The Value of Information in an (R,s,Q) Inventory Model

In this paper we compare three methods for the determination of the reorder point s in an (R; s; Q) inventory model subject to a service level constraint. The three methods di er in the modelling assumptions of the demand process which in turn leads to three di erent approximations for the distribution function of the demand during the lead time.The rst model is most common in the literature, and assumes that the time axis is divided in time units (e.g. days).It is assumed that the demands per time unit are independent and identically distributed random variables.The second model monitors the customers individually.In this model it is assumed that the demand process is a compound renewal process, and that the distribution function of the interarrival times as well as that of the demand per customer are approximated by the rst two moments of the associated random variable.The third method directly collects information about the demand during the lead time plus undershoot, avoiding convolutions of stochastic random variables and residual lifetime distributions.Consequently, the three methods require di erent types of information for the calculation of the reorder point in an operational setting.The purpose of this paper is to derive insights into the value of information; therefore it compares the target service level with the actual service level associated with the calculated reorder point.It will be shown that the performance of the rst model (discrete time model) depends on the coe cient ofvariation of the interarrival times. Furthermore, because we use asymptotic relations in the compound renewal model, we derive some bounds for the input parameters within which this model applies. Finally we show that the aggregated information model is superior to the other two models.