The expectation-based loss-averse newsvendor

Abstract

Abstract. We modify the classic single-period inventory management problem by assuming that the newsvendor is expectation-based loss averse according to Kőszegi andRabin (2006, 2007). Expectation-based loss aversion leads to an endogenous psychological cost of leftovers as well as stockouts. If there are no monetary stockout costs, then the loss-averse newsvendor orders a quantity lower than the quantity ordered by a profit-maximizing newsvendor. If there are positive monetary costs associated with stockouts, then the loss-averse newsvendor places suboptimal orders, which can be either too high or too low.

The expectation-based loss-averse newsvendor under supply uncertainty

“…Finally, if the salvage value of the product increases, the newsvendor will order more. This is consistent with the EBLA model in [1], where demand is uncertain.…”

“…However, Reference [7] mention that rational expectation determines the reference point. The concept of choice-acclimating personal equilibrium (CPE) defined in [7] was applied by Herweg [1] who analysis EBLA newsvendor model under random market demand. Newsvendor who faces uncertain supply just like the one who confronts uncertain demand, they both place their order before the uncertainty is resolved.…”

“…Thus, it is important to apply this hypothesis on newsvendor model. Reference [1] studies a newsvendor problem with uncertain demand under the framework of expectation-based loss-averse (EBLA). Yet, as we know, supply uncertainty is another source of uncertainty confronted by the decision maker in reality.…”

“…Our study is a natural extension of [1] by assuming that a newsvendor faces supply uncertainty instead of demand uncertainty. More specifically, we consider a newsvendor who knows the demand, yet the supply is uncertain.…”

A Review of Behavioral Decision Making in the Newsvendor Problem

“…Table 3 summarizes our findings. Valuation of outcome is derived from the amount of deviation from chosen reference point Newsvendor will be riskseeking (q > q * ) over negative deviation (Loss domain) and risk-averse (q < q * ) over positive deviation (Gain domain) Schweitzer and Cachon (2000), Herweg (2013), Nagarajan and Shechter (2013), Long and Nasiry (2014), Uppari and Hasija (2014), Zhao and Geng (2015), Li et al (2017) Newsvendor will be charged extra penalty for each unit of leftover inventory, so he would order less than optimal quantity (q < q * )…”