Vendor managed inventory and cost reduction in an international two level supply chain

The current economy demands companies to find new ways to grow and increase shareholder value. Companies that are not developing or finding competitive advantages are eventually being pushed out of the global competitive environment. One of the ways to gain an edge over the competition is to reduce inventory without negatively impacting operating costs or customer service. This paper describes an analysis performed for the implementation of a vendor managed inventory system in a two-level supply chain. A model for calculating replenishment quantities is presented, and problems such as incentives to deviate from optimal inventory levels and imbalanced savings are addressed. Vendor Managed Inventory (VMI) is a planning and management system that optimizes the supply chain performance in which the vendor is responsible for maintaining the customer's inventory levels but is not necessarily tied to inventory ownership. Under the traditional "serially linked" supply chain business model, when a customer needs a product, it places an order against the vendor. The customer is in control of the size and timing of the order being placed. Under the VMI business model, popularized by Wal-Mart, the vendor receives data from the customer regarding sales and stock levels, and thus it can decide when to initiate the replenishment procedure. By receiving timely information directly from the customer, the overall inventory requirements are reduced because the vendor's need for excess stock to buffer against uncertainty is eliminated. In this research, we analyze the impact of using a VMI system in a two level supply chain in which material flows in both directions. Specifically, we partnered with TRW Safety Restraint Division to analyze the supply chain formed by their Mesa, AZ and Chihuahua, Mexico facilities which are currently working under the traditional business model. We hypothesized that the implementation of a VMI system in both facilities would reduce their overall supply chain costs. In our experiment, we develop a mathematical model to calculate the ordering quantities from one plant to another based on their demand forecasts, safety stock levels, and a fixed ordering frequency. Experimental data taken from TRW shows that by implementing a VMI system, they can reduce their inventory levels by 40% on average. The research also addresses common problems faced by VMI systems such as deviation from optimal inventory levels and imbalanced savings. As it has been previously discussed in the literature, the vendor working under VMI might try to obtain individual gratification and therefore deviate from the overall optimal behavior which minimizes the cost for the supply chain as a whole resulting in decreased savings. We present an analysis that shows that for the considered problem a VMI system in conjunction with consignment eliminates the incentive of the vendor to deviate from the overall optimal behavior. In the case of imbalanced savings, we consider the possibility of transfer payments. We also consider the possibility of counterbalancing savings by using VMI for SKUs of both facilities. In this way, both plants can enjoy from the benefits of having their inventory managed and owned by the vendor.