| Summary: | 碩士 === 國立臺灣大學 === 工業工程學研究所 === 94 === Many high-tech manufacturing industries can be characterized by intense competition, high capacity investment cost, rapid technology advancement and uncertain demand. These characteristics tend to weed out weaker firms and drive an industry toward an oligopoly as it passes the embryonic stage. Examples of such industries include semiconductor manufacturing and liquid crystal display manufacturing. In an oligopoly, a firm can not afford to make capacity decisions based on cost-benefit analysis alone; potential actions of the opponents must also be taken into consideration.
In this paper, a gaming analysis method is described for designing capacity strategies in oligopoly competition with high irreversible investment cost and high demand uncertainty. This paper focuses on semiconductor manufacturing. Industry data of average selling price and manufacturing costs are utilized in the analysis. An optimal capacity model that a firm might undertake while disregarding the opponent’s action is first described. This model is based on tradeoff between over-capacity and under-capacity when the demand is highly uncertain and representable as a Brownian motion process and is used to determine myopic capacity decision of individual firms. In the second part of the paper, mathematical formulas are derived for modeling the behavior of the market leader and follower. Finally, a gaming analysis method is presented to analyze the gaming interaction between the players of the game and to compare the outcomes of aggressive and conservative strategies by the leader and the follower. The paper illustrates two results: both the players will take aggressive strategies; the follower cannot keep long-term success for its revenue appreciably lower than the leader.
|
|---|
