Credibility factors - Intro to Ratemaking...

[Old Man]

Formula 3.6b on page 71 of the text has "max". I'm using last year's JAM study guide. It originally used "max" in the formula also, but errata was posted saying it should be "min". I haven't found an errata list for the text. Which is correct, "min" or "max"?

[Alya]

Formula 3.6b on page 71 of the text has "max". I'm using last year's JAM study guide. It originally used "max" in the formula also, but errata was posted saying it should be "min". I haven't found an errata list for the text. Which is correct, "min" or "max"?

it should be min - there is no way to have credibility factor > 1. :)

[New at pd]

We also need to remember that for function to be valid for uses as a credibility function, it must be have a positive first derivative and a negative second derivative. A good one to keep in mind is sqrt(n/10000)
where n = # claims.

[August]

It is min. The 2002 JAM states "min" and I've got it confirmed from another source. Otherwise we could get a credibility greater than 1.00.