A Methodology For Stochastic Inventory Modelling With Arma Triangular Distribution For New Products

This paper proposes a stochastic inventory policy of continuous review with random demand described with temporal dependence through an autoregressive moving average (ARMA) model with explicative variables, of usefulness in new products without a history of demand data, assuming a triangular distribution. Optimization of the cost function related to the inventory model is obtained considering the expected value and variance marginal stationary of the demand per unit time and stochastic programming. The proposed policy is exemplified with real-world demand data from a Chilean hospital, where the demand of products (drugs) are correlated with other products and autocorrelated. The proposed methodology shows a useful tool for administrators who must decide optimal batch sizes and their reorder points when there is a low availability of demand data and is known to have a temporal structure.