Reforming the International Monetary and Financial System
Chapter

Comments: On Debt Management and Collective-Action Clauses

Editor(s):
Alexander Swoboda, and Peter Kenen
Published Date:
December 2000
Share
  • ShareShare
Show Summary Details
Author(s)
Pablo E. Guidotti

The 1990s have witnessed significant volatility in capital flows, especially in the second half of the decade. Volatility of capital flows to the private sector, in turn, has been closely associated with large swings in economic activity. For example, Figure 1 shows the recent relationship between capital flows to the private sector and GDP growth in Argentina.

Figure 1.Argentina: Private Capital Inflows and GDP Growth

Source: Argentine Ministry of Finance.

Some of the high volatility in capital flows may reflect the increased share of bond issues in total financing to emerging markets, and the corresponding decline in the relative importance of bank lending. In the case of Argentina’s capital account, bank lending has tended to behave countercyclically to the bond component. This phenomenon is illustrated in Figure 2.

Figure 2.Argentina: Capital Account of the Balance of Payments

(Billions of U.S. dollars)

Source: Argentine Ministry of Finance.

As part of the effort to deal with the volatility of capital flows in emerging markets, the international official community has provided significant financing packages to crisis countries, with the IMF as the principal institution in charge of the coordination and implementation of such initiatives. The use of large-scale packages of official external financing in response to disruptions in private capital flows, however, has increasingly come under scrutiny. As a result, various efforts—including the G-22 process—have been put into motion to search for ways to make the international financial system more resilient to shocks.

What Is the Debate About?

The current debate about crisis prevention and resolution, in my view, is reducible to three issues. First, prevention, namely, what policies reduce volatility. Second, moral hazard, namely, how to credibly reduce the size of official intervention. And third, if a crisis occurs, what mechanisms or institutions can be put into place to obtain an efficient resolution.

In this context, one aspect of the discussion has focused on how to better involve the private sector in the prevention and resolution of capital market crises. As part of what has been included under the heading of private sector involvement, various groupings—notably the 1996 Rey Report (see Group of Ten, 1996) and the 1998 G-22 report on international financial crises (Group of Twenty-Two, 1998b)—have analyzed the possibility of introducing innovations in bond contracts to facilitate restructuring, as well as promoting contingent credit or liquidity arrangements between governments and the private sector based on the successful experiences of Argentina, Mexico, and Indonesia in this regard.1 Eichengreen’s paper provides an excellent and comprehensive analysis of these proposals.

At present, views are divided. On the one hand, many are asking why the international community has not moved faster on implementation of existing proposals. On the other hand, market participants and emerging market issuers are becoming increasingly worried about the potential effects of such contract innovations on capital flows to emerging market economies.2 My own view is that this impasse reflects the fact that the current debate on private sector involvement has focused too narrowly on a set of technical proposals without clarifying adequately how their adoption relates to the set of issues set out above. To help clarify this relationship, I argue the following:

1. The most natural form of private sector involvement is through adequate debt and liquidity management policies.

2. Moral hazard (and, presumably, the need for large-scale financial packages) will be credibly reduced by strong preventive policies and adequate conditionality. I would not favor linking, in any way, official (IMF) policy to changes with respect to bond contracts.

3. Adoption of collective-action clauses should be seen exclusively as a mechanism to improve efficiency in crisis resolution.

Debt and Liquidity Management

Let’s consider first the issue of debt and liquidity management. If there is a lesson for emerging market economies from the past two years of global volatility it is the need for the public as well as the private sector to maintain adequate liquidity in the face of potential interruptions of capital flows. And liquidity is achieved through prudent debt management.3 While many of the following observations are also relevant for the management of private sector liabilities, I will focus mainly on public sector policy.

