Managing decentralized supply chains: Information, inventory sharing, and quality-based procurement

Xinghao Yan, Purdue University

Abstract

In this dissertation, two research topics are investigated: the demand information asymmetry in a decentralized inventory sharing supply chain and the quality-based supplier selection. For the first research topic, previous literature on inventory sharing has assumed that retailers’ demand information is known to each other. Such an assumption may not capture the reality for decentralized supply chains. In this thesis, we study an inventory sharing game between two retailers who privately hold demand information, non-cooperatively determine their order quantities, but cooperatively share inventory with each other. Characterizing retailers’ Bayesian Nash equilibrium order quantities when no demand information is shared between the retailers, we find that sharing demand information does not always benefit retailers. However, a coordination mechanism through the involvement of the manufacturer can be developed if the retailers share true demand information with each other before they each make ordering decisions. Unfortunately, an incentive compatibility analysis shows that the retailers have the incentives to untruthfully reveal their demand information under such a coordination mechanism. A truth-inducing scheme is then designed which, together with the coordination mechanism, achieves the coordination of the manufacturer-retailers system and guarantees an all-win situation for all parties involved. To our knowledge, this paper is the first attempt to study inventory sharing systems with asymmetric information. It also provides a coordination mechanism for a decentralized inventory sharing system that applies not only to the full information setting, but also to an asymmetric information setting. In this dissertation, we also extend the 2-retailer decentralized inventory sharing system to a more general n-retailer system. Similar to the 2-retailer system, we develop a retailer coordination mechanism, nRCM, for the n-retailer system. We first design nRCM under the scenario in which the retailers’ demand information is common knowledge. Then we consider the asymmetric retailers’ demand information scenario. We conduct the incentive compatibility analysis and find that under nRCM, the retailers do have incentives to report their demand information untruthfully. Finally, we also design a truth-inducing scheme to ensure each retailer tell truth when he reveals his demand information. Different from the 2-retailer system, nRCM not only implements a transhipment benefit allocation among n retailers, which is much more complicated than 2-retailer system, but also ensures that the retailers fully share their residual supply/demand and do not form any coalition, which are two most important issues in n-retailer inventory sharing system. As for the second research topic, we explore quality-based supplier selection when quality can be improved and quality itself also exhibits some uncertainty. To hedge against the quality uncertainty, the buyer can choose from two different approaches in selecting the supplier. With a quality requirement approach (QR), the buyer keeps her first-mover right by posting quality requirement to the potential suppliers and making her supplier-selection decision based on the reactions from the suppliers. With a quality promise approach (QP), the buyer gives up her first-mover right by asking each supplier to provide quality promise, based on which she makes her supplier-selection decision. With either approach, quality can be improved (with costs associated) by the suppliers and/or by the buyer. We investigate how these two different approaches affect the suppliers’/the buyer’s efforts in quality improvement and the cost of the supply chain under the setting. We find that due to competition and decentralization, (1) The first-mover’s objective is always aligned with that of the whole supply chain, while the second-mover aims at minimizing his/her own cost; (2) While fully exploiting the first-mover’s cost effectiveness in quality improvement, QR and QP restrict the second-mover’s cost effectiveness, but in different ways; (3) The first-mover right does not always benefit the buyer. In fact, if the supplier is more efficient in quality improvement, giving up the first-mover right (QP) is likely to benefit the buyer more. On the other hand, if the buyer is more efficient in quality improvement, then keeping the first-mover right (QR) may be more beneficial; (4) As for the whole system, if the supplier (buyer) is more efficient in quality improvement, QP (QR) more likely induces better quality and lower system cost (hence higher social welfare). These result are different from those in the previous literature in which no competition is considered.

Degree

Ph.D.

Advisors

Zhao, Purdue University.

Subject Area

Management|Operations research

COinS