Incorporating crossed classification credibility into the Lee-Carter model for multi-population mortality data (Q784458)

The paper focuses on a crossed classification credibility structuring of the Lee-Carter model, in the case of multi-population mortality schemes. After presenting notations and basic concepts concerning crossed classification credibility modeling, the classic Lee-Carter model is revisited and reformulated according to an approach based on crossed classification credibility formulation for multi-populations. This modeling choice constitutes the main advancement compared to the classic Lee-Carter model, where the time index follows a time series process. As part of this approach, the parameter estimation procedure is presented in detail, as well as the mortality forecasting process appropriate to the model. Numerical illustrations concretely show the application and development potential of the model. Finally, a stochastic version of the model is proposed.