In this regard, Argentina has developed a consistent strategy based on four principles.4 First, the average maturity of the public debt has been lengthened in recent years to about nine years. Second, the Treasury has aimed at prefunding a fraction (typically, one-fourth) of its financing requirements in the following year. Long maturities and prefunding ensure that the public sector has flexibility for dealing with potential disruptions in capital markets. The government’s ability to produce low yearly financing requirements, together with the capacity of suspending its borrowing program in tough times, reassures markets that the government will not convalidate unduly high risk premia, and will not be forced to crowd out the private sector at the worst moment. Third, liability management—debt exchanges and contingency lines in favorable times—has enabled the government to smooth out its amortization schedule and has made short-term issues in difficult times possible. Fourth, the development of a domestic debt market based on a predictable debt issuance program, implemented through market makers, has been based on the growth of the domestic institutional investor base—pension and mutual funds. In turn, the growth of institutional investors is a direct consequence of the privatization of the social security system.

More generally, the preceding approach suggests that IMF conditionally could be enhanced by emphasizing the concept of liquidity and debt maturity as a central element of fiscal policy. This would serve to link the traditional objective of conditionality—that is, achieving long-term sustainability of fiscal policy—with the need to manage short-term restrictions imposed by markets.

To provide a specific example, consider the following equation describing the evolution of the ratio of public debt to GDP, b:

where d is the budget deficit as a proportion of GDP and n is the (assumed constant) growth rate. In a steady state, sustainability implies the following relationship between long-run growth and the budget deficit:

For a given long-run (potential) output growth rate, equation (2) characterizes a relationship between the stock of debt and the maximum (structural) budget deficit consistent with a nongrowing debt-to-GDP ratio. Although useful, this relationship says nothing about the borrowing requirement associated with the fiscal deficit in any given year. And we know that markets tend to look closely at that variable in periods of turbulence. Thus, we can define the yearly borrowing requirement (net of any prefunding) as a proportion of GDP, x, in the following way:

where m is the average maturity of the public debt. (It is assumed for the example’s sake that amortizations are uniformly distributed over time.) Equation (3) simply states that the yearly borrowing requirement is the sum of the deficit plus amortizations.

Suppose we are interested in a measure of the sustainable deficit that is consistent with not exceeding a target yearly borrowing requirement. Equations (2) and (3) imply the following relationship between sustainable budget deficit and average debt maturity:

For a given x and n, equations (2) and (4) provide fiscal sustainability criteria, which relate the budget deficit and public debt objectives to the maturity structure of the public debt. In particular, to ensure that a country has an adequate liquidity position vis-à-vis capital markets, as measured by the yearly borrowing requirement (net of any prefunding), there is an inverse relationship between average debt maturity on the one hand, and the sustainable budget deficit and long-run debt-to-GDP ratio on the other. The shorter the maturity of the public debt, the smaller the maximum allowable deficit and long-run debt-to-GDP ratio.

Implications for Official External Financing

The above relationship illustrates one way of introducing debt and liquidity management as a central element in the design of fiscal conditionality. An important implication of the above framework relates to the role of official external financing and, hence, to the potential for moral hazard. In particular, by defining a set of policies that constrains the size of the yearly net borrowing requirement, the potential use of official external financing would be both limited and effective in preventing contagion. This observation may be particularly applicable to the IMF’s newly established Contingent Credit Line (CCL).

Consider, for instance, a country—such as Argentina—with a debt-to-GDP ratio of 35 percent. If the average debt maturity is nine years and the budget deficit is 1 percent of GDP, then the gross yearly borrowing requirement would represent about 4.9 percent of GDP. Provided an adequate prefunding strategy covering future borrowing needs equivalent to 2 percent of GDP, a potential access under the CCL of up to 300 percent of quota would represent additional resources for about 2.9 percent of GDP.5 Thus, availability under the CCL would ensure that, because of an adequate debt-management policy, the country would be able to weather a total interruption of its access to the capital market for a period of one year. This represents an obvious example where the size of the potential official financing envisaged under the CCL would be effective in deterring contagion.

