previous article in this issue | next article in this issue |
Document Details : Title: Robust Credibility Author(s): GISLER, Alois , REINHARD, Peter Journal: ASTIN Bulletin Volume: 23 Issue: 1 Date: May 1993 Pages: 117-143 DOI: 10.2143/AST.23.1.2005104 Abstract : Outlier observations caused by big claims or by an event producing a series of claims are a special problem in ratemaking and in tariff calculation. The authors believe that combining credibility and robust statistics is the right answer to this problem. The main idea is to robustlfy the individual claims experience by using a robust estimator 7], instead of the individual mean X̄, and to look at the credibility estimator based on the robust statistics {T,:l = 1, 2, .. }. Choosing a particular influence function leads to data-trimming with an observation-dependent trimming point. |