Inventory control under demand with non-stationary variance

摘要

Recent studies suggest that increasing variability of customers’ demand undermines validity of traditionalrnassumptions used in inventory management. This paper develops reorder point inventory models under demand withrnnon-stationary variance. Autoregressive conditional heteroskedasticity models are used to model and forecast thernnon-stationary variance. The elaborated models are evaluated by their experimental comparison with traditionalrnmodel using both simulated and empirical demand data. These models are aimed to improve efficiency of managingrnA class inventory items under highly variable demand.