The idea for the IBNR Robot was conceived in 2016. By early 2017, the IBNR Robot has been used in pricing and reserving exercises. IBNR Robot has been shared on various conferences and company visits.
Nicholas Actuarial Solutions is a technology based actuarial solutions provider. In 2016, we participated in various IBNR related projects. It came to our attention that existing methods of estimating IBNR and its range has many shortcomings. Technically, these estimations are often biased, unstable and inconsistent over time and over different companies and classes of business. More significantly, the process is inefficient, using up too much actuarial human resources.
The IBNR Robot is an application of the actuarial methodologies employed in Cognalysis Reserving System™, a reserving software owned and distributed by Christopher Gross Consulting, Inc, also distributed outside North America by Nicholas Actuarial Solutions.
The philosophy underlying the IBNR Robot is different due to its fully automated process, and hence additional pragmatic methodologies such as outlier truncation and variance compensation adjustment are applied.
The IBNR Robot reads in paid claims, incurred claims and earned premiums as data. It then combines commonly used loss reserving methods eg link ratio method and Bornhuetter-Ferguson method with statistical optimization techniques and pragmatic solutions to arrive at parameters such as age-to-age factors including tail factors, seed loss ratios, resultant optimal IBNR methodology and individual business class as well as company aggregate IBNR estimates and its range.
It can complete in minutes what would otherwise take a day for the most efficient actuary to perform. At Nicholas Actuarial Solutions, we believe this is just the beginning of artificial intelligence and process automation in the field of actuarial science.
Get in touch through or drop us a message below.
Alvin Choong, FIA CERA
Easy to use, efficient, and extremely quick. A useful supporting tool for practicing reserving actuaries.
Dr Ken Sennewald, DAV
It was a very interesting demonstration of the IBNR Robot. The great additional benefit of using IBNR robot is the methodological blending of the various approaches into an optimum result. The automatic creation of reports is impressive and a huge time saver.
Potentially, an improvement could be to derive a Monte-Carlo based distribution of the test variable.