ADP Week 4

12345678910111213141516171819202122232425262728293031
Across
  1. 2. Capital requirement that varies with measured risk exposures, rather than a flat amount.
  2. 3. Average loss given losses exceed the VaR level
  3. 9. risk Risk of rare, extreme outcomes in the far ends (“tails”) of a loss distribution.
  4. 11. “Tail heaviness”/peakedness relative to normal (interpret carefully
  5. 12. risk  The risk that actual outcomes differ from the model’s expected value even when the model structure and its parameters are correct, due to inherent randomness in the underlying lossgenerating process.
  6. 13. distribution A probability distribution describing potential losses over the model horizon.
  7. 14. model A quantitative model that estimates how much capital an organization needs to absorb losses over a defined horizon at a specified confidence level.
  8. 17. Measures of dispersion around the mean.
  9. 19. capital Minimum capital required by regulators under prescribed rules (e.g., standardized formulas or approved internal models).
  10. 22. inference Updates prior beliefs with data to produce a posterior distribution of parameters.
  11. 25. interval Range of plausible parameter values given sampling uncertainty (frequentist concept).
  12. 27. Blending individual experience with collective experience based on statistical weight.
  13. 28. Changing variance over time or across observations.
  14. 30. adequacy Whether available capital is sufficient relative to required capital under the chosen standard/model.
  15. 31. of capital Assigning total required capital to business units/risks (e.g., Euler allocation) for performance and pricing.
Down
  1. 1. A description of probabilities for all possible values of a random variable.
  2. 4. Property that an estimator converges to the true value as sample size grows.
  3. 5. How risks move together
  4. 6. benefit Reduction in required capital due to imperfect dependence across risks.
  5. 7. Loss threshold not exceeded with a given probability over a horizon (a quantile of the loss distribution).
  6. 8. limit A quantitative boundary (e.g., max VaR, max PML, max exposure) used to keep risk within appetite.
  7. 10. 50th percentile.
  8. 12. risk The risk that the parameters used in the model are wrong or imprecise because they are estimated from limited, noisy, biased, or nonrepresentative data (or expert judgment).
  9. 15. A fixed (unknown) quantity defining a distribution (e.g., mean, sigma).
  10. 16. period Expected time between events exceeding a given level (e.g., “1in200year loss” ≈ 0.5% annual probability).
  11. 18. Measure of joint variability
  12. 20. An observation far from others
  13. 21. testing Assessing outcomes under extremebutplausible scenarios not fully captured by historical data.
  14. 23. capital Capital required to remain solvent at a chosen confidence level (often aligned to a target credit rating), based on the organization’s own risk view (not accounting rules).
  15. 24. Asymmetry of a distribution.
  16. 26. How probable the observed data are under specific parameter values.
  17. 29. event A lowprobability, highseverity event (market crash, megacat, mass tort, etc.).