Strategies for Modeling Loss Development: Curve Fitting, Credibility, and Layer Adjustments

This paper discusses some strategies to better handle the modeling of loss development patterns. Some improvements to current curve- and distribution-fitting strategies are shown, including the use of smoothing splines to help the modeled patterns better fit the data. A strategy is shown for applying credibility to these curves that produces results that are well-behaved and that can be implemented without the use of Bayesian software. Next, it is shown how the fitted models can be leveraged to help determine the optimal look-back period to use for selecting LDFs as well as to calculate the parameter and process error distributions. Lastly, a technique is demonstrated for making adjustments to LDFs for different limits, loss caps, and attachment points.