Normal quantile and conditional tail mean

Separate the two summaries

The quantile locates the loss cutoff at a confidence level. The upper conditional mean averages losses above that cutoff under the positive-dispersion continuous normal law. It must therefore lie beyond the threshold.

Use the original zero-mean ten-unit model

Retain signed location effects

Adding the same mean loss shifts both values equally; it does not remove their difference. A sufficiently negative mean can make both summaries negative while the normal law still permits positive losses.

Keep the continuity condition

The conditional-tail description here relies on the continuous normal law. An atom-containing finite model can require fractional boundary inclusion for the exact selected-tail measure. Keep the referenced definition with the number.