| Summary: | 碩士 === 佛光人文社會學院 === 經濟學研究所 === 92 === Abstract
This study attempts to forecast the abnormal return ratios for the stock issued by TSE companies via the open market information. Therefore, this study incorporates the abnormal return ratios of stock price as dependent variables and the accounting ratios as independent variables; and selects the significant independent variables via step-wise regression, thereby constructing a model to forecast the abnormal return for stock prices. This study has attempted to compare accounting ratios and forecast future price changes through both models. As far as independent variable selection is concerned, the percentage of the changes between long-term liabilities and equity, percentage of the changes of net income, percentage of the changes of sales and inventory and the ratio of quarter-end net value per share are most capable of forecasting the abnormal return. As to the comparison between forecast methods, this study has concluded that OLS model was superior to Logit model; and OLS model outperforms Logit with respect to accuracy and return ratios.
|
|---|
