Modelling (s,Q) inventory systems: Parametric versus non-parametric approximations for the lead time demand distribution

In the practice of inventory management the identification of the shape of the lead time demand (LTD) distribution based on limited data is a difficult task. This paper compares the costs effects of modelling the LTD in a continuous-review (s,Q) inventory model, by using several parametric approximations against a so-called ‘kernel density’ approximation. In a simulation study it is shown that the proposed kernel density approach, with a local varying band width, holds promising results. Cost differences between several parametric approximations and the kernel density approach can be shown to be in favour to the last approach, especially in situations where service levels are high and the LTD distribution has certain peculiarities.