PROFIT OPTIMIZATION IN THE PRICE SETTING NEWS-VENDOR MODEL WITH POISSON DEMAND

Abstract

The Price Setting Newsvendor (PSN) model is one of the fundamental models for Operations Research and Management Science. In its most basic version, the Price Setting News-Vendor model allows retailers, who sell their products over a single selling period, to determine the optimal ordering quantity and selling price that maximize their expected profits. The set of products covered by the News-Vendor model can range from perishable food and bakery items to short life cycle items such as seasonal fashion goods or even electronics. In our model, we represent the demand on these products by a Poisson distribution, as some of them can face a low customer appeal due to their availability in a wide assortment or their occasional use. We also consider substitutable retail products, that are horizontally differentiated variants, under an additive-multiplicative demand setting and a logit consumer choice model. Under these settings, we propose a coordinate ascent algorithm that finds a local maximum of the profit function. Moreover, we devise optimality conditions that allow for checking whether the computed solution is a local or global optima. These conditions are used to develop a method that allows escaping local solutions that are not global. We validate our results by conducting various numerical experiments with random inputs of the model parameters. Finally, we evaluate the effectiveness of our approach compared to existing optimization heuristics applied under similar problem settings, and we suggest areas for improvement in future research.