4.7 Stress Testing

4.7 Stress Testing

Question 1

PE2018Q41 / PE2019Q41
An implementation principle recommended by the Basel Committee to banks for the governance of sound stress testing practices is that stress testing reports should:

A. Not be passed up to senior management without first being approved by middle management.
B. Have limited input from their respective business areas to prevent biasing of the results.
C. Challenge prior assumptions to help foster debate among decision makers.
D. Be separated by business lines to help identify risk concentrations.

Answer: C
Learning Objective: Describe weaknesses identified and recommendations for improvement in: The use of stress testing and integration in risk governance.

The Basel Committee states “At banks that were highly exposed to the financial crisis and fared comparatively well, senior management as a whole took an active interest in the development and operation of stress testing… stress testing at most banks, however, did not foster internal debate nor challenge prior assumptions…” Therefore, the Basel Committee recommends that prior assumptions used in stress testing be challenged to ensure that the stress test best captures the potential for extreme scenarios given current market conditions.

---

Question 2

PE2019Q50
A manager at an asset management firm relies on a $\text{VaR}$-based risk measurement system that calculates VaR for each of the firm’s portfolios as well as an aggregate firm-wide $\text{VaR}$.The CRO proposes implementation of a stress testing approach to supplement the $\text{VaR}$ system. Which of the following statements best supports the CRO’s proposal?

A. In practice, stress tests utilize a great number of scenarios while $\text{VaR}$ measures rely on just a few scenarios to create their loss estimates.
B. Stress testing makes it possible to capture dependencies between asset classes in specific scenarios that cannot be captured well through a $\text{VaR}$-based system.
C. Stress testing is more accurate than a $\text{VaR}$-based system in predicting the probability of losses at a point in time.
D.While stress testing is similar to $\text{VaR}$, it is restricted to using only distributions of macroeconomic variables to generate its predictions.

Answer: B
Learning Objective: Describe the various approaches to using VaR models in stress tests.

The main purpose of value-at-risk ($\text{VaR}$) measures is to quantify potential losses under “normal” market conditions, where normal is defined by the confidence level, typically $99$ percent. In principle, increasing the confidence level could uncover progressively larger but less likely losses. In practice, $\text{VaR}$ measures based on recent historical data can fail to identify extreme unusual situations that could cause severe losses. This is why $\text{VaR}$ methods should be supplemented by a regular program of stress testing. Stress testing is a non-statistical risk measure because it is not associated with a probability statement like $\text{VaR}$.

One other reason to stress test is that $\text{VaR}$ measures typically use recent historical data. Stress testing, in contrast, considers situations that are absent from historical data or not well represented but nonetheless likely. Alternatively, stress tests are useful to identify states of the world where historical relationships break down, either temporarily or permanently.

A is incorrect. $\text{VaR}$ utilizes a great number of scenarios while stress testing focuses on just

C is incorrect. This is a description of $\text{VaR}$.

D is incorrect. Stress testing may employ scenarios that are not generated by distributions and probabilities in general do not play a prominent role.

---

Question 3

PE2022Q39
A team of risk analysts is conducting an independent review of a recent stress test. As part of this process, the team evaluates the stress testing models used as well as the supporting assumptions in the models. Which of the following, if found in the review, should the analysts identify as the greatest deficiency relating to the models or their assumptions?

A. Relationships between core and peripheral variables in the model are found by regressing their past behavior during stressed market conditions only.
B. Credit risk losses are modeled by mapping historical default rate data provided by credit rating agencies to estimates of gross domestic product.
C. The stress test does not account for the likely responses of other financial institutions to changes in the core variables of the test.
D. A scenario used in the stress test was developed using reverse stress testing.

Answer: C
Learning Objective: Explain key considerations and challenges related to stress testing, including choice of scenarios, regulatory specifications, model building, and reverse stress testing.

C is correct. In addition to a scenario’s immediate impacts, analysts should consider its knock-on effects, or the impacts of how firms (particularly other financial institutions) respond to the scenario.

A is incorrect. When using regression to determine how peripheral variables respond to changes in a scenario’s core variables, it is important to recognize that the focus is on the relationship between variables in stressed market conditions rather than normal market conditions.

B is incorrect. This is an appropriate method for modeling credit risk losses in a stress test.

D is incorrect. Reverse stress testing can be an input to the work of a stress testing committee in selecting scenarios for investigation.

---