Chapter 18 : Models (1)

Question 1

What are the requirements of a good model?

Answer 1

CLERICAL ADVISORS

Capable of refinement
Length/expense of run not too long/ high
Easy to understand
Rigorous
Independent verification of outputs
Clear results
Adequately documented
Large range of implementation methods

All significant features allowed for
Developable
Valid
Inputs to parameter values appropriate
Sensible joint behaviour of variables
Output workings are communicable
Reflects risk profile
Simple whilst retaining key features

Question 2

What are some basic features of life insurance models?

Answer 2

1. Projection cashflows and profit (cost of supervisory reserve and solvency margins for profit flows)
2. Allowance for interaction and correlations
3. Allowance for guarantees and options
   4.Projection frequency and time period

Question 3

When is a stochastic model preferred?

Answer 3

1. When we need to cost guarantees and options (withdrawal & mortality levels as well)
2. To see a dbn of outcome not a single estimate.
3. Interaction between variables.
4. To estimate a probability

Question 4

When is a deterministic model preferred?

Answer 4

1. Sensitivity testing to get an approximation to a stochastic result.
2. Where results obtain would be similar to a stochastic model
3. As a check on a stochastic model

Question 5

What is risk-neutral calibration of a stochastic model?

Answer 5

Aims to replicate the market prices of actual financial instruments as closely as possible. Typically used for valuation.

Step 1: Choose a number of financial instruments that a price is known for.
Parameters are chosen such that average PV of cashflows = close to known market price

Question 6

What is real-world calibration of a stochastic model?

Answer 6

Aims to use realistic long-term expectations

Normally used for projecting into the future. Model parameters= expectations of the future

Question 7

What are the 4 different types of models

Answer 7

1. Profit testing
2. New business
3. Existing business
4. Full office model

Question 8

What are the steps of modelling?

Answer 8

1. Specify purpose
2. Stochastic/Deterministic
3. Data and assumptions
4. Model points
5. Project cashflows
6. Run model to achieve a satisfactory results
7. Test sensitivities