Marshall Flashcards
Why qualitative approaches are preferred when populating a correlation matrix
Quantitative approaches are costly, require data, and a lot of time
Bolt-on approach to determining risk margins
Separate analyses are completed to develop central estimate of insurance liabilities; does not involve single unified distribution
Claims portfolio
Aggregate portfolio for which the risk margins must be estimated
Valuation classes
Portfolios considered individually as part of the risk margin analysis
Claim group
Group of claims with common risk characteristics
Systemic risk
Risks that are common across valuation classes or claim groups
Two sources of systemic risk
Internal
External
Independent risk
Risks that occur due to inherent randomness in insurance process
Two sources of independent risk
Parameter
Process
Two sources of risk can be fully analyzed using bootstrapping or stochastic CL
Independent
Historical external systemic
Two sources of risk that cannot be fully analyzed using bootstrapping or stochastic CL
Internal
Future external systemic
Why traditional modeling techniques cannot capture all sources of uncertainty
Since models fit past data, only able to remove past episodes of external systemic risk, not future
Three sources of internal systemic risk
Specification error - due to model cannot perfectly model insurance process
Parameter selection error
Data error
Three external risk categories
Claim management process change Expense Event (Economic and social) (Legislative) (Latent claim)
Correlation between types of risk
Independent - uncorrelated with any other
Internal - tends to be correlated between valuation classes
External - correlation may exist between risks that belong to similar external systemic risk categories
Two distributions to calculate risk margins
Normal
Lognormal
Scenario testing for risk margin analysis
Used to determine how key assumptions would need to change to produce risk-loaded central estimate
Internal benchmarking for independent risk
Portfolio size - larger portfolio, lower volatility
Length of claim run-off - longer run-off, more time for random effects to have impact
When external benchmarking is useful
When little information is available for analysis
General hindsight analysis
AvE
Internal benchmarking for internal systemic risk
Classes with homogenous claim groups should have similar CoVs
Long-tailed portfolios should have higher CoVs
Internal benchmarking for external systemic risk
Long-tailed portfolios should have higher CoVs
