La posibilidad de un índice de longevidad para la Unión Europea bajo las directrices de solvencia II

Resumen

We believe that a proper calibration and a well-defined algorithm are necessary for the construction of an adequate longevity index, regarding a given country or population, but the adequacy of a generalised longevity index it is less clear to us. Further to this, in our view, the use of a longevity index benchmark is not always suitable. At the same time, while the mathematical model is important for the process of obtaining the index. Our investigations advocate the importance of the data collection source justification, the data cleansing and improvement, as well as, the analysis of the data. Our work follows the guidelines set by Solvency II regarding the risk of an aging population and we understand that is the way to build a LifeMetrics that meets best-estimate´s requirements of objectivity, transparency, robustness and accuracy. With the aim of simplifying the results of the research done, we only present the results for three countries which we consider the most significant ones, such as: England and Wales, USA and Germany. We obtained an index of longevity for those three different countries using HMD database in order to use data from their own government statistics entity. We have found that the differences between the results obtained for each country applying its own model and those obtained by us using the Lee-Carter model, are slightly insignificant, what indicates that our Lee-Carter model is acceptable enough. Therefore, we have used this model for the construction of a longevity index for four new countries, such as: Italy, Spain and Japan. Once all the results were obtained using homogeneous criteria for each of the countries observed, and after the analysis of the estimation of the evolution of mortality and its improvement through the calendar years, we 1 Catedrático de la UPV/EHU. amancio.betzuen@ehu.eus. 2 Profesora de la UPV/EHU. amaiajone.betzuen@ehu.eus. Este artículo se ha recibido en versión revisada el 28 de septiembre de 2015. La posibilidad de un índice de longevidad para la UE … Anales 2015/73-106 74 found that the improvement in mortality is not homogeneous. It is not distributed equally, nor by gender, nor by age sections, or by calendar year, nor has the same percentage of improvement in each country analysed. The outcome of our research allows us to conclude that, a common index for all the European Union countries would not reflect, with sufficient guarantee, the longevity risk features of each European country. We also believe that, a longevity index should not represent a static scenario, as it would appear from the document submitted by Life and Longevity Markets Association (LLMA), it rather should extend to a dynamic scenario with the aim to reduce the risk of an aging population that generates long term benefits’ payments in the future. The periodic calculation of a longevity index does not seem the best solution. Following these guidelines we have built a longevity index for Spain, taking into account the guidelines of Solvency II related to the risk of aging. As a result of our research, we conclude that a unique and dramatic increase of mortality improvement is not advisable. The improvement is not constant; it depends on the age and calendar year considered in the estimation. And we also add that it also depends on the contractual timing of the event. In addition, apart from the purge we have carried out through all the steps when obtaining the data and the results of the research, we have also analyzed the volatility of the results of the mortality improvement over the recent calendar years, obtaining results that we could qualify as very acceptable. These represents an additional guarantee of the reliability of the longevity index we obtained.