eventualities.
- Take corrective action: Whenever the supervisor notices a material departure from laid down policies and procedures in the stress testing program, they are obliged to speak up and initiate immediate corrective action.
- Conduct additional stress tests: Such additional stress tests should be performed within a bank’s jurisdiction and should complement the existing stress scenarios.
- Consult widelySupervisors should complement their own knowledge with that of other industry experts. That way, they can easily identify gaps in knowledge or other vulnerabilities in the stress testing program of the banks under their supervision.
- Evaluate capital and liquidity needs: As outlined in Basel II, a bank should take into account both its capital and liquidity needs while conducting stress tests. It’s the responsibility of the supervisor to evaluate whether these conditions have been met.
One of the recommendations put forth so as to improve stress testing scenarios is that:
- Stress tests should encompass a maximum time horizon of one year so as to take into account emerging issues and the risk posed by newly introduced products
- Test scenarios should always be supported by similar events in history to safeguard relevance
- Banks should employ reverse testing to understand their risk exposures better and test the effectiveness of existing risk mitigation strategies
- Banks should only test extremely severe scenarios that have the potential to threaten the bank as a going concern. Other risk management measures like the VaR should be used for mild events
The correct answer is C.
Under reverse stress testing, a firm should start from a predetermined outcome, followed by an identification of events that could lead to the outcome. Finally, the firm should evaluate the effectiveness of its hedging/risk mitigation measures to deal with the outcome.
A is incorrect. Stress scenarios should encompass a range of time horizons, not just the traditional 1-year maximum.
B is incorrect. Scenarios should not be based purely on historical events. Instead, they should be based on potential (hypothetical) future events, emerging risks, and the new products getting introduced to the market.
D is incorrect. Stress scenarios should consist of both mild events and sufficiently severe events that have the potential to threaten a firm’s very existence.
Which of the following actions are supervisors required to perform so as to improve the stress testing program of a bank?
- Only consult internal sources to preserve the independence of the firm
- Additional stress tests to complement the existing tests
- Take corrective action when there are material departures from laid down policies/procedures
- II and III
Supervisors should complement their own knowledge with that of other industry experts. That way, they can easily identify gaps in knowledge or other vulnerabilities in the stress testing program of the banks under their supervision. Such consultations are not limited in scope (both external and internal sources are acceptable).