CIA MfAD Flashcards
Definition of MfAD
Reflect the degree of uncertainty of the best estimate assumption. It is the deviation of actual from expected xp resulting from:
• Error in estimation
• Change of expected xp from non-anticipated influences
• Statistical fluctuations
Desirable risk margin characteristics
- Less that is known about the current estimate and its trend, the higher the risk margin should be
- Risk with low freq/high sev should have a higher risk margin than risk with high freq/low sev
- Longer contracts should have longer risk margin
- Risk with wide Pr distribution should have higher risk margin than risks with narrow Pr distribution
- If emerging xp reduce uncertainty, risk margin should decrease
Identify three PfADs
- Claims development
- Recovery from Re Ceded
- IRR
Provide examples where a large MfAD is appropriate (compared to the best estimate assumption)
- Less confidence in the best estimate assumption
- Approximation with less precision is being used
- Event assumed is farther in the future
- Potential consequence of the event is more severe
- Occurrence of the event is more subject to stat fluctuations
Provide examples where it would be appropriate to have a margin of 20% for claim development
- Significant changes due to tort reform
- Introduction of new LOB
- Significant change expected in future claims
- Financial crisis and its effect on longer tailed LOB
Calculate investment return MfAD according to the “explicit quantification of 3 margins” approach
Total Margin = Asset/Liab Mismatch Margin + Timing Risk Margin + Credit Risk Margin
Calculate asset/liability mismatch risk margin
Coverage Ratio * [(Asset D - Liab D) / Liab D] * (Interest Rate Movement in Runoff Period)
***Coverage ratio = policy liabs / invested assets
Calculate timing risk margin
L/(1+d’)^D = L/(1+d)^D’
MfTR = d-d’
d' = Discount rate adj for timing risk D' = Changed duration of liabs
Calculate the credit risk margin
Extra yield on a corporate bond compated to a risk-free bond
MfAD disclosure considerations
- Complexity of the concept
- Importance of the concept to users
- Sophistication of users
NOTE: These are same as AA disclosure considerations
Types of Quantile approaches
- Multiples of Standard Deviation: Simple and practical
- Percentiles or Confidence Levels: Most common
- Conditional Tail Expectation: Better for more skewed distributions
Calculate investment return MfAD according to the “weighted formula” approach
MfAD = iPM – iAM
iAM = interest rate for discounting after MfAD
= min[iPM, iRFM*(1-k)]
iPM = interest rate for discounting, prior to MfAD
iRFM = interest rate of risk-free bonds that matches the payout of liabs
k is the % by which iRFM would need to be adj to reflect a plausible shortening of the uncertain duration of the claim liabs
Provide techniques to derive MfADs
- Deterministic techniques
* Stochastic techniques are not expected to be so high that the Pr(unfavorable dev)
State one thing PfAD do not cover and one thing it covers when using stochastic models
- Do not cover the stat volatility arising from the model
* Covers the uncertainty in whether the actuary has the right model or the right parameter
Provide and briefly describe general considerations for claim dev MfADs
Any situation where there is a significant change in:
• Insurer’s operations (claim management, UW, other operations)
• Data on which the estimate is based (volume of losses, homogeneity)
• LOB (Length of tail, latent claim, liab exposure)
Provide considerations for MfADs from Re ceded
- High Ceded LR
- High Ceded commission rates
- Significant Unregistered Re
- Significant Re under liquidation
- Significant Re with weak financial condition
- Significant disputes with Re
Provide the several different types of risk addressed for investment return rates
- Mismatch risk between payment of claims and availability of liquid assets
- Error in estimating payment pattern for future claims
- Asset risk including credit/default and liquidity risk
Provide considerations for MfADs for investment return rates
- Matching of assets and liabs
- Low asset quality
- High loss of capital
- High asset default risk
- Determination of interest rates
- Concentration by type of investments
- Economic conditions (recession)
Which type of product will a stochastic modelling approach be benefiting the most ?
Products with skewed loss distributions with low freq/high sev • Stop loss Re • CAT insurance risk • Credit, warranty, mortgage insurance • Long tailed LOB
Discuss documentation of MfAD deviations
- Actuaries to document the considerations that were critical in their selections of MfADs
- Actuaries conducting stochastic analyses document what components are modeled as random variables as well as primary assumptions
- Documentation for both explicit and stochastic techniques would include support for key decisions
Discuss why normal distribution would not be appropriate in PC Insurance
- Rarely enough risk
- Risks are usually correlated
- Risks are rarely symmetric*
*Positive skewness: Distribution with a high Pr of having no claim and decreasing Pr as the claim amount increases
List features a risk margin methodology should have
- Consistent methodology for lifetime of the contract
- Assumptions consistent with current estimates
- Consistent with sound pricing practices
- Vary by product based on risk diff by product
- Consistently determined between reporting periods
- Consistently determined between entities
- Consistent with IASB objectives
- Consistent with regulatory solvency and other objectives
