Nonparametric Estimation of Extreme Quantiles with an Application to Longevity Risk

A new method to estimate longevity risk based on the kernel estimation of extreme quantiles truncated age-at-death distributions is proposed. Its theoretical properties are presented and a simulation study reported. The flexible yet accurate conditional having survived certain age fundamental for evaluating lifetime insurance. Our proposal combines parametric with nonparametric sample information, leading obtain an asymptotic unbiased estimator alternative different right tail shape, i.e., heavy or exponential tail. estimating continuous temporary annuity also shown. We illustrate our application official statistics population in Spain.