An Actuarial Note on Credibility Parameters

in this paper the relationship between the Bayesian credibility parameter, k, and the classical credibility standard for full credibility, F, is examined from a practical standpoint. A very useful “rule of thumb” is developed. For most practical applications one can determine the F that roughly corresponds to k, and vice versa. First convert k to a number of claims, if necessary, by multiplying by an expected frequency. Then take F equal to approximately eight times k. A few other interesting results are also derived. Among them is the effect of misestimating the Bayesian credibility parameter k. The results of using credibility are relatively insensitive to misestimates of k.