Where Are They Now? An Analysis of the Life Cycle of Convertible Bonds

• Date: 01 August 2013
• DOI: http://doi.org/10.1111/fire.12012
• Published date: 01 August 2013

The Financial Review 48 (2013) 489–509

Where Are They Now? An Analysis of the

Life Cycle of Convertible Bonds

Timo Korkeamaki∗

Hanken School of Economics, Bank of Finland

Timothy B. Michael

University of Houston - Clear Lake

Abstract

We study the determinants of the life convertible bonds’ life span issued between 1980

and 1998. About 60% of the bonds survive either to a call or to their maturity. The issuers of

the remaining bonds are delisted during the life of their bonds. Calls and delistings shorten the

average life span of convertibles from the original 17 years to an effective life span of only

seven years. Issuer’s post-issuance performance and investment behavior affect the effective

life of convertibles. Our results support the sequential financing hypothesis, as bonds issued

by firms with speedier investment schedules have shorter life spans.

Keywords: convertible bonds, callability

JEL Classifications: G13, G30, G32

1. Introduction

Two points in the life cycle of convertible bonds have received lively research

interest, namely their birth in issuance, and their “death” through a call. However,

∗Correspondingauthor: Hanken School of Economics, Bank of Finland, P.O. Box 479, FI-00101 Helsinki,

Finland; Phone: (358) 40 3521 308, Fax: (358) 9 431 33 393; E-mail: timo.korkeamaki@hanken.fi.

C2013, The Eastern Finance Association 489

490 T. Korkeamakiand T. B. Michael/The Financial Review 48 (2013) 489–509

besides recent studies by King and Mauer (2012), and Grundy and Verwijmeren

(2012), few papers consider the entire life cycle of convertible bonds and the joint

determinants of their tenure. We aim to fill that gap with this study. We track the life

of a convertible bond sample that was initially issued between 1980 and 1998, and

study determinants of the effective life of those bonds.

Our study has two main goals. First, we seek to provide information concerning

factors that affect the end of life for these bonds. How common is it for them to be

called versus how many of them mature as bonds? How many of them cease to exist

for other reasons, such as mergers and firm failures? Second, we observe certain

design features in convertible bonds, and relate those design features to existing

literature that motivates their usage. Features such as conversion premium, coupon

rate, maturity,callability and call protection all come with potential costs and benefits.

We seek to evaluate how the use of these design features is related to the subsequent

behavior of the issuing firms. The existing studies on call policy of convertiblebonds

focus on factors such as yield differential and conversion premium as determinants

of call policy. Our focus is different, as we measure the calendar time that the bonds

remain active beyond their call protection period. Our results should be of interest

to convertible bond investors such as commercial banks, whose ability to remain

invested in the issue is limited after a forced conversionto equity, and to hedge funds

holding a convertible bond as one leg in a convertible arbitrage portfolio. For both

groups, an unexpected shortening of the life of a convertible bond would present

unexpected transaction costs.

Our results indicate that many convertible bond issuers are delisted during the

life of the convertible. Out of our 955 sample bonds, only 60% remain activeto either

a call announcement or to original maturity. Over 65% of the delistings, counting

for about 26% of the original sample, occur due to mergers, while about 12% of

the original sample is either liquidated or dropped by the stock exchange. Both

calls and delistings have the effect of significantly shortening the effective life span

of convertible bonds. For the overall sample, the effective life span falls from an

original time to maturity of about 17 years to a realized life span of about seven

years. In observing the connection between the design of convertible bonds and their

eventual life cycle, we exclude bonds whose issuers are delisted, as we assume that

mergers and bankruptcies are not foreseen at the time of the issuance.

We find that features in the original bond design have a limited effect on the

length of the bond’s life. Bonds with longer original maturities survive longer as

bonds, and bonds with higher (original) yield advantage tend to be called sooner.

Call protection type and conversion premium fail to affect the bonds’ life cycle in

a consistent manner. Issuer’s performance after issuance, however, has a significant

effect on the life span of the convertible. Higher post-issuance stock returns shorten

the life of convertibles, and improvements in an issuer’s credit rating have a similar

effect.

Most interestingly, firms with high capital expenditures subsequent to issuance

call their bonds sooner. That finding is consistent with the sequential financing