Application of Two-Stage Stochastic Programming for Multi-Product Inventory Optimization in a Nigerian Pharmaceutical Firm

Abstract

Inventory management remains a key challenge in pharmaceutical supply chains, especially in emerging markets like Nigeria. Unpredictable demand, poor infrastructure, and limited budgets often cause stock imbalances. Traditional inventory models work well in stable environments but struggle to manage demand and lead time uncertainties. This study uses a two-stage stochastic programming model to improve multi-product inventory management for Miracle Kudura Pharmacy, a Nigerian retail and wholesale distributor. Based on 37 months of sales, holding, and ordering cost data for over 100 products, the model incorporates uncertainty through probability distributions grounded in real figures. In the first stage, it sets preliminary order quantities without knowing actual demand. In the second, it updates these quantities as actual demand becomes clear. The initial stage reduced inventory costs to ₦1,224,000, demonstrating cost-effective efficiency. The second stage, however, reported a total cost of ₦1,435,000, showing a trade-off between cost savings and flexibility. Sensitivity analysis revealed that demand for products such as Clarten and Postinor varied widely, while items like Tribotan were more stable. These findings highlight the importance of stochastic models in managing costs, service levels, and risks which are crucial factors in pharmaceuticals where shortages impact public health. Overall, the study highlights how stochastic optimization offers practical, adaptable solutions for managing pharmaceutical inventory in resource-constrained environments.