| In many supply chain environments, the transportation providers and product manufacturers absorb the inherent variability caused primarily by uncertain customer demand. In this dissertation we explore models where a supply chain agent, a distributor for example, contractually commits to a specific sequence of deliveries over a finite horizon. We term such an arrangement a “committed delivery” strategy. While such commitments limit the ability of the distributor to react to random demand, the manufacturers and transportation providers can efficiently use their resources, resulting in increased system profit. In exchange for such a contractual agreement, the distributor can expect a reduction in purchase and/or transportation cost for units acquired via committed delivery. Motivated by a distributor of industrial food products, we develop economic models for a distributor facing price-sensitive demand and given an opportunity to commit.; Commitment is most beneficial to the manufacturer and carrier when the delivery quantity and frequency are fixed. For this case, we develop expressions for both optimal price and commitment level. Using a constant elasticity expected demand function, we develop expressions for the overall percentage increase in profit due to commitment and the percentage increase in profit due to the cost reduction (rather than improved pricing) in terms of the discount and the elasticity of demand only. We decompose the positive effect of commitment into a “cost reduction” effect and an “improved pricing” effect. We find that for small to moderate commitment discounts, the cost reduction effect dominates, even when demand is highly elastic.; In certain environments, the manufacturer and carrier may be willing to allow the distributor some flexibility in the commitment level. We consider the case where the distributor receives a discount for commitment but may commit to a sequence of time-varying quantities. This problem can be viewed as stochastic production planning with lost sales. The multi-dimensional decision space makes obtaining optimal solutions difficult. We develop efficient heuristics that perform very well in our computational experiments. Furthermore, our experiments show that the value of flexibility in commitment is significant, especially when the commitment discount is large relative to the holding cost. |
