Today, the field of life insurance (i.e. mortality as well as disability insurance and pensions) is undergoing significant changes. On the one hand, there is an increasing demand for segmentation and better risk assessment. On the other hand more comprehensive data sets have become available. The presentation identifies some difficulties with techniques that are widely used to cope with these challenges.

In the first part of the presentation, tree-based methods are suggested and discussed as an interesting alternative. In particular a “hybrid” approach (using regression trees for a classification situation) is proposed. The main advantage of this approach is its ease of interpretability and its inherent transparency.

The method appears to be particularly suitable for the identification of the risk factors in complex situations like disability insurance. In its second part, the presentation looks at the application of tree-based methods to disability data. In a case study based on German insurance data, the interdependencies of influence variables and their impact as risk drivers for disability probabilities are analysed. The identification of occupational classes with homogeneous risk profiles and the difference of male and female rates are of special interest.