The Information Contents of Accounting Earnings-the Bits of Information Analysis

• Main Authors: Yu-Shen Huang, 黃于慎
• Other Authors: Jin-Jou Dai
• Format: Others
• Language: zh-TW
• Published: 2001
• Online Access: http://ndltd.ncl.edu.tw/handle/20458031346779364186

碩士 === 朝陽科技大學 === 財務金融系碩士班 === 89 === The study selected 207 companies listed in the Taiwan Stock Exchange to investigate the relations between stock returns and operating income, non-operating income, and scale, and the relative importance of the later three variables during the period of 1994:1-2000:3. We also investigated the difference of earning information structure of the companies in each industry. The approach of Theil’s bits of information was applied. The Bits of Information of each indepent variable presents the relative importance in explaining the variation of stock returns. The method can depict much more realistic picture than the regressive coefficient of OLS method does. According to the Regression Analysis, operating income, non-operating income, scale were statistically significant explanatory variables for stock returns. The Bit of Information Analysis and the Information Inequality Across Regressors Analysis, showed that operating income dominated the information of stock returns in most companies. And non-operating income and scale were the second and third important variables. Across all industries, the informational content of non-operating income was the highest for the Cable industry, and the lowest for the Paper industry. The informational content of scale was the highest for the Paper industry and the lowest for the Cable industry. The informational content of operating income was the highest for the Tourist industry and the lowest for the Plastic industry. Across all industries, Cable industry and Steel industry are with the higher dispersion of bits of information, Construction industry and Car industry are with the lower dispersion of bits of information.