This workshop will focus on the use of Actuarial Case Reserves to solve many of the problems with triangle development (changes in mix, settlement rate, and case adequacy). By using predictive modeling techniques to develop objective algorithms for these reserves, reserving triangles can be re-built to be much more reliable and transparent (think Berquist-Sherman, but with a scalpel instead of a chainsaw). While the approach makes use of individual claim development models, framing the results back into triangles provides demonstration of the resulting algorithm’s effectiveness and stability. Benefits of using these algorithmic reserves extend not only to creating estimates of total reserve need, but to actuarial pricing work as well.
While the focus of this half-day workshop will be on the creation and use of Actuarial Case Reserves, we will also discuss the concept of Policy-Level IBNR reserves as an additional triangle layer, reflecting estimated not-yet-reported claims.
No prior predictive modeling experience is necessary, but there will be two hours of pre-work to maximize the effectiveness of time spent during the workshop itself. Users will be provided access to temporary Azure workstations for a week prior to the workshop until a week after the workshop so that they can continue to explore these concepts with the sample data and tools provided.