Now take the same example, but assume that average debt maturity is one year. In that case, the yearly borrowing requirement (net of prefunding) would amount to 34 percent of GDP. The same potential access under the CCL would represent only 8.5 percent of the yearly financing requirements. In such a situation, the CCL would most likely be ineffective in deterring contagion, and the volume of funds required for it to be effective would be inappropriately large.6

The above example illustrates how a constructive private sector involvement through prudent debt management and liquidity policies can serve to strengthen the role of official external financing, while at the same time reducing the potential for moral hazard associated with unduly large financing packages. It is important to stress that in my analysis no distinction should be drawn between external and domestic debt. In this regard, provided that keeping price stability is a policy objective, it is easy to show that all the arguments about sustainability apply to total debt rather than to its components.

Collective-Action Clauses

The above suggestions are predicated on the desirability of improving the resilience of emerging market economies to capital flow volatility, but when crises do develop the relevant focus shifts from prevention to efficient resolution. In this context, as I mentioned at the outset, proposals to include wider collectiveaction clauses in bond contracts have been at the center of the debate.

The first important issue to understand is the extent to which the introduction of such clauses represents a significant rather than a marginal innovation on existing bond contracts and, thus, what issuers should expect in terms of their marginal borrowing cost and their market access. As Eichengreen correctly argues, the notion of collective-action clauses already exists in bonds issued according to U.K. law (British-style bonds). Such clauses do not, however, exist in bonds issued according to New York law (American-style bonds). To the extent that emerging market countries already issue British-style bonds, we would be faced with a rather marginal innovation that should not, in principle, produce significant alterations in the pricing of emerging market debt.

Table 1.Argentina: Public Debt as of March 31, 1999
Billions of US$Percent
Total public debt113.6100
Bonds in foreign currency72.764
N.Y. law17.816
U.K. law12.011
Other9.28
Bradies (N.Y. law)17.916
Domestic market (Bontes, Bocones, Bonex, etc.)15.814
Bonds in pesos7.77
Loans29.626
Short-term T-bills3.63
Source: Argentine Ministry of Finance.

It is useful to look at the recent experience of Argentina, which has been a large emerging market issuer in the international capital market. Table 1 shows the composition of the Argentine Republic’s outstanding public debt divided into the following categories: non-Brady American-style bonds, British-style bonds, other external bonds, Brady bonds (restructured debt subject to New York law), and domestic bonds (in foreign currency subject to Argentine law).7 The Appendix describes the main legal features applying to non-Brady American- style and British-style bonds.8

As shown in Table 1, both American-style and British-style bonds account for a significant portion of Argentina’s outstanding public debt. In order to assess whether the governing legislation affects significantly the pricing of bonds, Figure 3 shows the behavior of the spread differentials (over U.S. treasuries) of domestic bonds versus those, respectively, of American-style and of Brady bonds during the Asian, Russian, and Brazilian crises. Figure 4 illustrates the behavior of the spread differential between American-style and British-style bonds as well as the corresponding sovereign spread over U.S. treasuries. Bonds are chosen that have comparable (although not identical) characteristics in terms of duration and liquidity.

Figure 3.Spread Differentials Over U.S. Treasury N.Y. Law-Argentine Law

(Basis points)

1 Domestic bonds in foreign currency: Treasury bonds in U.S. dollars 2002.

2 Non-Brady American style: Republic of Argentina 2003.

3 Brady: Floating rate bond.

Figure 4.Spreads Differentials Over U.S. Treasury U.K. Law-N.Y. Law

(Basis points)

1 on-Brady American style: Republic of Argentina 2003.

2 British style: Republic of Argentina euro 2003 (swapped to dollars).

A preliminary analysis of Figures 3 and 4 suggests that global volatility affects the relative price of the different bond categories, although caution is required because several factors impinge on the liquidity properties of different bonds.9 While there appears to be a systematic worsening of bonds issued under Argentine law, relative to those issued under New York law, in response to an increase in global volatility, there does not, however, appear to be a systematic change in spread differentials between British-style and American-style bonds.

