Skip to content

Normal loss assumption and normality evidence

A fitted mean and dispersion do not on their own establish that losses are normal.

On this page

Read the original calculator setup

The tool takes a declared normal loss law with supplied mean and standard deviation. Its calculation answers what the tail summaries are under those assumptions. It does not test the data or choose the law.

Retain the data-review questions

Different records
RecordEvidence or task
Declared lawExplicit normal assumption and compatible parameters
Parameter estimateObserved data and estimator
Distribution assessmentShape, tails, diagnostics and limitations
Exposure modelLoss construction and nonlinear pricing basis

Avoid a sample-size shortcut

A large observation count does not automatically make the original loss distribution normal. A central-limit statement about an average or sum needs its conditions and does not silently transform every individual observed loss law.

Use the appropriate follow-up

If empirical tails or exposures contradict the approximation, investigate a suitable loss model or additional scenarios. Record the unresolved issue rather than relabeling a mathematical result as validated market risk.

Further references