Research Of Robust Optimization Decisions Of Supply Chains Under Supply Risk And Overconfidence

Prior to the COVID-19 outbreak,supply chain managers generally focused on the risk of uncertainty from the market demand side and ignored the risk of supply uncertainty to the supply chain.In fact,supply risk has gradually become a serious problem and evolved into a major cause of supply chain inefficiency.Besides,since it is difficult to obtain information about the true distribution of uncertain supply in real life,the traditional stochastic program-ming that assumes that uncertain supply follows a specific distribution is not suitable.At the same time,there are cases in supply risk management where policymakers have an overcon-fident preference for product supply uncertainty.Moreover,there are three different forms of overconfidence: overestimation,overpositioning,and overprecision,where overestimation and overprecision behaviors have completely different effects on decision makers.Based on this,a distributionally robust optimization approach is used to investigate how firms mitigate sup-ply risk and the impact of decision maker’s overestimation and overprecision on the pricing and ordering decision of the supply chain when there is supply uncertainty risk with unknown distribution.Specifically,this topic focuses on the following four research work:(1)Based on the price-setting newsvendor problem,we discuss the effects of emergency supply and responsive pricing strategies on mitigating supply risk and the strategic relationship between them,when companies are faced with uncertain supply and limited distributional in-formation of stochastic yield.We find that under the mean-variance ambiguity set,when the unit emergency purchase cost is relatively small,the firm prefers to use emergency supply and the two strategies are e strategic substitutes.Otherwise,responsive pricing is better and the two strategies are strategic complements.In addition,under the mean ambiguity set,respon-sive pricing does not work,and emergency supply and responsive pricing are always strategic complements.(2)Based on the price-setting newsvendor problem,we explore the impact of retailers’ overconfidence behavior on product pricing and ordering decisions under demand or supply risk with unkown distribution.By constructing a mean-variance ambiguity set and adopting the mean-preserving but variance-reducing transformation to characterize overconfidence behavior,we find that under different sources of uncertainty(demand or supply),overconfidence affects product retailer price and retailer’s ordering quantity differently.Moreover,when the source of uncertainty is demand,different demand models lead to different results.Finally,we obtain that overconfidence remains a cause of cognitive bias for decision makers in the framework of price-setting newsvendor.(3)We extend the above price-setting newsvendor problem to a supply chain system,which consists of a supplier that suffers from stochastic yield and an overconfident retailer.In partic-ular,we simultaneously examine two types of overconfident behavior of the retailer: overesti-mation and overprecision.An overprecision decision maker underestimates the variance of the random yield; an overestimation decision maker overestimates the mean of the random yield.For an uncertain market environment where only the mean and variance of the stochastic yield are known,we develop an optimal robust decision model for the supply chain and find that these two overconfidence behaviors have completely different effects on channel members.The re-tailer’s overconfident behavior leads to a loss of profit for itself but benefits the supplier.More-over,both overconfident behaviors can reduce the double marginalization effect and coordinate the supply chain system.(4)The effects of two overconfident behaviors of suppliers or retailers on the pricing and or-dering decisions of supply chain members are examined for a supply chain system consisting of a supplier that suffers from random yield and two competing retailers.By constructing a mean-variance ambiguity set of random yield and a mean-increasing variance-decreasing approach to portraying overconfidence behavior,we find that although the two overconfidence behaviors have significantly different effects on equilibrium,they both reduce the double marginaliza-tion effect.In addition,we find that(i)overprecision always benefits suppliers;(ii)retailers’ overprecision and overestimation behaviors do not necessarily hurt themselves;(iii)suppliers’ overprecision has no effect on supply chain members,while their overestimation hurts all par-ties in the supply chain;(iv)while competition among retailers can improve the performance of supply chain members,overconfidence may diminish this positive effect.Our findings alert policymakers that the overconfident behavior of particular supply chain members(i.e.,suppli-ers or retailers)may have quite different effects.Moreover,in the case of retailer competition,overconfidence can be a positive driver for supply chain members.