Use of Markov Process Theory and Thiele's Differential Equation in Practical Claims Reserving

A straightforward derivation of Thiele’s differential equations for the moments of present values of a payment stream governed by a discrete time Markov process is presented. Two different applications of the theory are then discussed. The first deals with the problem of rehabilitation on disability annuity claims and the second bases predictions on case estimates. Emphasis is put on straightforwardness in deriving the computational scheme and exemplifying the features relevant in an implementation.

Keywords: Claims reserving, disability annuities, rehabilitation, Markov process theory, Thiele’s differential equations.