Skip to main content
eScholarship
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

Barter Exchange: Modeling, Analysis, and Participant's Strategy

  • Author(s): MIN, ZHAO
  • Advisor(s): Shen, Max
  • et al.
Abstract

In this study, we considered modern corporate barter platforms. We evaluated the probability of a participant on the platform being able to find an exchange partner(s) and his/her waiting time before being able to barter. Four stochastic models that characterize corporate barter platforms operating under different exchange mechanisms were analyzed. These models enable participants to concurrently barter with multiple partners in various exchange volumes. In addition, we prove that the platforms studied in the models are stable in terms of queueing theory, i.e., they do not expand such that there are infinite numbers of waiting participants. It was found that, under one of the exchange mechanisms (called online with pairwise exchange mechanism), the participants' waiting time is exponentially distributed. The expectation and variance (or their bounds) of waiting time were analyzed under other mechanisms also. Participants' preference was investigated based on the probability of being able to exchange and the corresponding waiting time under each exchange mechanism. Based on the participants' preference, suggestions for improving a platform's performance and profitability are provided accordingly.

Continued from the above results, participant's strategies under online with pairwise exchange mechanism are investigated. For a participant who implements EOQ model to purchase production materials, and who is willing to try barter to get rid of excessive inventories and barter-in production materials, a concise sufficient condition is developed to identify if it is beneficial to participate in a barter exchange. With the satisfaction of sufficient condition, if the participant cannot be matched with exchange partner(s) upon arrival, we suggest he always wait on the barter platform for future opportunities. While waiting, if his regular procurement and production plan does not permit back-order in production materials, a good strategy for the procurement of production materials during the waiting time is maintaining the usual practice with the regular supplier, and do not change the procurement activity until he is able to barter-in production materials. Additionally, if back-order in production materials is allowed during the waiting time on the barter platform, then a better temporary procurement plan with back-order may be available compared with no back-order allowed. We found a sufficient condition that justifies if back-order will be beneficial, and the sufficient condition becomes necessary as well if the participant's cost function (including ordering cost, inventory holding cost, and back-order penalty cost) is convex. Last but not least, though simultaneous trade-in and trade-out is a common practice in barter, we recommend that participants separate trade-in and trade-out processes to reduce the total waiting time. The reduction in waiting time may outweigh negative effects from the separation of trade-in and trade-out activities.

Main Content
Current View