The Cross section of Expected Returns、Firm Attribute Variables and Momentum Effect in Taiwan

• Main Authors: HUANG,WEI-SHIN, 黃偉信
• Other Authors: ZHAN,JIA-CHANG
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/34929472760333531555

碩士 === 東海大學 === 企業管理學系碩士班 === 90 === The period of the study was January 1996 to December 2001, and the firms selected were all companies listed on the Taiwan Security Exchange. This paper analyzes the influence from past return and expected return on momentum. Then, using this concept to anlayze the difference of momentum effect among the firm attribute. There are two parts in this study. The first part is the momentum effect on past return. We use Fama and MacBeth（1973）cross-secetion regression model to analyze the momentum effect among total, high-tech industy and non high-tech samples. Then we condsider the consistent of past return, and test mometum effect again. The second part we use expected return that obtain from the average coefficients of regressor multiply past return to form momentum portfolio. According to frim attribute variables, we separate samples into subsamples. Then test the difference of momentum effect among these subsamples. As for the influence of past return on momentum effect：We find that the momentum effect exist in total and high-tech subsamples roughly. But the momentum effect on non high-tech subsamples is weaker. The consistent variable of past return has less impact on return. Therefore the momentum effect still exist in total and high-tech subsamples. And the momentum effect on non high-tech subsamples is still weaker. The momentum effect on high-tech is stronger than non high-tech. As for the influence of expected return on momentum effect：We use expected return to form 5 portfolio. Then we perform momentum strategies that buy the stocks with best expected return and sell the stocks with worse expected return. The return of the momentum strategies on toal, high-tech and non high-tech are positive. And the expected return model of non high-tech has stronger momentum effect than high-tech. As for the influence of firm attribute on the momentum effect of expected return model，Size：Small cap stocks of total samples has weaker momeutm effect than large cap stocks. But there are stronger mometum effect on small cap stocks of high-tech and non high-tech. BE/ME：Low BE/ME stocks of total samples has stronger mometum effect. But low BE/ME stocks has weaker mometum effect in high-tech and non high-tech. IF we condsider both Size and BE/ME：small cap and low BE/ME stocks have the maximum momentum effect. In samll cap subsmples, low BE/ME stocks in total and high-tech samples has stronger mometunm effect than high BE/ME stocks. And the difference of mometum effect between low and high BE/ME stocks in high-tech is stronger than total samples. In large cap subsmaples, low BE/ME stocks in non high-tech samples has stronger momentum effect than high BE/ME stocks.Turnover ratio：High tunover ratio stocks in total and high-tech subsamples have weaker mometum effect. But high turnover ratio stocks of non high-tech subsamples have stronger mometum effect.