Credibility for Severity [Discussion]

This is an inspiring paper very clearly written and well presented. I hope that the point made by Mr. Hewitt comes home, namely that credibility is theoretically justifiable and eminently practical. The main contribution of this paper is the explicit application of general credibility techniques to the factorized pure premium, the factors being expected frequency and expected severity. So far, most other applications have been made by either applying the same techniques to the expected frequency alone or directly to the pure premium, without considering its product form. Credibility, Experience Rating Plans/Territorial Rating