Tradeoff between expected reward and conditional value-at-risk criterion in newsvendor models

Two common approaches to addressing risk in the newsvendor setting are to maximize the probability of achieving a target profit and the newsvendor's expected utility, respectively. In this paper we introduce a weighted mean-risk objective. In particular, we consider the tradeoff between the expected profit and conditional value at risk (CVaR). The CVaR criterion measures the average value of the profit falling below a quantile level which is commonly known as the value at risk (VaR). We derive the optimal order quantities and discuss comparative static properties in terms of optimal order quantity, the wight used in the objective function and the degree of risk aversion of the newsvendor.