On this page
Start with the scope of the estimate
A reviewer needs the exposure or loss-record identity, common horizon, loss unit, sign convention, source and estimation method. If the numbers are an original hypothetical example, say so; do not describe them as observed risk.
Give a precise original example
For the declared −10/0/10/50 model with 30/40/25/5 percent probabilities, the lower 90% loss quantile is 10 units and exact-tail ES is 30 units. The selected ES mass is 10%; probability strictly above VaR is 5%. This is a model summary, not a guaranteed limit.
Make the next review actionable
| Issue | Evidence or action |
|---|---|
| Input calibration | Provide observations or retain assumption label |
| Quantile convention | State cumulative rule or interpolation |
| Horizon comparison | Supply compatible multi-period record |
| Aggregation | Show joint probabilities and dependence basis |
| Unexpected result | Reconcile signs, units and selected mass |
Keep professional claims separate
Explaining this arithmetic does not by itself qualify a person for a risk role or certify a regulatory model. Retain the specific unresolved issue and the evidence needed to resolve it.