Analytic and bootstrap estimates of prediction errors in claims reserving

We consider an appropriate residual definition for use in a bootstrap exerc ise to provide a computationally simple method of obtaining reserve predict ion errors for a generalised linear model which reproduces the reserve esti mates of the chain ladder technique (under certain restrictions which are s pecified in the paper). We show how the bootstrap prediction errors can be computed easily in a spreadsheet, without the need for statistical software packages. The bootstrap prediction errors are compared with their analytic equivalent from other stochastic reserving models, and also compared with other methods commonly used, including Mack's distribution free approach (M ack, 1993. ASTIN Bulletin 23 (2), 213-225) and methods based on log-linear models. (C) 1999 Elsevier Science B.V. All rights reserved.