Evaluating Long-Term Service Performance under Short-Term Forecast Updates

Abstract

Managing product availability in a cost effective way has always been a major challenge faced by inventory managers. We study the problem of a firm selling a perishable product with short-term demand patterns and a long-term service target using the newsvendor framework. The newsvendor determines his long-term order at the first stage, and revises the order according to a short-term forecast update at the second stage. He also evaluates a long-term service target for his overall performance on product availability across all possible forecast updates. We characterize his optimal inventory policy that minimizes the expected inventory cost while meeting the long-term service target. Both in-stock rate and fill rate targets are examined.