Elsevier

Journal of Corporate Finance

Volume 24, February 2014, Pages 149-157
Journal of Corporate Finance

Why are conversion-forcing call announcements associated with negative wealth effects?

https://doi.org/10.1016/j.jcorpfin.2013.10.003Get rights and content

Highlights

  • We find negative stock price responses to predictable convertible calls.

  • Signaling rationales cannot explain this phenomenon.

  • Greater hedge fund involvement leads to less short selling in response to the call.

  • Greater hedge fund involvement is associated with a reduced stock price reaction.

  • Price pressure and not signaling underlies negative call announcement effects.

Abstract

We analyze call announcement returns taking into account two recent developments in the convertible bond market: the inclusion of dividend protection clauses in convertibles' terms, and the high fraction of convertible issues purchased by hedge funds. Calls of dividend-protected convertible bonds are predictable, yet we still observe a negative stock price reaction that cannot be explained by signaling. Greater hedge fund involvement prior to a call means less short selling in response to the call and we document a reduced price reaction. We conclude that price pressure and not signaling underlies the negative announcement effect of convertible bond calls.

Introduction

Companies that issue convertible bonds often include a call provision in the bond's terms and conditions. A call provision gives the issuing company the right to redeem the bond before its maturity date at a pre-specified call price. Survey evidence in Graham and Harvey (2001) suggests that the ability to call and force conversion is important, since the most important reason given for issuing convertibles is that they are seen as an inexpensive way to issue delayed common stock.

A number of empirical studies find that companies that call their convertible bonds experience a negative stock price reaction. A summary of previous studies in Table 1 shows that the estimated effect for in-the-money calls varies between − 0.58% and − 2.13% depending on the sample period, the methodology, and sub-sample investigated.1

The two principal rationales for the negative announcement effect are the signaling model of Harris and Raviv (1985) and the price pressure effect investigated in Mazzeo and Moore (1992) and traced to short-selling in Bechmann (2004). Harris and Raviv (1985) model an equilibrium in which managers truthfully signal by calling convertibles only when their private information is unfavorable.2 A decision to call is then rationally perceived by investors as bad news and induces a negative stock price reaction, while forgoing an opportunity to force conversion is a signal of favorable inside information.

Bechmann (2004) argues that price declines around convertible calls can be explained by short sellers entering the market at the time of a call. Convertible hedge desks try to lock in arbitrage profits by buying convertibles from long-only investors and shorting to hedge their equity risk. Underwriters also hedge their exposure by shorting stock. Bechmann (2004) concludes that the short sales triggered by call announcements create substantial price pressure.

In this paper we study two recent developments in the convertible bond market that a priori might be argued to have a mitigating effect on the negative abnormal returns accompanying call announcements. Both of these developments allow us to provide new evidence on the relevance of the signaling and the price pressure hypothesis. Additional evidence is important as the conclusions of prior studies have been mixed. For example, Datta et al. (2003) observe short-term price reversals after convertible bond calls, but still conclude that convertible bonds convey bad news implicit in their long-term stock price underperformance. On the other hand, Bechmann et al. (2014-in this issue) use high-frequency data to compare the announcement effects of in-the-money and out-of-the-money calls and find evidence for the price pressure hypothesis as the stock market reaction to in-the-money calls is associated with trade order imbalances.

The first development in the convertible bond market that we examine is that most convertible bonds issued today are dividend-protected. When a convertible is dividend-protected, the conversion ratio increases in the event of a dividend payment so as to hold constant the market value of the shares into which the bond is convertible. Grundy and Verwijmeren (2013) document that prior to 2003 convertible bonds were generally not protected against dividend payments while from 2003 on dividend protection of convertible bonds has become the norm. More than 95% of the convertible bonds issued since 2003 are dividend-protected. As a result, rationales for call delay that crucially depend on bondholders forgoing dividends if they do not convert are nullified when the bond is dividend-protected.3 Grundy and Verwijmeren (2013) document the empirical absence of call delay for dividend-protected convertibles. The absence of call delay has made calls of protected bonds a near perfectly predictable function of the share price. Because of the predictability of these calls, the Harris–Raviv signaling model cannot be used to explain any negative price response that accompanies these calls since perfectly predictable announcements cannot provide signals.

The second potentially mitigating effect on the price reaction to call announcements comes about because a major part of the buy-side of the convertible bond market has changed from long-only investors to hedge fund investors.4 Convertible bond funds hedge their equity exposure by shorting the underlying stock and, as a result, there is already substantial short selling before any call. The larger the fraction of the bond already held by convertible arbitrageurs at the time of the call, the smaller the additional short-selling in response to a call and the smaller any short-selling induced price pressure.

Comparing calls of dividend-protected and non-dividend-protected convertible bonds, we find that during the period 2000 through 2011 conversion-forcing call announcements are associated with similar negative stock price reactions for both groups. There is no significant difference between the price reaction to the predictable call of a dividend-protected convertible and the price reaction to the less predictable call of one of its unprotected cousins and we therefore conclude that signaling rationales cannot explain the price reaction to convertible calls. This leaves the short-selling induced price pressure examined in Bechmann (2004) as the potential explanation of the negative price response to announcements of convertible bond calls. We use data on hedge fund involvement in convertible issues to show that greater hedge fund involvement prior to the call relates to less negative call announcement effects. Hedge funds have already established short positions before the call. Therefore, our results are consistent with negative call announcement effects being a reflection of increased short selling-induced price pressure in response to the call rather than a response to the release of bad inside information.

