The following communication is not an original paper, but the summary of an article published by the author in collaboration with H. Zimmermann in the Bulletin de l'Association des Actuaires Suisses, L, pt. 2 (1950). As the method of reserve calculation by grouping described below assumes a fundamentally new point of view and combines substantial advantages with great simplicity, it may, however, interest those actuaries who have no access to the original paper.

As is well known the individual valuation of every single policy, even for a medium-sized portfolio, is a long and arduous task. Therefore different methods for the valuation of policies grouped in certain ways have been in use for a long time. In this respect it is possible to distinguish methods giving the exact total of the individual reserves (e.g. the so-called Altenburger or Karup or valuation constant method) and those leading to approximate results (e.g. the Lidstone or Z-method). But most group methods have the evident disadvantage of a grouping which is in other respects inappropriate for the classification of the individual policy within the portfolio. Thus, in the valuation constant method, groups of equal age attained have to be formed, and the Z-method requires groups of the same remaining term. Yet one would much prefer a grouping which necessitates no sorting of the policies contrary to their natural grouping. Thus the putting together of all policies with the same year of entry for the purpose of collective reserve calculation is in some way a natural manner of grouping, and this was the procedure used in the t-method formerly proposed by the author.