Chapter 21: Use of Models in ERM

Question 1

The uses of models in ERM PADDA PC EFEC:

Answer 1

• Product pricing
• Advice given on risk management
• Day-to-day running
• Decision making
• Allocation of capital
• Project evaluation
• Checks and controls on the business
• Economic value of a company assessed
• Future earnings of the business estimated
• Effect of risk management on the business
• Capital modelling – MCR and economic capital

Question 2

Causes of inappropriate model/parameter selection HICE AOP

Answer 2

o	Data issues: 
	Heterogeneity in the data 
	Insufficient data 
	Completeness of data may lack
	Errors in historical data 
o	Distribution selection
	Alternative distributions not properly investigated  
o	Parameter selection 
	Oversimplification

Question 3

Features of a good actuarial model for ERM application DJ ADI SCARFS

Answer 3

• Documented
• Joint behaviour should be sensible
• Appreciate and communicate workings
• Development and refinement can be done
• Inputs and outputs should be valid for the business modelled
• Implementation methods should be variable
• Independent verification
• Simple, but not overly simplified
• Consistency with past events
• Amendable – after model analysis has been done
• Robust, but still sensitive to plausible future scenarios
• Fit for purpose
• Shortcomings clearly stated

Question 4

Steps to develop and apply a model PDF PPGT PIROS

Answer 4

• Purpose defined
• Data – collect and group
• Form of model – parameters and variables
• Period of model
• Parameters and correlations
• Goodness of fit assessed
• Project future cashflows
• Trade off in cashflow frequency considered (reliability vs. run time)
• Interactions between variables allowed for
• Run model with past data/stochastic simulations
• Output generated
• Sensitivity and stress testing

Question 5

How model output should influence decision making in ERM CAPERI

Answer 5

• Corporate risk policy applied
• Alternative decisions identified
• Preferences – deciding between risks on the efficient frontier
• Economic value added by a specific mitigation or action
• Risk appetite tested
• Intuition and judgement

Question 6

Factors to consider when choosing a model PANDA P SPEC

Answer 6

•	Purpose of the model 
•	Appetite for risk of the stakeholders
•	Nature of the risk 
o	Size
o	Patterns in the risk
o	Extreme values 
o	Correlations with other risks 
•	Data - HICE
•	Alternative models to consider
•	Parsimony principle

Question 7

Merits of Deterministic model LECE COS

Answer 7

Less Capital intensive
Explainable - since no distributions are applied

Clarity on scenario tested
Easy and quick to design

Carefully consider which scenarios will be tested
Only point estimates produced
Some scenarios may be missed

Question 8

Merits of a stochastic model WAQA SLICAH

Answer 8

Wider range of scenarios tested
Assess financial guarantees/assumptions tested

Quality result / Quantify risks better - Statistical analysis performed on the results
Allows for uncertainty and covariance between parameters

Spurious accuracy
Longer construction and run time
Interpretation and communication difficulty
Complex programming/ Costly
Additional capital intensity
Higher risk of model and parameter error due to complex nature