Chapter 12 A Framework for Assessing Risk Margins - K. Marshall et al.
Importance of qualitative analysis
- Parameter and process variance model with stochastic model
\(\star\) Three components of internal systemic risk 12.4
\(\star\) Score against best practice (Table 12.2, 12.3 and 12.4)
Calibrate to CoV for each component where CoV \(\in [5\%, 25\%]\)
- 2 Hindsight analysis
Check for internal consistency of the CoV
\(\star\) 7 Different risk categories
- Certain lines are impacted more than other by a given risk
CoV: Use benchmark similar to internal but select CoV directly
Correlation
The 3 main risk sources are independent of each other and therefore can be sum with square root rule (12.2)
- Pros and cons quantitative method for correlation
Independent (12.3):
Assume independence across lines, weight by liabilities
Internal (12.4):
Base on correlation matrix \(\Sigma\), again weighted by liabilities
-
Correlation between each valuation group and risk categories \(\Rightarrow\) then roll up to the risk categories and assume they are independent of each other
Risk Margin: (12.7)
- \(\alpha\)-tile (12.8)
\(\star\) Addition analysis
Sensitivity, scenario testing
\(\star\) Internal benchmarking, important to know the relationships and consistency, been heavily tested in the past
External benchmarking
Hindsight and mechanical hindsight
Regularity of review