SHORT-TERM PURCHASE-INVENTORY DECISION MAKING UNDER HIGHLY FLUCTUATING PRICES

The problem of purchasing in the open market when (a) there is a deadline, (b) a single price quotation is given for each period and it remains stable during the period, (c) each price quotation is a random variable, and (d) the purchaser is faced with a buy-or-wait decision has all the characteristics of a stopping-rule problem. The applicable stopping-rule solution in such a case takes the form of a multiple reservation price policy; i.e., there exists a separate reservation price for each period. Only if the purchasing horizon is infinite (no deadline) does the stopping-rule solution yield a single reservation price applicable for all periods. Even then the cost for generating each price quotation, the cost c, must be the same for all periods and it must be positive. Because all practical purchasing problems usually have a deadline and because the cost of postponing the purchase for one period is a saving equivalent to one day of inventory holding costs, the single reservation price solution of the optimal stopping-rule does not apply to typical purchasing problems.;The performance of each model using both uniform and Laplace distributions as generators of periodic price quotations is measured, documented, and compared. The influence of price variability and percent inventory holding costs on purchasing performance of each model is studied. Clear guidelines are then provided in order to assist the user in selecting and implementing an appropriate purchasing strategy. Finally, the interaction of purchasing and inventory control models and the resultant net savings are determined and discussed.;This research first formulates a single reservation price model not based on the optimal stopping-rule approach (Model I). The model is then broadened and converted to a multiple reservation price type (Model II) and it is shown that the broadened model coincides with that obtained using the optimal stopping rule.