本論文主要探討零售商銷售季節性商品之動態定價議題。由於季節性商品的市場改變快速，且存貨的前置時間長，無法在短銷售期內作彈性的補貨動作，故零售商必須在期初即決定存貨訂購數量的多寡。因此如何在有限的銷售期間有效的利用價格機制是很重要的，較佳的價格機制可以改變消費者對產品的需求或促進顧客的購買意願以增加產品售出的機率。簡單的說，如何從消費者剩餘中來獲得最大的銷售利益，是目前各企業亟欲追求的目標。本研究以Bitran and Mondschein（1997）所提出的離散時間型定價策略（Periodic Pricing Review Policy）為主軸，首先會對動態定價而形成的價格組合做一空間性概念的陳述，將價格資料視覺化，使整個價格趨勢及變動能一目了然，有助於了解各個主要因子之間相互影響的關係，並且延伸至價格曲面之估計。除此之外，本研究考慮到傳統動態規劃求解方式的繁雜與耗費時間，而發展出一個以估計迴歸曲線的方式來代替傳統最佳化計算價格的演算法，以追求更簡化便捷的決策過程，並以達到近似於最佳的總期望利潤為目的，提供管理者一個方便且能實用在做定價決策過程時的參考與依據。 In this thesis we study optimal pricing strategies for seasonal products in retailing. Because the market of seasonal products is changing fast and the lead time of inventory to replenish is long, retailers have to decide the number of order quantity at the beginning of planning horizon. Therefore, how to use the pricing policies effectively during a short sale season is important. Superior pricing policies could have positive influence on products’ demand and the probability of purchasing. Briefly, getting maximum revenue from consumer surplus is the goal that most enterprises would like to pursue. In this research, we use the Periodic Pricing Review Policy from Bitran and Mondschein (1997) as a benchmark. We first present a concept of three-dimensional of price sets from dynamic programming and try to visualize the optimal price data to see an entire pricing trend and variation. It’s useful to know the interactions and relationship among main factors. Furthermore, we develop a heuristic procedure for finding near-optimal prices with regression to replace the traditional optimal dynamic pricing method which is complicate and time-consuming. The benefits of the heuristic are more convenient and simpler to implement. Moreover, our heuristic provides a near optimal expected profit at the same time. The aim of this paper, in sum, is to provide a practical reference material and basis of pricing decision process for managers.
