台股報酬率的預測模式的分析與比較

• Main Author: 鄭建成
• Other Authors: 張延彰
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/12017000744180384389

碩士 === 國立新竹教育大學 === 應用數學系碩士班 === 104 === ABSTRACT The major problem which investors the most care about is whether investment could make money or not. For investors, the explanatory variables whether can have a good prediction effectiveness on the returns on Taiwan stock is clearly even more important. In this study, the observed values were collected from 271 Taiwan firms from 1996 to 2013. A threshold regression model proposed by Hansen(1999) were used to test the threshold effect and to investigate whether earnings per share within a certain range, total assets growth rate, operating profit per share and net growth rate on the effect of Taiwan stock returns, and then the results were compared with the traditional regression model. The empirical result shows total assets growth rate and operating profit per share are statistically significant and has a positive relationship with stock returns in both traditional regression and threshold regression in models of 2012 and 2013. Net growth rate is statistically significant and has a positive relationship with stock returns in both traditional regression and threshold regression in model of 2014. Net growth rate is statistically significant and has a positive relationship with stock returns in both traditional regression and threshold regression in model of 2014. Earnings per share (EPS) is statistically significant and has a positive relationship with stock returns only in traditional regression in models of 2012 and 2014.