Skip to main content
Journal Article

United we stand or divided we stand? Strategic supplier alliances under order default risk

Management Science 62 (5): 1297–1315
Xiao Huang, Tamer Boyaci, Mehmet Gumus, Saibal Ray, Dan Zhang (2016)
Management sciences, decision sciences and quantitative methods; Product and operations management
Cooperation, competition, supply risk, coalition stability, supplier alliances
We study the alliance formation strategy among suppliers in a framework with one downstream firm and n upstream suppliers. Each supplier faces an exogenous random shock that may result in an order default. Each of them also has access to a recourse fund that can mitigate this risk. The suppliers can share the fund resources within an alliance, but they need to equitably allocate the profits of the alliance among the partners. In this context, suppliers need to decide whether to join larger alliances that have better chances of order fulfillment or smaller ones that may grant them higher profit allocations. We first analytically characterize the exact coalition-proof Nash-stable coalition structures that would arise for symmetric complementary or substitutable suppliers. Our analysis reveals that it is the appeal of default risk mitigation, rather than competition reduction, that motivates cooperation. In general, a riskier and/or less fragmented supply base favors larger alliances, whereas substitutable suppliers and customer demands with lower pass-through rates result in smaller ones. We then characterize the stable coalition structures for an asymmetric supplier base. We establish that grand coalition is more stable when the supplier base is more homogeneous in terms of their risk levels, rather than divided among a few highly risky suppliers and other low-risk ones. Going one step further, our investigation of endogenous recourse fund levels for the suppliers demonstrates how financing costs affect suppliers’ investments in risk-reducing resources and, consequently, their coalition formation strategy. Last, we discuss model generalizations and show that, in general, our insights are quite robust.
© 2016 INFORMS
Journal Pages