Micro-level parametric duration-frequency-severity modeling for outstanding claim payments

Unlike collective models, individual models have the advantage of keeping attributes each claim intact. We propose a three-component parametric model that uses this information in form explanatory variables. The first component predicts delays between occurrence, report, and closure using survival models. For second (frequency) third (severity) components, we use generalized linear splice Moreover, elapsed time report claims is converted into an exposure variable count model. Finally, discuss estimation procedures, make predictions, compare results with other data set from major Canadian insurance company.