BAYESIAN ESTIMATED MODEL OF STRUCTURAL MARKET SHARE UNDER INCOMPLETE INFORMATION

• Main Authors: HSU, TSU-HSIANG, 徐子祥
• Other Authors: LI-CHUNG JEN
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/89588148307703473917

碩士 === 國立臺灣大學 === 國際企業學系研究所 === 86 === In practice, it is always a myth that getting correct information of marke t share. Although there are a lot of strategy provided to create competitive a dvantage or core competence, managers still can*t plan the right strategy. The reasons are that many strategies decided are based on the company*s market po sition: market share. However, most managers obtain the information by guessin g. Therefore, it motivates us to do this research, and think of the question: "How to update structural market share permaBasically, our research is built b y bayesian approach and hypothesize that managers only have one-period informa tion of market share and incomplete current data. The model is divided into si x stages, and the first stage is divided into six steps. In the first stage, w e update current market-share data and from the second to the sixth stage, we simulate data to converge the estimates in bayesian structure. Finally, use th e concept of efficiency to prove that bayesian approach is better than na*ve o ne. In shoOur contribution of research are in the following aspects.1. Manage the company*s market position dynamically in short term, instead of the tradi tional concept of "ABC management". 2. Make a marriage between high benefit and low cost by computer language and software package, because you can get th ese information in a few minutes. 3. Complete the preceding literature about market structure and make the strategy management more practical.4. Dimension is free, and it can be applied to some short-term issue, such as marketing of election.