On this page
Use the two original constructions
Each of five period losses has marginal mean 2 and dispersion 10. The CSV compares independent periods with a shared standard normal shock. Both models sum period losses arithmetically; no compounding or position changes are assumed.
Record the aggregate basis
Independent periods give mean 10 and dispersion √5×10. A perfectly shared shock gives mean 10 and dispersion 50. The same period count and marginal moments therefore do not determine the same aggregate variance.
Retain dependence questions
| Field | Evidence needed |
|---|---|
| Aggregation | Additive losses or another definition? |
| Periods | Identical or differing parameters? |
| Joint law | Independent normal, shared shock or another model? |
| Covariance | Which cross-period terms? |
| Exposure | Static or changing loss construction? |
| Status | Declared example or assessed data? |
Keep the download’s role clear
The worksheet is a static review record, not a time-series estimator or a horizon-scaling service. If the supplied independence is unsupported, preserve the unresolved question instead of presenting one row as measured risk.