RDP 2003-02: Do Collective Action Clauses Influence Bond Yields? New Evidence from Emerging Markets 1. Introduction

There has recently been much debate over crisis prevention and resolution. Although much has been done on crisis prevention, it may be inevitable that there will be occasional sovereign debt crises such as those experienced in recent years by Argentina, Ecuador, Russia and Ukraine. Crises such as these are usually extremely costly for both investors and the citizens of the affected countries, and have prompted debate over what changes could be made to the way that financial markets and the international community deal with sovereign debt crises. However, there is little agreement over what reforms are desirable. The debate appears to hinge on different views as to how to meet the goal of facilitating the speedy restructuring of unsustainable debt burdens in a way that minimises both the dislocation to the affected country and the loss of value to creditors, while ensuring that borrowers that can repay their obligations do so.

Official proposals have focused on either a contractual or a statutory approach to the problem. A contractual approach would involve borrowers placing clauses in their debt contracts that would spell out the various procedures that might be followed in the event that debt servicing problems arise, including procedures for qualified majorities of bondholders to change the payment terms of bonds (see Taylor (2002)). These clauses would draw heavily on the collective action clauses (CACs) that already exist in many bonds issued in the Euromarket. Proposals for a statutory approach are continuing to evolve as the International Monetary Fund (IMF) develops its proposal for a sovereign debt restructuring mechanism (SDRM).^[1] The purpose of the SDRM would be to move the framework for dealing with sovereign debt problems closer to the bankruptcy frameworks that exist within national economies. An important aspect of these two approaches is that neither would force restructuring on creditors or borrowers: they would simply provide a framework by which a super-majority (say 75 per cent) of creditors could come together and agree to change the payment terms if they viewed it as being in their collective interest.

Despite support for both the contractual and statutory approaches from the international organisations and major industrial countries, there has been little support for either approach from either investors or emerging market issuers. Investors (and the financial markets more generally) were for a long time strongly opposed to a contractual solution (wider use of collective action clauses), arguing that the status quo of voluntary exchange offers for dealing with debt problems was sufficiently flexible to deal with debt servicing problems.^[2] Although private sector organisations have recently been more supportive of a contractual approach and have drafted a set of model CACs,^[3] the strength of this support is unclear, with some arguing that they are unwilling to discuss moving toward a contractual approach as long as the statutory approach remains on the table.

Emerging market borrowers have also tended to oppose most reforms, with some citing concern that reforms would increase their borrowing costs. In many respects the lack of support from borrowing countries appears to be related to opposition from financial markets. Indeed, Boorman (2003) notes that despite the absence of credible evidence that collective action clauses have increased spreads ‘the private sector seems to be going around to emerging market countries and trying to scare the hell out of them about the fact that either the use of collective action clauses or the SDRM will lead to an increase in spreads’. An important recent development, however, was the inclusion of CACs in a global bond issue by Mexico on 26 February 2003, which we discuss further in the conclusion.^[4]

This paper makes no judgment on the relative benefits of a contractual versus a statutory approach.^[5] Instead, it presents new evidence, based on market prices, about the way that markets have viewed contractual clauses that can facilitate debt restructuring. In particular, since a significant proportion of the outstanding stock of sovereign bonds – from mature as well as emerging market countries – already includes CACs, we are able to analyse the pricing of bonds with and without CACs.

This paper adds to the literature on the pricing of CACs in two regards. First, it provides an additional test as to the historical pricing of CACs. In particular, previous empirical evidence has suggested that the use of CACs in a bond issue did not increase the cost of borrowing for that particular bond.^[6] Previous evidence has also shown that the use of CACs did not affect bond pricing in the secondary market after the issuance of the bond.^[7] However, if the use or non-use of CACs is something that markets consider as relevant in pricing bonds (e.g., because it signals something about repayment probabilities), then we would expect that when a new bond issue occurs that involves a change in the contractual terms (towards or away from CACs), this should result in a change in the value of the existing stock of issuance from that issuer. We test this hypothesis via an event study and find no evidence that decisions about the contractual terms of new bond issues have affected the pricing of the existing stock of debt. This is an additional piece of evidence that CACs have historically not been viewed as a factor that influenced bond yields.

Second, this paper provides a very recent update on the pricing of bonds with and without CACs. Although Becker, Richards and Thaicharoen (forthcoming) have shown that the use of CACs had no impact on the pricing of a large sample of bonds in the secondary market in mid 1998 and mid 2000, much has happened since then. In particular, the November 2001 IMF proposal for the SDRM and the default by Argentina – the largest emerging markets issuer – have highlighted the problems of debt crises, which may have resulted in new focus by market participants on the role of CACs in pricing emerging market sovereign bonds. We therefore test whether there is a significant difference in the pricing of bonds with and without CACs as of 31 January 2003. The results indicate that there is still no evidence that CACs have had an economically or statistically significant impact upon bond pricing in the secondary market. Accordingly, we conclude that market participants have still not – as of early 2003 – focused on which bonds have CACs, or that they do not believe that the presence or absence of CACs should influence the value of a bond.

The rest of the paper is organised as follows. Section 2 provides some background on the nature and use of CACs. In Section 3 we present the results of the event study of decisions about CACs and their impact on the pricing of the outstanding stock of debt over 1994–2003. Section 4 contains the tests of the pricing of the CACs in a large sample of bonds in the secondary market on 31 January 2003. Section 5 concludes.

Footnotes

See IMF (2003) for further details of the proposed sovereign debt restructuring mechanism. [1]

See e.g., Chamberlin (2002) for a summary of private sector views from the head of one leading trade association. [2]

See the 31 January 2003 letter from the heads of seven private sector organisations to G-10 Finance Ministers, available at <http://www.emta.org/ndevelop/Final_merged.pdf>. [3]

Global bonds are bonds issued simultaneously into the US market, Euromarket and other markets. This appears to be the first sovereign issue into the US market with CACs, although there are instances of bonds issued into the Euromarket under New York governing law that include CACs (see Footnote 9). [4]

One obvious difference, however, is that a statutory approach would be a more comprehensive solution, since it would overcome problems such as aggregation of different bond issues (see e.g., Boorman (2003)). [5]

See Becker, Richards and Thaicharoen (forthcoming) and Tsatsaronis (1999). By contrast, Eichengreen and Mody (2000) suggest that CACs decrease borrowing costs for higher-rated issuers and increase borrowing costs for lower-rated issuers, though Becker, Richards and Thaicharoen suggest these results may be due to data problems and the instrumental variables technique used by the former to control for possible endogeneity in the use of CACs. [6]

See Becker, Richards and Thaicharoen (forthcoming), Petas and Rahman (1999), and Dixon and Wall (2000). [7]