管理學院: 財務金融學研究所指導教授: 胡星陽吳穎昌Wu, Ying-ChangYing-ChangWu2017-03-032018-07-092017-03-032018-07-092016http://ntur.lib.ntu.edu.tw//handle/246246/274153本文的目的是檢定股價操縱的出現機率受到哪些市場交易面、股權結構、以及公司財務之因素影響。使用邏輯斯(logistic regression)迴歸分析，並得到計算股價受到操縱機率的模型。使用我國以經過法院裁定且定罪的股價操縱案件，並另外依照操縱案個股的市值、週轉率、股價淨值比的馬氏距離綜合指標挑選與操縱個案相近的個股作為對照組。以逐步迴歸找出適合的模型，再經過移除解釋力不足之自變數的步驟，基本上，股價波動度越高，越容易受到操縱；董監和外資持股比越高，受到操縱的機率越低；個股Beta值越高，受到操縱的機率越低；週轉率越高，受到操縱的機率也越高；負債比越高的公司，也越容易受到操縱。The purpose of this study is to test what kind of market, ownership, and financial factors will increase or decrease the probability of illegal stock price manipulation. Logistic regression is employed to model the probability of stock price manipulation. The samples this study uses is the illegal manipulation cases in Taiwan, which had been accused, as manipulation samples. The benchmark sample, which never involved price manipulation and have similar market value, turnover and price-to-book ratio, were selected by Mahalonbix distance. The model was built by a loop step of stepwise logistic regression and removing independent variables which is lack of explanatory ability. Basically, the stocks with larger return volatility, higher Beta, higher turnover, higher debt ratio and lower board and foreign investor holding will have a larger probability of price manipulation.1184976 bytesapplication/pdf論文公開時間: 2017/7/25論文使用權限: 同意有償授權(權利金給回饋本人)股價操縱股價操縱案之決定因素股價操縱之機率拉高倒貨邏輯斯迴歸模型挑選Stock price manipulationfactors of stock price manipulationprobability of stock price manipulationpump-and-dumplogistic regression model selectionthesis10.6342/NTU201600617http://ntur.lib.ntu.edu.tw/bitstream/246246/274153/1/ntu-105-R03723046-1.pdf