Over the last decade, a vivid market for pension block transactions has developed. Initially concentrated on the UK, market activities are spreading into increasingly more countries. Like any life and health insurance business, pension block transactions are connected to a multitude of biometric risks, some more and some less well-researched.
Among the most obvious risks are the mortality base and trend risks. Several actuarial societies have published tables and projections, or even projection toolkits. Some – in particular the British IFoA – have even involved themselves in research on stochastic projections. We also find a lot of related academic research. A striking aspect is that much of the existing research on mortality projections considers the situation of individual countries in isolation. Some even go so far to fit, in addition to the parametrization of a model, also the very structure of the model to the data of the one country under consideration. We take the position that this is a dangerous approach. While for the short term we obviously expect any projection to continue the country-specific dynamics and trends, our view for the mid- to long term is that in many cases the demographic and medical dynamics driving the trends are fundamentally similar. Thus, using only country-specific long-term projections seems misleading. We suggest to take a more global perspective.
Less obvious at first sight are the biometric risks related to dependents. In many defined benefit pension funds, “any spouse” surviving the pensioner will continue to receive benefits after the pensioner’s death. This exposes the risk carrier to a multitude of risks, which are often underappreciated – the proportion of married individuals among the pensioners, re-marriage, and the age difference, to name a few. We discuss approaches to a better understanding of such risks.