Modeling supplier risks using Bayesian networks

Abstract

Purpose

To counteract the effects of global competition, many organizations have extended their enterprises by forming supply chain networks. However, as organizations increase their dependence on these networks, they become more vulnerable to their suppliers' risk profiles. The purpose of this paper is to present a methodology for modeling and evaluating risk profiles in supply chains via Bayesian networks.

Design/methodology/approach

Empirical data from 15 casting suppliers to a major US automotive company are analyzed using Bayesian networks. The networks provide a methodological approach for determining a supplier's external, operational, and network risk probability, and the potential revenue impact a supplier can have on the company.

Findings

Bayesian networks can be used to develop supplier risk profiles to determine the risk exposure of a company's revenue stream. The supplier risk profiles can be used to determine those risk events which have the largest potential impact on an organization's revenues, and the highest probability of occurrence.

Research limitations/implications

A limitation to the use of Bayesian networks to model supply chain risks is the proper identification of risk events and risk categories that can impact a supply chain.

Practical implications

The methodology used in this study can be adopted by managers to formulate supply chain risk management strategies and tactics which mitigate overall supply chain risks.

Social implications

The methodology used in this study can be used by organizations to reduce supply chain risks which yield numerous societal benefits.

Originality/value

As part of a comprehensive supplier risk management program, organizations along with their suppliers can develop targeted approaches to minimize the occurrence of supply chain risk events.

Citation

Lockamy, A. and McCormack, K. (2012), "Modeling supplier risks using Bayesian networks", Industrial Management & Data Systems, Vol. 112 No. 2, pp. 313-333. https://doi.org/10.1108/02635571211204317