預約產能定價之研究

Abstract

本論文替ㄧ個擁有多客戶、資本密集的半導體代工廠，建立一預約產能問題的模型並解決預約產能定價之決策。為了使產能利用滿載，同時適當地分配產能予客戶，代工廠會要求客戶為預備產能付預約金。預約金定價是一個取捨平衡的問題，太高的預約金會造成客戶傾向保守而導致較少的預約量，相反地，太低的預約金則引發較多的預約量。然而，在決定預約量的當下，未來發生的需求數量卻是未知的隨機變數。本研究發展了兩個決策模型，一個是買方的決策，另一個則是賣方的決策。在某個特定的預約價格之下，買方的決策模型會為他決定一個最佳的預約數量，預約價格與預約數量之間的功能性關聯可因而決定。利用這樣的功能性關聯，賣方的決策模型即可以得到一個使得賣方利潤最佳化的預約產能價格。

IThis thesis formulates and solves a pricing decision for booking the capacity of an IC foundry. IC foundry, a capital-intensive factory, has many customers. In order to fully utilize capacity and appropriately allocate capacity to customers, IC foundries may ask customers to pay a booking fee for reserving capacity. Pricing the booking fee is a trade-off issue. Higher booking fee would make customers more conservative and induce a lower booking quantity. Conversely, lower booking fee would induce a higher booking capacity. However, the future demand quantity, at the time point of making booking decision, is a random variable. This research develops two decision models, one for capacity buyers and the other for the capacity seller. For a particular booking price, the first decision model can be used to determine the booking quantity which optimal for buyer. The functional relationship between booking price and booking quantity can therefore be determined. Using such a functional relationship, the second decision model can compute the booking price that is optimal for the capacity seller.