Media Hr|bluebox – Lapse Analysis Using CART

Hr|bluebox – Lapse Analysis Using CART

uploaded October 13, 2022 Views: 81 Comments: 0 Favorite: 1 CPD

Early lapses of life insurance policies can be a great nuisance to primary insurers. Typically, the agent or broker receives a commission for each policy sold, which can be higher than the annual premium paid by the policy holder. If the policy holder then lapses before the end of the first year, the primary insurer suffers a financial loss, even though no claim was made. hr|bluebox is a data analytics service offered by Hannover Re with the goal of decreasing the early lapse rate. There is no universal solution to this problem, as the reasons why policy holders lapse differ between products and markets. In order to identify the drivers of early lapse in a specific portfolio, a crucial step is therefore to detect and describe high-lapse or low-lapse segments in the existing data. 

For this purpose, we developed a modified version of the Classification and Regression Trees (CART) algorithm that is targeted not at individual prediction, but at segment detection. The results are easy to interpret and validate and thus form a useful basis for decision making. In this presentation, we will give you an overview of the hr|bluebox approach, talk about our experience when applying this approach and share with you some lessons learned.  



Content groups:  content2022


There are no comments yet. Add a comment.