Optimal decisions for the loss-averse newsvendor problem under CVaR

Optimal decisions for the loss-averse newsvendor problem under CVaR
复制标题

DOI:
10.1016/j.ijpe.2015.03.019
复制
发表时间:
2015-06
影响因子:
12
通讯作者:
Xinsheng Xu;Z. Meng;Shen Rui;Jiang Min;Ji Ping
Xinsheng Xu;Z. Meng;Shen Rui;Jiang Min;Ji Ping
中科院分区:
工程技术1区
文献类型:
--
作者:
Xinsheng Xu;Z. Meng;Shen Rui;Jiang Min;Ji Ping

文献摘要

被引文献

相似文献

现有的关于报贩问题的文献大多集中在选择最优订货量以使风险中性报贩的期望利润最大化。然而,一些研究指出,在实践中,管理者对订货量的决策总是偏离利润最大化的订货量,这被称为报贩问题中的决策偏差。然后,许多研究引入了其他偏好,而不是报贩的风险中立来解释这种决策偏见。因此,本文将损失厌恶引入到报贩决策偏差的研究中。本文对报贩的损失规避问题提出了新的定义——遗留损失,遗留损失可以定义为超额订单的损失,也可以定义为到期销售损失的短缺罚款。规避损失的报贩的最优决策可通过以下三种方式获得:(i)最小化预期遗留损失;(ii)通过在风险管理中结合CVaR度量来最小化遗留损失CVaR; (iii)最小化预期遗留损失和遗留损失CVaR的组合;并与其他已有结果进行了比较。此外,我们还给出了最优决策的一些性质以及它们之间的关系。给出了一些数值算例来说明所得结果。
Most of the existing literature about the newsvendor problem mainly focused on choosing an optimal order quantity to maximize the expected profit of a risk-neutral newsvendor. However, some studies pointed out that in practice managers׳ decisions as to order quantities always deviate from the profit-maximization order quantity, which is referred to as decision bias in the newsvendor problem. Then, many studies introduced other preferences rather than risk neutrality of the newsvendor to explain such a decision bias. This paper thus introduces loss aversion to the research of the newsvendor decision bias. We propose a new definition to the loss-averse newsvendor problem–legacy loss, which is defined either as the loss for excess order or the shortage penalty for lost sales when sales time is due. The optimal decisions of a loss-averse newsvendor are obtained through the following three ways: (i) minimizing the expected legacy loss, (ii) minimizing CVaR of legacy loss by combining the CVaR measure in risk management, (iii) minimizing the combination of expected legacy loss and CVaR of legacy loss; and are compared with other existing results. Besides, we present some properties of the optimal decisions as well as relations among them. Some numerical examples are given to illustrate the obtained results.