A Note on Full Credibility for Estimating Claim Frequency [Discussion]

While receiving greater attention in the past few years, credibility standards still lack clarity and structure. This becomes all too apparent when it is realized the standards in common use today are intended for claim frequency only, but are routinely applied to pure premiums. In addition, the assumptions underlying the traditional Poisson claim frequency process are rarely fulfilled. These assumptions include homogeneity of risks and randomness of claims (implying both an 'accidental' nature and mutual independence).