可轉換公司債發行動機之連續融資假說實證

Abstract

本研究之主旨為驗證Mayers的連續融資假說，研究樣本為在台灣證券交易所掛牌之公司於1994年至2003年的可轉換公司債發行宣告。連續融資假說提出公司發行可轉債之動機為節省發行成本與控制過度投資問題，因此本研究擬探討三個主題：第一，使用事件研究法檢驗股票報酬率對宣告發行可轉債的反應；第二，採用GARCH(1,1)模型檢驗股票報酬率之波動度是否因宣告發行可轉債而改變；第三，以橫斷面迴歸探討影響宣告效果的因素，檢驗發行公司的特徵是否符合連續融資假說之推論。

本研究實證顯示臺灣上市櫃公司宣告發行可轉債時，該公司普通股產生顯著負異常報酬率。另外，公司股價報酬率之波動度在宣告前後並無顯著不同。就橫斷面迴歸而言，實證顯示以下幾點： 一、集中度之係數不顯著，故此實證無法支持連續融資假說。 二、長期負債比率之係數為正且接近10%顯著水準下顯著，此實證與連續融資假說所認同之公司特徵一致，但亦同時符合Stein的後門權益假說。 三、公司規模之係數為負且在5%顯著水準下顯著，此實證與連續融資假說一致，且同時支持減輕資訊不對稱問題之觀點。 四、干擾變數於10%顯著水準下顯著，顯示事件期的干擾事件會影響宣告效果。 五、相對發行規模為正且在10%顯著水準下顯著，顯示相對發行規模越大者，則對市場之訊號強度越強。

綜上所述，本研究實證顯示樣本公司有兩特徵與連續融資假說一致，但此二特徵亦符合減輕資訊不對稱問題之觀點，而有關集中度之實證則不支持連續融資假說，因此臺灣公司發行可轉債之動機是否符合連續融資假說仍值得進一步研究。

The objective of this thesis is to re-examine the sequential-financing hypothesis by employing data of convertible bond issuances from firms listed at the Taiwan Stock Exchange from 1994 to 2003. The sequential-financing hypothesis, proposed by Myers, argues that convertible debt financing is motivated by a desire to minimize security issue costs and mitigate the Jensen’s overinvestment problem. There are three major topics. First, we use event-study methodology to explore stock price responses to the announcements of convertible debt offerings. Next, we use GARCH(1,1) specification to examine if the volatility of stock returns is changed due to the announcements. Third, we run cross-sectional regressions to investigate the factors which might influence the announcement effects and provide some evidence for the sequential-financing hypothesis.

We find that announcements of convertible debt offerings are, on average, associated with significantly negative abnormal returns. In addition, the pre-announcement volatility of stock returns is not significantly different from the post-announcement volatility. The results of the cross-sectional regressions are as follows. 1.The coefficient for the focus variable is insignificant, so our evidence does not support the sequential-financing hypothesis. 2.The coefficient for the long-term debt ratio is positive and significant at the 10% level. This result is consistent with the sequential-financing hypothesis as well as Stein’s backdoor-equity hypothesis. 3.The coefficient for the firm size variable is negative and significant at the 5% level. This evidence is consistent with not only the sequential-financing hypothesis but also the view of mitigating the informational asymmetry problem. 4.The coefficient for the contamination dummy is significant at the 10% level. This suggests that the contaminated events during the announcement period affect the announcement effects. 5.The relative issue size is positive and significant at the 10% level. This implies that the lager the relative size of issue, the stronger the strength of a security's signal to the market.

In summary, we show that two characteristics of our sample firms are consistent with the sequential-financing hypothesis, but they are also consistent with the view of mitigating the informational asymmetry problem. However, our evidence on the focus variable does not support the sequential-financing hypothesis. Therefore, our evidence overall shows mixed support for the sequential-financing hypothesis in explaining the motivation of using convertible debt financing for Taiwanese firms.