On this page
Read the original record
The default is a declared normal loss law with mean 0 and dispersion 10, in one hypothetical horizon and common loss unit. Confidence is 95%. No observed exposure or data source is implied.
Interpret each quantity
| Quantity | Value | Interpretation |
|---|---|---|
| VaR | 16.448536 | Upper loss quantile |
| ES | 20.627128 | Average of worst 5% |
| Quantile coefficient | 1.644854 | Standard normal cumulative crossing |
| Tail coefficient | 2.062713 | Standard normal selected-tail average |
| Selected probability | 5% | Complement of confidence |
Make a controlled parameter comparison
Changing the mean shifts both loss summaries equally. Changing dispersion scales them by different coefficients. Changing confidence changes the selected quantile and tail mass. Retain a new assumption record for each comparison.
Keep standardized and original coordinates distinct
The density’s x coordinate is z, not the original loss amount. Convert a coordinate to loss using μL+σL z. The graph’s endpoints are display choices rather than maximum losses. A zero-dispersion input uses a constant limit instead of that standardization.