Impact of nested inventory allocation policies in a newsvendor setting

We model an inventory management setting in which the decision maker first uses newsvendor model to decide on the amount of ordered perishable inventory for a fixed consumption period, based on the best available forecast of demand at the time of ordering. After a relatively long lead time, the consumption period starts and she has to assign the received inventory to two priority customer classes given the -newly updated- rate of arrival of each class. The assignment of inventory requires two simultaneous decisions: 1) the reservation quantity for the high priority class and 2) the choice of inventory allocation mechanism (Standard Nesting SN or Theft Nesting TN); to minimize the expected units short of the high priority class while minimizing the wasted inventory at the end of the fixed consumption period. We assume that some partial information about the bottom line impact of a shortage in high priority customer class compared to the other can be conjectured. For both inventory allocation mechanisms, we then calculate the monetary benefit for all feasible reserved quantities to identify the optimal reserved quantity. We derive closed form expressions for the expected number of units short in each demand class under SN and TN allocation mechanisms. We showcase the management of electricity smart meter inventory in a multi-year implementation project consisting of multiple fixed consumption periods. Numerical experiments and graphical interpretations feature the optimum allocation policy and the cost minimizing reserved quantity under such policy.