Exploring the Effects of Sharing Information on Expected Cost under the VMI Model

探討VMI模式下之資訊分享對預期成本的影響

本文藉由VMI 模式，探討供應商與零售商之資訊分享程度對供應商預期成本的影響。本文假設建立顧客需求是ARMA(1, 2) 的時間序列，不同程度的資訊分享為：
完全沒有資訊分享、部分資訊分享及完全資訊分享等模式，對供應商存貨運送量變異數、最適運送量上限（optimal delivery-up-to level）及預期成本的影響。同時，在不同的資訊分享程度下，供應商存貨運送量變異數、最適運送量上限及預期成本對前置時間（l）、當期銷售與前期銷售的相關係數（ρ）、前一期和二期的誤差對顧客當期銷售的影響程度（θ1 與θ2）等四個參數，來進行敏感度分析，並利用數值模擬加以驗證。其結果有下列主要發現：(1) 分享需求資訊不僅會讓供應商運送量穩定，亦降低供應商的最適運送量上限及預期成本。(2) 在每一種程度資訊分享下，各參數對供應商之運送量變異數、最適運送量上限及預期成本均有正向的影響。(3) 資訊分享程度與所有參數對於降低預期成本都有加乘效果。其中又以l 與ρ的影響最顯著。換言之，當l愈小或是ρ愈大時，資訊分享愈多對降低預期成本效果最為顯著。

This paper was to discuss the effects of information sharing between the vendors and the retailers. Assuming the underlying demand process was followed by auto-regressive moving average ARMA(1,2) in vendor managed inventory (VMI). Considering different information-sharing levels, such as no information sharing, partial information sharing, and full information sharing, with the factors of the lead time, the correlation coefficient between demand prediction and current demand, and the correlation coefficient between prediction error terms and demand, this research was to find the importance of single and multiple effects on the vendors’ delivery variance, delivery-up-to level, and expected cost. Through the sensitivity analysis, the effects from the single and multiple factors at different information-sharing levels could be observed. Moreover, the numerical simulation examined the findings as follows. (1)Information sharing stabilized delivery variance and delivery-up-to level as well as reduced expected cost. (2)At any information sharing level, all factors influenced the vendor’s delivery variance, delivery-up-to level, and expected costs. (3)At any information sharing, the factors of a smaller lead time and a larger coefficient between demand prediction and current demand had more significant effects on minimizing inventory and cost.