| Summary: | 碩士 === 國立成功大學 === 財務金融研究所 === 101 === This study aims to examine the announcement effect on convertible bond (CB) issues by using the methodology of Event study. In addition to examine the announcement effect on all CB issues in the research period, we divide the full sample into seven subsamples in terms of issuing place and issuing purpose and examine the announcement effect on these subsamples. We not only adopt the traditional one-factor pricing model, but also involve three kinds of multi-factor pricing models to estimate the expected stock returns, hoping to measure the abnormal returns in the announcement period more precisely. Besides, in addition to examine the announcement effect on original samples, we further involve the matched sample test for the purpose of controlling some unknown variables that might influence the research results.
Our results reveal that the announcement effect of CB is significantly negative in Taiwan Market. When all CB issues are further separated in terms of issuing place and purpose, we find significantly negative effect for European CB and CB issued with mixed purposes. CB issued with purpose to finance capital expenditure schemes exhibit significantly positive effect. As to CB issued with the purpose of refinancing previous debts, we find negative and extremely significant effect in the original sample. But this kind of announcement effect became insignificant in the matched sample test. Domestic CB and CB issued with purpose of supporting normal business operations or supporting possible M＆A programs all exhibit insignificant announcement effect. Finally, the cross-sectional regression analysis reveal that the variable issue maturity is positively related to abnormal returns, inversely, both relative size of issue and leverage ratio are negatively related to abnormal returns in the announcement period.
|
|---|
