Chapter 29: Risk measurement and reporting Flashcards
+
features of a risk that would make it appropriate to model stochastically as opposed to deterministically
- has a high score (high frequence or severity) and there is a high prioristy to assess carefully
- has a high variability of possible outcomes
- has a lot of experience data on which to base probability distributions
- relates to financial guarantees or options
- involves a mismatching of assets and liabilities
‘single event’ insurance
one where there can only be a maximum of one claim
ways to allow for operational risk within an organisation
- a broad brush approach that does not perform any detailed analysis
- scenario analysis
why would we go for scenario analysis instead of stochastic modelling?
- risks might not be suitable for mathematical modelling
- Distribution would need so many subjective parameters so that value of using it would be eroded.
Steps in scenario analysis
- Grouping risks into broad categories - all risks involving fraud, all risks involving systsems - input from management may be needed
- development of a plausible adverse scenario - so that it is possible to quatify the impact of the risk, the scenario must be representative of all risks in that group.
- calculations of the consequences of the risk event occuring for each scenario - include redress paid to those affected, fines correcting systems and records and opportunity costs
- total costs calculated are taken as the financial cost of all the risks represented by the chosen scenario
What is the limitation to scenario testing ?
it can only quantify the severity of a scenario and not the probability
The limitation to scenario testing is that it can only quantify the severity of a scenario and not the probability
What are some of the ways to go around this limitation
- The organistation may use their capital model to determine the probability of an equivalent scenairio occuring
- they may have an idea of the probability of the scenario occuring and use this in conjuction to the severity to help caliberate or validate the capital model
stress scenario test
scenario testing + stress testing
two types of scenario stress test
- to identify the “weak areas” in the portfolio and investigate the effects of localised stress situations by looking at the effect of different combinations of correlations and volatilities
- to gauge the impact of maor market turmoil affecting all model parameters, while ensuring consistency between correlations while they are “stressed”
*
reverse stress testing
- construction of a stress scenario that just allows the firm to be able to continue to meet its business plan
- identifying a scenario which would just be enough to stop the company fulfilling its strategic plan
Stochastic modelling as an extension of stress testing
- full stochastic model with all the variables that give rise to risk being incorporated as probability distributions
- and a full set of dynamic interactions between the variables specified
- the model can then determine the capital necessary to just avoid ruin at any probability level
ways to limit the complexity of a stochatic model
- restrict the duration (or time horizon) of the model
- limit the number of variables modelled stochastically and use a deterministic approach for the other variables - variables that only have an adverse effect when they move in one direction can be modelled stochastically e.g mortality and longevity
- carry out a number of runs with different single stochastic variable and then a single deterministc run using all the worst case scenarios together
limitations to stochastic modelling
- difficult to specify and build
- the run times of multiple variables become impractical
Explain why the effect of multiple risks may be less than the sum of individual risks
- Impact of diversification or less than perfect (even negative) correlation
- less than perfect correlation means that they are unlikely to occur at the same time
Different ways of aggregating risk
- stochastic modelling
- simple formulae if risk events are fully dependent
- correlation matrices
- copulas