CvmortalityMult: Cross-Validation for Multi-Population Mortality Models

This article presents the CvmortalityMult R package, a novel tool designed for modelling, forecasting and evaluating mortality models for several populations. The package facilitates the fitting and forecasting of multipopulation mortality models, providing accurate projections in an increasingly interconnected world characterized by minimal or no borders between countries. By incorporating different cross-validation (CV) techniques, the package allows for the assessment of the forecasting accuracy of multipopulation mortality models for specific countries or regions within a country. Through an empirical application to Spanish regions, we demonstrate the efficacy and simplicity of the CvmortalityMult R package in selecting and evaluating multipopulation mortality models. By providing accessible tools for mortality modelling, forecasting and testing, this package stands out as a valuable resource for advancing the understanding and forecasting of mortality trends across diverse populations. Its contributions extend to enhancing decision-making in critical fields such as life insurance, public health, and pension plan sustainability.