Chapter 29 Flashcards
(14 cards)
Define a sensitivity test
How sensitive is the model to one variable.
Define stress testing
Projection of financial condition of a company after a specific extreme adverse event over a period of time. Also deterministic. Subset of scenario analysis but of tail events. Cost the most.
Define a scenario analysis
Deterministic method of evaluating risk. Looks at fin impact of a plausible, possibly adverse, set of events, like economic recession.
How would a scenario analysis be done?
- Group risk exposures into broad categories
- Develop a plausible adverse scenario for each group
- Translate the scenario into assumptions like duration of service impacted or % of systems affected
- Then estimate the financial impact
Define a stress scenario test
Scenario analysis identifies the factors that are impacted under a chosen scenario, then a stress test is applied to these factors.
The overall stress scenario test combines the individual factor stress tests, and this is done simultaneously to allow for inter-relationships
Define reverse stress testing
Construction of a severe stress scenario that JUST allows the firm to be able to continue its business plan
Describe the two types of stress scenario tests
Localised:
Pinpoints vulnerabilities in pf by changing a few key variables
You change the volatility and correlations
It tells you where pf is fragile and effectiveness of pf diversification
Systematic/global:
What happens if the entire financial system is under stress — a market-wide meltdown. Think of it like a crisis simulation.
Assume global shock. So shock all key variables in coherent way
It tell you whether pf can survive a full blown global shock
How can operational risks be evaluated?
Difficult to quantify bc most are rare and contribute little to the aggregate risks, they are also impractical to quantify. But there are some ways to assess and allow for it:
- Scenario analysis
a. Break down operational risks into multiple categories, say 15, like fraud, money laundering, etc.
b. Then consider the cost of an adverse plausible scenario - Broad-brush analysis without any details analysis
a. Includes a factor-based approach taking part of the prems and provisions (Solvency II)
How can risks be aggregated numerically?
- Independent risks can be summed
- Correlation matrices then the capital requirement= root(sum(cisRjRi)
- Copulas
- Sum the capital requirement for each risk if they are fully independent (could be inefficient use of capital)
What are the advantages and disadvantages of using copulas?
Advantages:
1. Modelling Tail Dependence
2. Flexibility in Marginal Distributions:
o Copulas separate the modelling of marginal distributions from the dependence structure, enabling the use of different distributions for different risks (e.g., normal for one line and Pareto for another).
3. Capturing Non-Linear Relationships
4. Better Risk Aggregation Accuracy
5. Allows for partial dependency
Disadvantages:
1. Copulas require significant statistical expertise and computational power to implement and calibrate.
2. Data Intensity
3. Choice of Copula and parameterisation isn’t trivial
4. Results are less intuitive, difficult to communicate
5. Requires lots of data
What are the advantages and disadvantages of using correlation matrices?
Advantages:
1. Simplicity and Familiarity:
2. Ease of Implementation:
3. Lower Data Requirements:
4. Regulatory Acceptance:
Disadvantages :
1. Inability to Model Tail Dependence
2. Assumes Linear Dependence
3. Sensitive to outliers
4. Instability:
5. Subjective assumptions
How would you evaluate the risk of poor decision-making?
Difficult to model mathematically, and difficult to fit a full probability distribution bc of # of subjective factors involved
* Group risk exposures into broad categories like business risk
* Develop a plausible adverse scenario for each group
* Translate the scenario into assumptions like duration of service impacted or % of systems affected
* Then estimate the impact of the occurrence of the risk like extent of loss of capital
Involve senior management
How to evaluate the risk of extreme market movement?
- Use financial stress tests to investigate the impact of the movements on A and L
- Use lower equity value of a different discount rate
- Consider asset correlations and volatilities carefully bc they are often observed simultaneously
How to evaluate the risk of an option biting?
- Model it stochastically since its a financial risk that is measurable
- To reduce model complexity, only model 1 to 2 variables stochastically like investment returns
- Then model the rest of the variables deterministically
- OR model each risk stochastically, one at a time while the rest are deterministic
- Then run 1 deterministic model using all the worst-case scenarios
- This will determine the effect of interactions between the variables
