Why Do Firms Issue Convertible Bonds Evidence From The Field PDF Download

This thesis makes three main contributions to the literature on convertible bond financing. First, we provide a new theoretical explanation for convertible bond financing. Unlike the existing theory, our new theory provides a rationale for the issuance of both callable and non-callable convertible bonds. We also undertake empirical tests of the implications of the new theory and find that the new theory is supported by the empirical evidence. Second, we empirically examine the way in which firms choose the design of convertible bonds and investigate the effect of financial constraints on the firms' convertible design decision. Consistent with our new theory, we find that the design of convertible bonds is influenced by both adverse selection costs and financial distress costs. Moreover, we find that the design of convertible bonds for relatively constrained firms is determined in a different manner from the design of convertible bonds for relatively unconstrained firms. Our findings suggest that taking into account the effect of financial constraints is important in the understanding of convertible design decisions. To the best of our knowledge, our study is the first to document the effect of financial constraints on choice of convertible design. Third, we empirically examine two alternative explanations for the late call of a convertible bond: the "optimal" call theory of Butler (2002) and the financial distress costs theory of Jaffee and Shleifer (1990). In contrast to the existing evidence reported in Altintig and Butler (2005), we find that the observed late calls cannot be explained by the effect of the notice period as incorporated in the optimal call theory of Butler (2002). The observed conversion premium is much higher than Butler's optimal conversion premium. On the other hand, we find strong empirical support for the financial distress costs theory. Firms do not make a conversion-forcing call until the conversion premium is large enough to avoid a failed conversion, which could give rise to financial distress. We find that by the time a call is made, the probability of failed conversion is very small and the cross-sectional variation in the conversion premium is mainly explained by potential distress costs.