An efficient entropy-based stopping rule for mitigating risk factors in supply nets

Potential supply-net risk factors include capacity issues, currency volatility, design changes, frequent changes in tax regulations, unsafe information systems, and port shutdowns. Such risk factors make it challenging for decision makers to design efficient risk-management procedures and minimize the operational costs associated with mitigating these risks. Since complete information about all risks is generally not available, we utilize an axiomatic approach, based on the notion of entropy, to develop an efficient solution procedure. The strategy involves developing a stopping rule that enables the decision maker to decide online whether or not to continue investing in the acquisition of information about risk factors. The problem is formulated as a non-linear integer optimization model. We develop sufficient conditions for the entropy function such that a unique global solution is obtained. A case study that addresses risks associated with the transportation of aquatic products in refrigerated containers demonstrates the superior performance of the stopping rule relative to a standard risk-assessment procedure. In addition, numerous computerized experiments are carried out, under different problem settings, to compare the stopping rule with an anticipative optimum. The cost incurred when using the stopping rule is found to be no more than 0.4% higher than the cost of the anticipative optimum (the lower bound for the objective). These findings clearly demonstrate the efficiency of the proposed stopping rule for a wide range of problem sizes.