需求不確定下的零售商決策實驗-關鍵績效指標之資訊傳遞效果-

Effects of Key Performance Indicators on the Decision Bias in the Newsvendor Problem under Demand Uncertainty

侯楚薇(Chu-Wei Hou)

指導教授 ： 楊奕農

中原大學/商學院/國際貿易研究所/碩士(2009年)

https://doi.org/10.6840/CYCU.2009.00745

摘要

供應鏈零售商決策問題 (Newsvendor problem)的發生，起因於零售商必須在面對市場消費者隨機的需求下銷售商品，假設在銷售季節前零售商訂貨的機會只有一次，因此缺貨時零售商並沒有補貨的機會且其將面臨缺貨損失、而殘貨時將面臨殘貨損失，雖然供應鏈零售商之最適訂貨量在理論上已根據預期利潤之極大值的條件下推導出來，近年來部分關於供應鏈零售商決策的實驗即使在最簡單的實驗前提假設下，仍發現受試者做了次佳的訂貨量決策，意即受試者的平均訂貨量在低成本時有過低、在高成本時有過高於最適訂貨量的現象發生。
過去部分實驗提供給受試者的資訊傳遞之內容多數為每回合商品售價、進貨成本、消費者之需求量以及利潤、累積利潤、總成本、總收入、缺貨量，以及殘貨量，但由於理論上在預期利潤極大化的條件下推導出的最適訂貨量中，隱含在低成本時，受試者平均每增加一單位的訂貨量，其單位殘貨損失相較於缺貨損失為小，且由於殘貨與缺貨發生之機率皆相同，因此受試者此時應增加訂貨量，而同理當受試者面對高成本時，平均每增加一單位的訂貨量，其殘貨損失相較於缺貨損失為大，因此其訂貨量決策便應減少訂貨量已達成利潤極大，而過去的實驗所提供的主要資訊是缺貨量與殘貨量似乎容易誤導受試者採用平均需求量為訂貨量之策略，而忽略在零售價格相同但成本不同下，缺貨損失與殘貨損失同一單位高(低)於平均需求量會有不對稱損失，故此為本研究提供的資訊傳遞之內容除上述實驗已提供之外，再增加缺貨損失及殘貨損失之原因。
本研究設計的兩種實驗中包含三種試驗變因，第一項試驗變因為在實驗中當增加資訊傳遞之內容為上述缺貨與殘貨損失，本研究稱之為-「關鍵績效指標 (Key performance indicators) 」，觀察受試者平均訂貨量是否會為關鍵績效指標之影響而往理論上之最適訂貨量靠近。第二項變因為關鍵績效指標與非關鍵績效指標之資訊內容傳遞之「先後順序」，研究在皆提供兩張表資訊供受試者做決策參考時，受試者先看到哪一張表的資訊傳遞是否會影響其訂貨量決策。第三項變因為在已提供「關鍵績效指標」的前提下，再增加資訊傳遞之內容-本研究稱之為「經營績效考核報告」提供給受試者時，受試者之平均訂貨量是否會受其影響而往理論上之最適訂貨量靠近。
實驗結果發現在已提供與過去實驗相同的資訊傳遞之內容下，額外提供試驗變因中的「關鍵績效指標」供受試者做決策參考時，平均而言仍未顯著影響其訂貨量，而第二項變因-關鍵績效指標之「先後順序」發現受試者在16-30回合且為低成本時，關鍵績效指標出現之「先後順序」對平均訂貨量之影響有顯著差異。最後再提供第三項變因的「經營績效考核報告」發現，無論是否提供受試者各種關鍵績效指標及績效考核報告，相較於僅提供累積利潤的情況，皆不顯著影響受試者的訂貨量決策。

關鍵字

實驗 ； 供應鏈 ； 零售商決策問題

並列摘要

In the newsvendor problem, a retailer sells products under demand uncertainty. This paper assumes that the retailer only has one chance to order products before the selling season, the retailer may face two circumstances. One is if he orders too little, he loses sales and no further replenishments are possible, and the other one is if he orders too much, he disposes of the excess stock at a loss. Although it is well known the mathematical model maximizes the expected profit by determining the optimal order quantity, in past classical laboratory studies have showed some decision makers order too low when they should be high in the high profit product and vice versa.
This paper observes the “information” that past laboratory studies offer experiments and suspects that the “information” may the reason why this newsvendor problem occurs. What past laboratory studies offer information to experiments includes the price and cost of the product, order quantity, realizations of demand, profit, cumulative profit, the overage and shortage of product, the total cost and revenue, and the forgone sales value. But in theory of the expected profit-maximizing order quantity implies that because the marginal loss is less than the marginal profit in the high profit product, the decision maker should order more; because the marginal loss is more than the marginal profit in the low profit product, the decision maker should order less. This paper suspects that past laboratory studies offer the overage and shortage of product may lead experiments to order in the mean demand and ignore when the forgone salvage and profit was higher (lower) than the mean demand have asymmetric losses under the same price and the different cost. Therefore, that’s why this paper offering the information includes not only that past laboratory studies have owned, but also the forgone profit and the forgone salvage.
This paper evaluates the effect of the information with two laboratory designs with three treatments. One treatment includes the information of the forgone profit and the forgone salvage and this paper called “Key performance indicators”, another is the order that experiments view information tables, and the other one is besides the key performance indicator table, the experiments can view the additional content of the information, called “The performance appraisal of management report”, includes the mean forgone profit, the mean forgone salvage, the mean profit, and the appraisal of above. The purpose of two laboratory designs is to help the decision maker’s order strategy approach the optimal order quantity through the information transformation.
The results show that whether this paper offers not only the content of the information that past laboratory studies have offered but “Key performance indicators” or “The performance appraisal of management report” is not significant help the decision maker’s order strategy approach the optimal order quantity. And the order that experiments view information tables is not completely irrelevant to decision maker’s order strategy.

並列關鍵字

Experiment ； Newsvendor Model ； Supply Chain

參考文獻

Benzion, U., Cohen, Y., Peled, R., & Shavit, T. (2008). Decision-making and the newsvendor problem: An experimental study. The Journal of the Operational Research Society, 59(9), pp. 1281-87.

Bolton, G. E., & Katok, E. (2008). Learning by doing in the newsvendor problem: A laboratory investigation of the role of experience and feedback. Manufacturing & Service Operations Management, 10(3), pp. 519-38.

Bostian, A. J. A., Holt, C. A., & Smith, A. M. (2008). Newsvendor "pull-to-center" effect: Adaptive learning in a laboratory experiment. Manufacturing & Service Operations Management, 10(4), pp. 590-608.

Fisher, M., & Raman, A. (1996). Reducing the cost of demand uncertainty through accurate response to early sales. Operations Research, 44(1), pp. 87-99.

Holt, C. A., & Laury, S. K. (2002). Risk aversion and incentive effects. The American Economic Review, 92(5), pp.1644-55.