Impact of Reference Point on Lose-Averse Newsvendor Problem

Newsvendor problem is a fundamental singleitem inventory problem with stochastic demands over a single period. In the traditional newsvendor model, the expected utility theory is introduced to measure the risk attitude of decision-maker. However, behavior of decisionmaker for many practical problems cannot be explained with the expected utility theory. As a result, prospect theory of the behavioral economics was adopted to describe the risk preference of decision-maker. In this paper, value function of the prospect theory was used to quantify the risk bias of the decision-maker. A widely accepted piecewise-linear form of value function was considered. The impact of the reference point on the optimal order quantity was, then, analyzed. The results show that the optimal order quantity of newsvendor problem on prospect theory is not a unique value. The order is in an interval varying with different reference point. Furthermore, several typical stochastic demand distribution density functions, for example, uniform distribution, exponential distribution were tested for the model. The results of numerical examples were consistent with some anomalistic findings of experiments in the literature.