Our findings relate to the general price pressure literature, as price pressure is not only associated with calls of convertible securities, but also with offerings of convertibles (e.g., Loncarski et al., 2009), mergers and acquisitions (e.g., Mitchell et al., 2004), and index revisions (e.g., Harris and Gurel, 1986). Most notably, a price pressure hypothesis and a signaling hypothesis also make up two alternative hypotheses in studies on convertible offerings and mergers and acquisitions. By controlling for dividend protection and hedge fund involvement, the tests in this paper are able to distinguish between signaling and price pressure as explanators of the negative announcement effect of convertible bond calls, and we conclude that price pressure is the main determinant of the negative announcement effect.

The remainder of the paper is organized as follows. Section 2 sets out the hypotheses and Section 3 describes the data. The empirical results are reported in 4 Results, 5 Conclusions contains our conclusions.

Section snippets

Hypotheses

In the Harris–Raviv signaling model insiders who expect the firm to do well will also expect voluntary conversion in the future. A decision to call and force conversion today is then a signal that insiders are not convinced future voluntary conversion will occur. The firm therefore calls to deprive the convertible holders of the insurance and income advantage they would otherwise enjoy as bondholders. In an equilibrium in which insiders care about both current and future stock prices, firms

Data description

We examine a set of 98 convertibles (identified from SDC) that were issued after January 1, 2000 and called in-the-money prior to June 1, 2011. We obtain this set by following the selection procedure in Grundy and Verwijmeren (2013). The set contains 62 non-dividend-protected bonds and 36 dividend-protected bonds.

As an example of dividend protection consider the 2004 convertible bond issue by Matria Healthcare. The prospectus states (page 31): “Subject to the terms of the indenture, we will

Results

In the calculation of the stock price reaction to call announcements, announcement dates are determined from Factiva. If a call is announced after the closing of the market, we take the next day as day 0. Our event window is [− 1, 0], which is in line with Bechmann (2004) and many other studies listed in Table 1. We focus on abnormal returns measured relative to a market model whose parameters are estimated over the post-announcement window of [+ 80, + 200] days. We are able to obtain the abnormal

Conclusions

Two important developments in the convertible bond market have the potential to shed new light on the cause of the negative announcement effect of convertible bond calls. The first development is the incorporation of provisions into convertibles' terms and conditions that protect the conversion value of convertible bonds against dividend payments. Dividend protection nullifies dividend-related rationales for call delay and makes the optimal call policy a predictable function of the stock price.

Acknowledgments

We thank an anonymous referee, Jo Danbolt, Frank Hong Liu, Geoffrey Poitras, Yulia Veld-Merkoulova, and the participants at a seminar at the University of Glasgow and at the Western Economic International Association Conference in San Francisco (June 2012) for their helpful comments and suggestions.

References (31)

  • K.G. Nyborg

    Convertible debt as delayed equity: forced versus voluntary conversion and the information role of call policy

    J. Financ. Intermed.

    (1995)
  • A.R. Ofer et al.

    Convertible call policies: an empirical analysis of an information-signaling hypothesis

    J. Financ. Econ.

    (1987)
  • B. Arshanapalli et al.

    New evidence on the market impact of convertible bond issues in the U.S.

    Financ. Lett.

    (2005)
  • P. Asquith et al.

    Convertible debt: corporate call policy and voluntary conversion

    J. Financ.

    (1991)
  • I.E. Brick et al.

    Stock price response to calls of convertible bonds: still a puzzle?

    Financ. Manag.

    (2007)
  • Cited by (6)

    • Arbitrage vs. informed short selling: Evidence from convertible bond issuers

      2020, Journal of Corporate Finance
      Citation Excerpt :

      Prior literature establishes that short selling may have an effect on announcement returns for corporate financial events. For example, price pressure due to short selling affects returns around merger announcements [Mitchell et al., 2004, and Liu and Wu, 2014], non-overnight seasoned equity offerings [Gustafson, 2018], and convertible bond call announcements [Bechmann, 2004; Bechmann et al., 2014, and Grundy et al., 2014]. We provide further evidence on the importance for academic researchers studying the announcement effects of corporate financial decisions to adjust announcement date abnormal returns for price-pressure-related declines due to arbitrage short selling.

    • In- and out-of-the-money convertible bond calls: Signaling or price pressureα

      2014, Journal of Corporate Finance
      Citation Excerpt :

      In conventional event studies, returns are aligned in event time such that the announcement of the call happens at day 0. The announcement effect for firm i is then calculated as the sum of the market adjusted returns for (at least) day 0 and day 1 (see, e.g., the overview of call announcement studies in Grundy et al., 2014-in this issue). The lack of data on the exact announcement time is the standard reason for including both day 0 and day 1 returns.

    • Convertible bond financing

      2014, Journal of Corporate Finance
      Citation Excerpt :

      The positive effects for out-of-the-money calls are better explained by signaling effects. Grundy et al. (2014--in this issue) exploit two recent developments in the convertible bond market to study in-the-money calls: the inclusion of dividend protection in the terms of the convertible bonds and the increased participation of hedge funds as buyers of convertible bonds. These developments create cross-sectional differences that allow the authors to conclude that the negative announcement effects associated with in-the-money calls are better explained by price pressure.

    View full text