The above observation suggests that, presently, the market does not penalize emerging market issuers on their British-style bonds. This provides support for the notion that, indeed, collective-action clauses represent a rather marginal innovation with respect to present practices. Thus, it is unclear why issuers should react negatively to the proposal of adopting British-style legislation in other jurisdictions. In fact, it could be argued that, by adopting legislation that would make restructuring more efficient in the unlikely event that default was inevitable, the associated losses to both debtors and creditors would be reduced. This, in turn, would tend to lower the (ex ante) marginal borrowing costs.

In my opinion, what drives emerging market issuers to be extremely cautious in their response to the proposal of widening the use of collective-action clauses is their perception that by adopting legislation that facilitates restructuring, future official external financing to emerging markets would either be reduced or tied to involuntary involvement of the private sector in restructuring initiatives.10 It is understandable, therefore, that the above proposal may be interpreted by the market as signaling a change in current official policy vis-à-vis emerging markets that would tend to increase rather than decrease marginal borrowing costs. Moreover, because international official policy in this area will continue to be subject to a significant degree of discretion, these changes may be seen as introducing a new element of uncertainty into contractual relationships between emerging market borrowers and the capital market.

This is why it is important that we focus, in the present debate, on collectiveaction clauses exclusively in terms of their potential contribution to making crisis resolution more efficient. In particular, this means that no link should be made between official external financing and the type of legislation applying to international bond contracts. Unless this issue is clarified and collective-action clauses become the international standard in industrial countries’ domestic and foreign bond contracts, it is unlikely that we will see their use expand rapidly among emerging market borrowers.

Appendix

This appendix summarizes the main requirements for the modification of the terms and conditions of non-Brady American-style bonds and British-style bonds, as well as the provisions dealing with the enforcement of creditors’ rights.

Non-Brady American-Style Bonds

Requirements for Modification of Terms and Conditions

Nonmaterial modifications. The fiscal agent and the issuer may agree, without the consent of bondholders, to any modification of the terms and conditions of the bonds that is of a formal, minor, or technical nature or is made to correct a manifest error. In addition, the fiscal agent may make any other modification that is, in his opinion, not materially prejudicial to the interests of bondholders.

Key modifications. Certain key modifications to the terms and conditions of the shelf bonds require the consent of every bondholder of the relevant series. These key modifications are as follows: (i) changing the due date for the payment of principal or any installment of interest; (ii) reducing the principal amount, the portion of such principal amount that is payable upon acceleration of the maturity, the interest rate or the premium payable upon redemption; (iii) changing the currency in which payments in respect of interest, premium, or principal are payable, or changing the place at which such payments are made; (iv) amending the definition of redemption event or the procedures provided therefore; (v) shortening the period during which the issuer is not permitted to redeem the bonds of such series if, prior to such action, the issuer is not permitted to do so; (vi) reducing the proportion of the principal amount of bonds that is required for a vote or consent of the holders to modify, amend, or supplement the fiscal agency agreement or the terms and conditions of the bonds or to make, take, or give any request, demand authorization, direction, notice, consent, waiver, or other action provided in the fiscal agency agreement or the terms and conditions of the bonds; or (vii) changing the obligation of the issuer to pay additional amounts.

Other modifications. Modifications to the terms and conditions of bonds, other than nonmaterial modifications and certain key modifications as described above, require the affirmative vote or written consent of bondholders holding not less than 66⅔ percent in aggregate principal amount of the relevant bonds outstanding at a bondholder meeting that satisfies quorum requirements. A duly convened bondholder meeting requires a quorum consisting of persons who can vote a majority in principal amount outstanding of the relevant bonds. If a meeting is adjourned due to lack of a quorum and is later reconvened, persons who can vote 25 percent in principal amount outstanding of the relevant bonds shall constitute a quorum for purposes of taking any action set forth in the notice of the original meeting.

