Abstract
In this paper, we discuss various credibility and modeling strategies for loss development factors. We present several improvements to the popular inverse power curve to help it better fit to the data. Using a basic approach to credibility weighting curves often produces results that do not lie in between the original curve and the overall average, as would be expected. We show a technique to deal with this issue. We also discuss how to model across continuous variables as well as show formulas for converting and modeling across different loss caps, retentions, and policy limits.
Keywords: Loss Development Factors, Credibility, Generalized Additive Models, Splines
Volume
Summer
Page
1-25
Year
2015
Categories
Financial and Statistical Methods
Statistical Models and Methods
Generalized Linear Modeling
Actuarial Applications and Methodologies
Ratemaking
Trend and Loss Development
Financial and Statistical Methods
Credibility
Publications
Casualty Actuarial Society E-Forum