考量企業存續機率最大化之商品交易決策

Abstract

本研究探討一個特殊的商品交易情境，在未來下游市場價格變動且不確定的情況下，經營者每期需以固定價格向上游採購固定數量，並依照當期的市場價格決定售出的數量以從中獲利。這個情境可以在不同的實務應用之中發現，一般是為了幫助社經較弱勢的供應鏈上游，例如農會產銷班會向農民以保價收購的方式採購農作物，再轉賣給下游。針對此情境我們建構一動態規劃(dynamic programming)最佳化模型，以提供最佳交易決策，但因為這種情境多出現在非營利組織，交易獲利的目的是要維持保價收購的機制，是要維護組織營運的永續，所以我們探討如何使經營存續機率(survival probability)最大化，而非傳統文獻中的利潤最大化。我們的模型除能提供短期的最佳交易決策，也能作為投入資金與營運產能設定等中長期決策的參考。

This paper determines an optimal trading strategy for a decision maker that needs to purchase a commodity from the upstream in a fixed quantity and price at each period, and resell to the downstream at a high market price. This trading problem is typically found in the non-profit organizations to support the poors in the industry upstream, and thus the objective is to survive, not the profits or business growth. We model this problem as a stochastic dynamic programming problem to maximize the survival probability; optimal strategy is determined and managerial insights are also discussed.