Supply chain risk network management: A Bayesian belief network and expected utility based approach for managing supply chain risks

Abstract

The paper develops and operationalises a supply chain risk network management (SCRNM) process that captures interdependencies between risks, multiple (potentially conflicting) performance measures and risk mitigation strategies within a (risk) network setting. The process helps in prioritising risks and strategies specific to the decision maker's risk appetite. The process is demonstrated through a case study conducted in a global manufacturing supply chain involving semi-structured interviews and focus group sessions with experts in risk management. Theoretically grounded in the framework of Bayesian Belief Networks (BBNs) and Expected Utility Theory (EUT), the modelling approach has a number of distinctive characteristics. It utilises a top-down approach of Fault Tree Analysis (FTA). Performance measures are identified first and subsequently connected to risks. A ‘probability-conditional expected utility’ matrix is introduced to reflect the propagation impact of interdependent risks on all performance measures identified. A ‘weighted net evaluation of risk mitigation’ method is proposed and the method of ‘swing weights’ is used to capture the trade-off between the efficacy of strategies and the associated cost keeping in view the decision maker's risk appetite. The approach adapts and integrates techniques from safety and reliability engineering (FTA), decision making under uncertainty (EUT), and multi-criteria decision analysis (swing weights). The merits and challenges associated with the implementation of interdependency based frameworks are discussed. Propositions are presented to elucidate the significance of modelling interdependency between risks and strategies.