Mod 15: Introduction to risk modelling Flashcards
Describe how methods of quantifying risk are commonly used in respect of the major risk types
Methods of quantifying risk
Quantitative models are commonly used in respect of: market & economic risks (including interest rate, foreign exchange and basis risks), credit risks, liquidity risks, insurance and underwriting risks.
Operational risks are currently largely quantified using scenario analysis and simulations due to the relative lack of data. However, organisations are increasingly collecting historic data and quantitative analysis will become more common. Nevertheless, worst case scenario analysis will remain important as it only takes one such scenario to bring down a company.
Dynamic financial analysis models an organisation’s risk exposures and the relationships between these risks to produce projected balance sheets and P&L accounts. An internal capital assessment is one such example. A Financial Condition Report is similar but is aimed at an assessment of the current solvency position of a company and its possible future development.
List the four main issues in risk quantification
Issues in risk quantification
1. difficulties in assessing possible emerging risks and future extreme events which, although often high severity and low probability, remain of interest and require mitigation. These are, by their nature, hard to predict but:
* an emerging risks register may help an organisation to respond .
* an organisation that learns from prior extreme events may be in a better position to react to future such events
* may be modelled using extreme value theory
2. data limitations (eg limited volume, heterogeneous)
* alternative data sources introduce issues
3. difficulties in assessing the interdependence of risks eg use of risk ranges or risk buckets
4. how to deal with unquantifiable risks −
(eg lower reliability)
State three approaches to calculating correlation and outline what is implied by negative, low positive and high positive correlation
Approaches to calculating correlation
1.linear correlation
2. rank correlation
3. tail correlation – perhaps based on the lowest and highest k% of a sample, another example is tail dependence
- Negative correlation implies (partially) offsetting risks.
- Low positive correlation implies (some degree of) diversification.
- High positive correlation implies concentration of risks.
©
Define linear correlation
vbk://9781035523344/page/454
Outline the advantages and disadvantages of linear correlation
vbk://9781035523344/page/456
Define two forms of rank correlation, giving the formulae for two statistics
vbk://9781035523344/page/458
Describe the properties of rank correlation ©
vbk://9781035523344/page/460
Outline three approaches to modelling risk using a deterministic model
Approaches to modelling risk using a deterministic model
1.
sensitivity analysis −involves varying each input assumptions one at a time to quantify its independent effect on the model’s output
2.
scenario analysis −involves changing multiple inputs simultaneously – each ‘what if’ scenario represents a plausible and internally-consistent set of future conditions
3.
stress testing −
similar to scenario or sensitivity testing but it focuses only on extreme scenarios or very large changes in input assumptions
Note that all three approaches quantify the impact but not the probability.
List three main reasons a company would wish to use sensitivity analysis
The main uses of sensitivity analysis
1. to develop an understanding of the risks faced
2. to provide an insight into the dependence of the output on subjective assumptions
−it can help to focus attention on the most important assumptions, and make clear the model’s reliance upon judgement
3. to satisfy a supervisor’s requirements −
eg Value at Risk is a commonly used risk measure within the banking and insurance sectors
©
List the main advantages and disadvantages of scenario analysis
The main advantages and disadvantages of scenario analysis
Advantages
* facilitates evaluation of the potential impact of plausible future events
* not restricted to consideration of what has happened in the past and so can assess vulnerability to high impact, low probability events
* provides useful additional information to supplement statistical models
* facilitates the production of action plans by assessing the possible impact both pre-and post-implementation of response strategy
Disadvantages
* potential complexity as a process
* reliance upon successfully generating hypothetical extreme but also plausible events
* uncertainty over whether scenarios are representative or exhaustive
* as with sensitivity analysis, no probabilities are assigned to any of the scenarios
Outline six important practical considerations when carrying out scenario analysis
©
Important practical considerations in carrying out scenario analysis
1. In practice, the terms scenario analysis, stress testing, scenario testing may be used interchangeably; stating the meaning helps avoid confusion.
2. The risk practitioner should assess the impact of both historical and hypothetical scenarios.
3. Scenario analysis is not solely a quantitative technique (eg qualitative applications in operational risk assessment).
4. As well as the financial impact, scenario analysis should assess practical implications, eg operational strain, regulatory interest, dividend affordability, consequences for customers.
5. The analysis should be appropriate and proportionate to requirements.
6. The accuracy of the analysis depends on the accuracy of the data and inputs used.
©
List the main advantages and disadvantages of stress testing
The main advantages and disadvantages of stress testing
advantages
* facilitates the production of action plans by assessing the possible stress impact both pre-and post-implementation of response strategy
* enables supervisors to compare the impact of the same stresses on differing organisations
* examines extreme events which might not otherwise be considered, eg if a stochastic approach was adopted.
Disadvantages
* assigns no probability to the events considered
* looks only at extreme situations, and so needs to be coupled with other techniques in order to understand the full range of outcomes
* the choice as to which assumptions to stress and the degree of stress(es) to consider is subjective
Outline two approaches to stochastic modelling
Approaches to stochastic modelling
Many simulations are run through a model. The results of these simulations give a distribution of potential outcomes.
1.
historic simulation (bootstrapping):
− each simulation is generated by direct reference to historical data
2.
forward looking approaches:
* Monte Carlo simulation: each simulation uses random numbers to generate input values (eg selecting randomly from each input variable’s probability density function)
* factor-based approach: causal links between variables are described explicitly within the model
* data-based approach: focuses more on modelling the key variables that provide the best fit to the data rather than the driving factors
Outline the main advantages and disadvantages of historical simulation (bootstrapping)
©
The main advantages and disadvantages of historical simulation
Advantages:
* applicable to many situations, if suitable past data is available
* does not require large amounts of data (sampling with replacement)
* does not require the specification of probability distributions
* reflects the characteristics of the past data (including non-linearity, non-normality, interdependencies etc) without the need for parameterisation
Disadvantages
* cannot be performed without relevant past data
* assumes that past data is indicative of the future
* does not consider correlations between past data items
* may underestimate uncertainty (based only on what actually happened not what could have happened
©
Outline the main advantages and disadvantages of Monte Carlo simulation
The main advantages and disadvantages of Monte Carlo simulation
Advantages
* software is widely available and is easily adapted and updated
* increasing the number of simulations increases the output’s accuracy (ie reduces estimation error)
* it is possible to simulate the interdependence of risks
* it is a widely understood technique
* it can be used to model complex financial instruments (eg with non-linear, non-normal payoffs), such as derivatives
Disadvantages
* the random selection of parameter values may lead to a set of simulations which is not representative of the full range of possibilities, unless the set is sufficiently large
* large sets of simulations may be time consuming to perform
