Statistical problems are addressed by optimally analysing bank and insurance data that are of a categorical nature. The purpose is to determine the multi- dimensional influences and behaviour patterns of variables in data (e.g. the lapse of policies or defection of clients). This is done within the framework of actuarial science.

Principles of categorical data analysis and survival analysis are combined to summarise factual implications resulting in accurate future predictions. In the case of specific characteristics such as policy lapses under consideration, the influence of predictors are quantitatively measured in terms of indices, specifying the global picture.  These indices are estimated and implemented in a very effective way.