Unit 4 Flashcards
ISO31000 - risk assessment =
Risk identification- what might happen?
Risk analysis- likelihood and impact?
Risk evaluation- so what? Is it within appetite?
Risk assessment considerations-top down vs bottom up. Top down advantage & dis
Advantages =
- enterprise wide approach
- most significant strategic risks and manageable number
- buy-in at top
- likely to be consistent methodology
Disadvantages =
- more focused on external risks
- limited internal operational risks/ interdependencies identified
- too superficial as SM believe can manage crisis
- new emerging operational risks not identified
Bottom up advantage and disadvantage
Advantages
- buy in at all levels
- risk impact beyond immediate op risks discussed
- op risk aware of local risks and causes
- vary methodology for norms / culture
Disadvantage
- little focus on external/ strategic risks
- time consuming
- too detailed and siloed
- new risks might not be reported
Risk assessment techniques
- Questions and checklists
- Workshops and brainstorming
- Inspections and audits
- Flow charts and dependency analysis
VAR definition
Measure of potential loss in a portfolio over a given time horizon within a given confidence interval assuming normal markets and No trading
VAR statement says
There is an x percentage probability that losses will not exceed y in the next n days
VAR description cont
Most of the effort in calculating VAR involves estimating the return distribution (gain or loss) values that the portfolio might return and the probability of each value
Methods of calculating VAR
Historical, co-variance, Monte Carlo
Historical- based on real historical data but the future might be different
Var- variance and co-variance
Assumptions:
- results are normally distributed ie bell shaped
- correlations between risk factors are constant
- sensitivity to changes of risk factors is constant
Doesn’t work well for non-linear. Strength- easy to calculate
Weakness- doesn’t reflect the historic distributions
VAR- Monte Carlo
Randomly selects trials based on a selected probability distribution. But how validate outcomes
VAR shortcomings
- Measured with estimation error- model makes assumptions therefore not precise statement
- VAR does not give any info about the severity of loss by which it is exceeded. Eg 95% confidence level losses not expected to exceed VAR 5 days out of 100, but on any of those 5 doesn’t say how big a loss
- VAR doesn’t describe the worst case loss but the worst case for a specified confidence level.
Expected Shortfall
For a given time period and confidence level, ES is the average loss that could occur in excess of the loss calculated by VAR over the same time period and using the same confidence level.
ES is NOT the worst case loss. Requires greater info about the extreme tail return distribution so harder to calculate
Stress testing and scenario analysis used for;
- assessing capital and liquidity requirements
- understanding the dynamics of the risk environment and therefore providing a tool for decision making
- challenging the output of VaR or other models
- informing senior management.
Types of stress test
Sensitivity- change factors in models
Historical events- what would happen to the portfolio if it happened today
Customised events- previous hurricane took a different path-lot of work
Multi year stress test- ‘what if’ changes to business plan-typically 3-5 years ahead
Reverse stress test- start from the point the business is unviable
Stress testing / scenario analysis prescribed by regulators
Domestic recession
Global financial crisis
Funding market shutdown
Interest rate fall
Natural catastrophe-hurricane
Global pandemic
