Pricing different limits for different periods

The ICRFS-Plus™ MPTF module can be used to calculate retrospective (and prospective) reinsurance with different attachment (limit) points for policy (accident) periods. Correlations between all cells are used when calculating means, standard deviations and percentiles. Different attachment points can also be applied across future calendar periods and policy (accident) periods.

In order to conduct this analysis, the data are first organized into different layered triangles corresponding to the different attachment points. For example, for attachment points 100K and 200K say, layered triangles are created Lim100K and Lim200K. In respect of the array Lim100K, for example, paid losses are in respect of individual losses limited to 100K. [With out loss of generality we confine the example to two layers].

A composite model is identified (designed) in MPTF which captures the volatility in each layer, and the two types of correlations (process and parameter) between the two layers.

Based on the identified model, the PALD module is used to generate simulated values by accident year, calendar year (and total) for each layer. The simulations incorporate correlations between the cells, within a layer, and between layers. These correlations induce correlations between any accident/calendar year in one layer with any accident/calendar year in the other layer.

A 'composite' array is constructed from the two simulated arrays based on the limits selected for the different accident years. For an accident year for which the limit is 100K, the corresponding simulated values are taken from the Lim100K array and so on.

This is shown below. Any statistic including percentiles can be computed from the simulated composite array. [Quite a few statistics can actually be obtained from the composite model itself].

Only ICRFS-Plus's unique MPTF modelling framework coupled with the PALD module facilitates such unique, creative, sound solutions to critical long tail liability issues.