LDF Curve-Fitting and Stochastic Reserving: A Maximum Likelihood Approach

An application of Maximum Likelihood Estimation (MLE) theory is demonstrated for modeling the distribution of loss development based on data available in the common triangle format. This model is used to estimate future loss emergence, and the variability around that estimate. The value of using an exposure base to supplement the data in a development triangle is demonstrated as a means of reducing variability. Practical issues concerning estimation error and extrapolation are also discussed.