Stochastic optimization of a cash supply chain

We develop a mathematical model of a financial institution from a logistics point of view aiming to A) find optimal decisions related with cash inventory and transportation, B) balance the cost of the service and the quality perceived for final users, and C) taking into consideration the stochastic behavior of the cash demand series. A Stochastic Mixed Integer Programming (SMIP) model that represents the situation is developed and solved by using the Sample Average Approximation (SAA) algorithm.  Financial transactions involving cash have a random behavior and therefore the cash flows are random variables. The stochastic nature of the demand series is captured directly into the mathematical model and its solution algorithm. Our results shows that it is possible to reduce combined operational and opportunity costs up to 30%, without sacrificing the service level, by making better cash inventory and flow decisions, and that the model developed can be used to address other decisions over the studied system. The cost savings were calculated by applying the results coming from the model to a particular past period of time and comparing the cost against that observed with the actual inventory and transportation policies implemented by the bank. Our project shows that optimization supply chain models applied in the past to manufacturing supply chains, can also be adapted and applied to model a network of financial institutions’ branches; transportation, inventory, stockout and opportunity costs are similar cost elements to both supply chains.