Module 25: Assessment of other risks

Question 1

Why are quantitative techniques not usually practical when assessing liquidity risk?

Answer 1

Due to lack of historical data on liquidity crises and the uniqueness of every organisation’s exposure to liquidity risk, quantitative techniques are usually not practical.

Question 2

The main methods of assessing liquidity risk: (2)

Answer 2

• scenario analysis - examining when and why expected cash outflows might exceed inflows, considering both short- and long-term scenarios
• stress testing - examining the effect on liquidity of an extreme event or change in a key assumption

Question 3

(Assessing liquidity risk)

Cash inflows include: (3)

Answer 3

• reasonably predictable revenues / income generated by assets
• proceeds from the sale of assets - may be highly uncertain (eg on forced sales)
• drawings upon sources of liquidity, eg raising capital

Question 4

2 measures of liquidity risk under Basel III

Answer 4

• Liquidity Coverage Ratio (LCR) - designed to ensure that banks can survive a one-month stress scenario.

• Net Stable Funding Ratio (NSFR) - designed to consider funding over a one-year time horizon.

Question 5

Demographic risk can be broken down into:

Answer 5

• level risk (or underwriting risk)
the risk that claims incidence and intensity is not as expected over the short-term, eg due to poor underwriting.

• reserving risk, which includes:

• – volatility risk, due to stochastic uncertainty in the experience of sub-populations
• – catastrophe risk - assumed to be downside only
• – trend risk (or cycle risk), due to exposure to potential longer-term changes in claims incidence and intensity.

Question 6

How might level risk be assessed?

Answer 6

Level risk may be assessed by combining both experience rating (determining the initial or central mortality rate) and risk rating (eg using GLMs)

Question 7

How can volatility risk be modelled?

Answer 7

Probabilistically or stochastically assuming some underlying statistical process.
As volatility risk varies by age, models are generally fitted by Poisson maximum likelihood estimation.

Question 8

High can catastrophe risk be modelled?

Answer 8

Catastrophe risk may be modelled by scenario analysis (considering sudden temporary increases in mortality) and capturing more complex dependencies using copulas.

Question 9

Non-life demographic risk can be broken down into: (2)

Answer 9

• level risk

- reserving risk

Question 10

Modelling non-life insurance risks involves some different complexities than demographic risks, including: (3)

Answer 10

• the intensity of claims also needs to be modelled - having greater uncertainty
• the possibility of more than one claim per policy
• the potential for each policy to move through (many) different states over its lifetime

Question 11

Define liquidity risk

Answer 11

Refers to a company not having sufficient short-term or cash-type assets to fund its short-term obligations.

Question 12

Funding liquidity risk

Answer 12

The risk of money markets not being able to supply funding to a business when required.

Question 13

Market liquidity risk

Answer 13

The lack of capacity in the market to handle asset transactions at the time when the deal is required (without a material impact on price).

Question 14

Outline why prediction cash outflows is particularly problematic for banks

Answer 14

Much of a bank’s liabilities will be in the form of deposits from customers who may withdraw their money with little or no notice.

Question 15

Liquidity analysis:

Describe 7 specific scenarios that should be considered

Answer 15

1. rising interest rates - eg a bank may find depositors transfer funds elsewhere in search of higher returns.
2. ratings downgrade - eg a bank may find depositors transfer funds to a more secure institutions.
3. large operational loss - resulting in a sudden reduction in cash-like assets.
4. large single insurance claim or a large set of claims from associated events - resulting in a sudden reduction of cash assets.
5. loss of control over a key distribution channel - resulting in a loss of expected revenues
6. impaired capital markets - equity investors or bondholders may be unable or be unwilling to provide fresh capital when required
7. sudden termination of a large reinsurance contract - leaving an insurer exposed to large cash outflows, but without expected inflows from the reinsurance contract.

Question 16

Define level risk

Answer 16

The risk that the particular underlying population’s claims incidence and intensity is not as expected over the immediate (short-term) future, eg due to shortcomings in the underwriting process.

Question 17

Define Volatility risk

Answer 17

Uncertainty with regard to the actual future immediate (short-term) mortality experience.

This arises due to only having a finite pool of policies.

As a consequence, it is not possible to measure precisely the past underlying rates of the underlying population and, going forward, the experience of sub-populations will exhibit statistical variations from that of the underlying population.

Question 18

Define catastrophe risk

Answer 18

An extreme form of volatility risk, eg the occurrence of a natural disaster resulting in a large number of deaths.

Question 19

Define trend risk

Answer 19

The risk of future (longer-term) changes in claims incidence and intensity.

Question 20

2 Distinct ways to determine the current underlying level of mortality

Answer 20

1. experience rating
2. risk rating

They may be combined by using credibility weightings.

Question 21

Experience rating

Answer 21

Involves examining the number of deaths in a portfolio of lives to determine the initial mortality rate or central mortality rate.

Question 22

Risk rating

Answer 22

Involves modelling the mortality rate (the response variable) of each homogeneous group as a function of the shared characteristics of the members (called covariates).

The model might take the form of a generalised linear model.

Question 23

Discuss 2 characteristics of the data upon which both experience and risk rating methods rely

Answer 23

Both rating methods rely on the data being:

• divided into homogeneous groups (eg male/female, employee type). There is a trade-off between the number of groups and ensuring the sample size in each is sufficiently large.
• collected over a period which is sufficiently long to generate adequate data, but not so long that the mortality rates could have varied greatly.

Question 24

How might volatility risk be modelled

Answer 24

Either probabilistically or stochastically, assuming some underlying statistical process,

eg a binomial or Poisson distribution.

Question 25

3 Stages of the Poisson Maximum Likelihood Estimation process

Answer 25

1. Calculate the expected number of deaths at each age (a function of the parameters)
   using the model to be fitted,
   and set this equal to the mean of a Poisson distribution.
2. Calculate the probability of the observed number of deaths at each age, based on the Poisson distribution derived above.
3. The fitted parameters are then obtained by maximising the likelihood function, ie the product of the probabilities (for all ages) that were determined in the preceding step.

Question 26

Discuss how trend or cycle risk for non-life insurance differs from demographic risks

Answer 26

More likely to correspond with the economic cycle (than for demographic risk), and so is best assessed using scenario analysis.

However, generally, non-life insurance risks have a shorter period of exposure than life insurance risks so longer term changes in risk factors are less important than a correct assessment of the risk factors themselves.