Events of Default—Enforcement of Rights

Bondholders. If the issuer fails to pay any principal or interest on any shelf bond or if the issuer declares a moratorium on the payment of principal or interest on public external indebtedness, each bondholder of such series may declare the principal amount of bonds held by it to be immediately due and payable. In addition, in the event of an occurrence of any event of default, including payment defaults, the holders of not less than 25 percent of the bonds of the relevant series by notice to the issuer at the office of the fiscal agent may declare the principal amount of all the bonds of the relevant series to be immediately due and payable.

British-Style Bonds

Requirements for Modification of Terms and Conditions

Nonmaterial modifications. The trustee and the issuer may agree, without the consent of bondholders, to any modification of the terms and conditions of the bonds that is of a formal, minor, or technical nature or is made to correct a manifest error. In addition, the trustee may make any other modification that is, in the opinion of the trustee, not materially prejudicial to the interests of bondholders.

Key modifications. Certain key modifications to the terms and conditions of the bonds require a simple majority vote at a bondholder meeting in which a quorum consisting of two or more persons who hold or represent not less than 75 percent in principal amount outstanding of the relevant series (or at any meeting that is adjourned for lack of a quorum and later reconvened, not less than 25 percent in principal amount outstanding of the relevant series) is present. These key modifications are as follows: (i) Amending the dates of maturity, redemption or other payment of principal of the securities, or any date for payment of interest; (ii) reducing or canceling the principal amount of the securities of any series; (iii) reducing the interest rate of any security or varying the method or basis for calculating such rate; (iv) reducing any minimum or maximum interest rate; (v) changing the method of calculating the amortized face amount in respect of zero coupon notes; (vi) changing the currency of payment of the securities; or (vii) modifying the provisions concerning the quorum required at any bondholder meeting or the majority required to pass an extraordinary resolution.

Other modifications. Modifications to the terms and conditions of bonds, other than nonmaterial modifications and certain key modifications as described above, require a simple majority vote at a bondholder meeting in which a quorum consisting of two or more persons who hold or represent a majority in principal amount outstanding of the relevant series (or at any meeting that is adjourned for lack of a quorum and later reconvened, two or more persons who hold or represent holders of the relevant series, regardless of the principal amount of such series they hold) is present.

Events of Default—Enforcement of Rights

Trustee. If any event of default with respect to the terms and conditions of the bonds occurs, the trustee (i) may, at its discretion, declare the principal amount of the relevant series of bonds to be immediately due and payable or (ii) shall, at the request of bondholders holding not less than 25 percent in aggregate principal amount of the bonds of the relevant series of bonds then outstanding, declare the principal amount of the relevant series of bonds to be immediately due and payable.

References

Innovations in bond contracts include mainly the so-called collective-action clauses, which provide for collective representation of creditors, majority action to alter the payment terms of bond contracts, and sharing of payments among creditors.

See, for instance, comments included in the Institute of International Finance’s June 24, 1999, press release.

Guidotti (1996) discusses the interaction between financing policies in the public and in the private sector, and argues for an economy-wide liquidity strategy that includes the management of the maturity structure of the public debt as well as liquidity policies in the financial system.

These principles are discussed more generally in the 1998 report by the G-22 Working Group on Strengthening Financial Systems.

In the specific case of Argentina, this would amount to about US$8.6 billion.

Even a maximum access equivalent to 500 percent of quota would represent only 14 percent of the yearly borrowing requirement.

Other external bonds include German domestic issues, Swiss franc and Spanish peseta bonds, and Japanese yen “Samurai” bonds.

The legal framework applying to Brady bonds is similar to that of American-style bonds. For other foreign and domestic bonds, in most cases, contracts do not contain provisions regarding the modification of the terms and conditions of the bonds. In effect, the absence of any procedure for amendment makes modification of the terms and conditions of the bond contracts impractical.

Factors such as size of the bond issue and distribution among different types of retail and institutional holders, for instance, may affect the short-term behavior of spread differentials.

Expectations of this sort have been raised by the recent experiences of Pakistan and Ukraine.

    Other Resources Citing This Publication