In the April edition of the Africa Newsletter, I wrote about the continuing controversy surrounding the interpretation of the form of the Pari Passu clause included in sovereign bond issues and the particular problems that had been created as a result of the remedies made available to the plaintiffs by the New York courts in the case of NML Capital, Ltd –v- Argentina. In this article I want to look at a different but related point concerning the inclusion of so-called “collective action clauses” (“CACs”) in sovereign debt instruments. Indeed, the US Court of Appeals for the Second Circuit recognized the link between these two issues when rejecting the appeal of the Republic of Argentina against the first instance decision of Judge Griesa, when they stated that many of the potential problems with hold-out creditors that a wider interpretation of the Pari Passu clause would create could in fact be resolved through the adoption of CACs in the relevant bond instruments.
So what is a CAC? Though that terminology, and the debate around it, assumed prominence in the context of the initial Greek debt crisis back in 2011/12, the concept has been around for a long time particularly in bond issues governed by English and Japanese law (it had fallen out of favour in New York governed instruments principally because the device had been prohibited for domestic issues by the Trust Indenture Act 1939). At its simplest, the clause provides for the terms of a bond to be varied subsequent to issuance by a decision of the bondholders themselves. That decision can be taken at a meeting of the bondholders or in writing but it requires a specified majority and , in the case of so-called Reserved Matters, that requirement is increased and linked to an additional requirement as to the size of the quorum required before the matter can be put to the vote. “Reserved Matters” cover those matters which go to the economic core of the bond; for example, the payment dates, the amounts payable, the currency of the bonds and the governing law. In these cases it is normal to require a quorum of two-thirds of the aggregate principal of the bonds outstanding with a voting threshold of 75% of the aggregate principal of the amount outstanding of the bonds represented at the meeting.
The application of such a variation mechanism means that, if the requisite majority have voted in favour of the change, then the opposing or abstaining minority are bound by the changes as well. Thus regardless of which interpretation of the Pari Passu clause the potential hold-outs may adopt, it will not assist them because they are forced to accept the changes agreed to by the majority so there are no different terms on the basis of which they can argue for pro rata or ratable treatment when it comes to payments. So far so good, but there is still the capacity for vulture funds or the like to ensure that they hold or control bonds which have an aggregate value which is such that they are able to block the passing of a resolution relating to a Reserved Matter in respect of any particular bond issue or series of bonds.
The problem for borrowers, and those creditors who are minded to agree to a debt rescheduling, is that not only can a minority of the creditors prevent changes being agreed in relation to a particular series, but the fact that one issue of bonds or series of bonds, does not sign up to a rescheduling package means that it is challenging to get bond holders under other issues or series to commit to a rescheduling since they will be concerned that not all creditors are sharing the pain of a haircut equally. Furthermore the borrower will continue to be challenged by the need to service the non-rescheduled debt thereby potentially straining its capacity to honour the obligations on the rescheduled debt and generally turn its credit position around. This is an issue not just for the borrower and its commercial creditors, but also for the official sector which has skin in the game and a desire to facilitate an orderly sovereign work-out. It is therefore no surprise that institutions such as the IMF and the EU have been at the forefront in trying to improve the range and efficacy of CACs.
The “solution” that is being pursued is a contractual one, in the sense that there is a recognition that in the context of international bond issues which are governed by the laws of countries other than the borrowing state or state entity, there is no easy way to deal with this through some form of legislation (in the recent Greek debt rescheduling, those bonds which were governed by Greek law were dealt with by the simple expedient of imposing the “solution” – the Greeks enacted legislation which for voting purposes aggregated claims across all affected domestic law issuances thereby eliminating the power of creditors to achieve a blocking position on any single issuance). Accordingly there have been a series of recommendations to the market and prospective borrowers as to the form of CAC that should be included in bond issues going forward.
Broadly, the proposed clauses for new bond issues seek to dilute the capacity of potential hold-out creditors to form a blocking minority. This is achieved by making the size of the investment required to form such a blocking group so large as to render it financially impossible for the likes of a vulture fund to acquire sufficient bonds. Of course it is worth pointing out that if the proposed rescheduling is so aggressive and unfavourable to creditors, and understates the real economic strength of the borrower, then the blocking minority is likely to coalesce as a result of the independent decisions of a diverse group of creditors regardless of any specific, disruptive, action taken by vulture funds.
The dilution is brought about by aggregating all the bonds issued by a particular borrower so that the size of the pool taken into account for determining the necessary thresholds is increased. So if a borrower has four current bond issues each with USD 100 million outstanding, and the voting is aggregated across all four, then on the basis that there is still a need to get 75% approval of a change to Reserved Matters then potentially a vote of bondholders holding more than USD 300 million in principal outstanding is required but more importantly, a minority that wishes to block the change has to hold at least USD 100 million rather than the USD 25 million that would be required to block changes to one of the bond issues (and thereby potentially derail the whole rescheduling as mentioned above) if there had been no aggregation.
The thinking and debate about aggregation began before the Greek debt crisis threw it into sharp relief. Initially the suggestions centred around the use of a two limb voting procedure. This required a percentage of the bondholders in each issue to agree to the proposed change but at a lower percentage than usually required for adjusting Reserved Matters, say 50% rather than 75%, with there then being a vote across all the bond issues, where the voting is aggregated to approve the changes, when the aggregated threshold for approval would be two thirds. This still leaves open the prospect of a blocking minority being put together at the first limb though the amount required to do that is higher than previously would have been the case.
Accordingly, there has now been developed a single limb proposal whereby there is only the one vote which is on an aggregated basis. This means that the influence that bondholders in one issue or series can exercise is dramatically reduced even to the point where technically one could have a situation where all the bondholders in an issue could vote against the proposal but still be bound by it because bondholders in other issues voted in favour. There is clearly scope here for oppression of the minority and much thought has been given as to whether this presents an inherent problem under New York or English law (being the most commonly adopted governing laws for international bond issues). The current thinking, based on the case law is that single limb voting procedures will be upheld by the courts in both jurisdictions provided that (and with some simplification of the analysis for both legal systems) the majority are acting in good faith and for the benefit of the bondholders as a whole. In order to ensure that these criteria are met, the form of clause used provides that the proposal put to the bondholders must be uniform for all the bondholders.
To further encourage creditor participation, clauses are recommended to require the borrower to be transparent as to its economic condition so that the bondholders are in possession of all material facts at the time they are called upon to vote. In addition, sovereign bonds now often include disenfranchisement provisions which exclude for voting and quorum purposes all bonds owned or controlled directly or indirectly by the borrower or its public sector entities. Such provisions are designed to limit the ability of a sovereign borrower to manipulate the voting process on a single limb voting structure.
So what of those African sovereign bond issues that I referred to in my earlier article and which I analysed to see to what extent the new thinking on the drafting of Pari Passu clauses had been adopted? Has the new thinking on CACs also been incorporated? The short answer is “yes” in relation to the more recent issues such as those of Ethiopia and Tunisia. In those issues the full panoply of provisions have been included giving the borrower the option to modify the particular series only but also use multiple series aggregation two limb voting and multiple series aggregation single limb voting. What they will not do of course is bind the bondholders on earlier issues- so the Tunisian issue of July 2014 does not have these mechanics but the issue in January 2015 does.
As the climate for emerging market debt gets cooler with the cold winds of a China slowdown, lower commodity prices and the rising value of the US dollar, combined with interest rate increases, it may not be long before those new formulations of both of the Pari Passu clause and CACS, are put to the test.