Mark Weidemaier and Mitu Gulati teach classes on international debt at the Universities of North Carolina and Virginia and co-host the podcast Clauses and Controversies. They would like to thank Heream Yang, Rishabh Sharma and Leah Berger for their research assistance.
Emerging sovereign-debt markets look increasingly bleak, and it seems likely that multiple restructurings are in the offing. That means aggregated Collective Action Clauses (CACs), a new contractual tool adopted around 2014 to ensure orderly restructurings, will get a serious road test. But we worry that early adoption in Ukraine — and in a remarkably smooth negotiation, no less — has revealed a serious design flaw. Or feature.
First, some basics. Under the new aggregated CACs, debtors seek creditor approval for a restructuring in one of two ways. The first is called the single-limb process, where the debtor offers the same deal terms at the same time to all holders of multiple bond series. If 75 per cent of that group agrees, the deal is approved. The second option is the “two-limb” process, where the debtor needs the approval of only ~66 per cent of the entire negotiating group, but must also win the approval of at least half of each individual bond series. At first many assumed the single-limb process would be used most often. After a default, the thinking went, investors would all accelerate their claims, leaving everyone in the same position regardless of their original maturities and coupons. As things turned out, investors holding bonds with different maturities and coupons do have different interests. So the two-limb mechanism has been used in every restructuring since the advent of these clauses.
This two-limb mechanism can create some problems.
Let’s first think about how the provision is ostensibly supposed to work. Assume there is a bond series or two dominated by investors opposed to the restructuring. Can they be left out of the aggregated vote? The answer is yes, at least in some cases. The bond documentation makes clear that the issuer may exclude a subset of bonds from the restructuring vote. Imagine the issuer says in advance that it proposes to restructure Series A through C, but not Series D (in which a particularly obstreperous investor holds a large stake). This process would work smoothly. If approved by 66.67% of the combined vote of the three series, plus 50% of each one, the restructuring would take effect.
But what if the issuer wants to wait until after the votes are in, and then decide which series to include in the restructuring?
Can it survey the votes and then “redesignate” the group of affected securities so it can meet the requirements of the two-limb voting process? It seems odd to allow such a thing. This type of redesignation undermines the purpose of the aggregated voting mechanism. We must presume the intent of requiring a 66.67% affirmative vote in the aggregate across all series was to make sure the restructuring enjoys broad support, while the requirement of majority approval by each series was designed to prevent the creditors from teaming up with the debtor against one would-be bagholder. But with redesignation, a restructuring that failed to garner broad support can nevertheless be imposed on a subset of creditors. And, unlike the example where the issuer clearly indicates in advance which series of bonds will be bound, investors have no way to know how the issuer will exercise its discretion until after the vote.
In late 2000, Argentina did this very thing with bonds governed by New York law. The bonds’ contracts dictated that the choice of which series to include in the restructuring would be “final.” That language seemed to explicitly prohibit redesignation after the vote. But Argentina, being Argentina, found a loophole. Its contracts allowed for less important (“non Reserved”) matters to be modified with only a majority vote. In collaboration with a subset of creditors, it modified the contracts to remove this barrier to redesignation. Put plainly, it used a majority vote to change a supermajority vote requirement. But that was Argentina. At the time, one could hope that other debtor countries, perhaps more concerned for their reputations as borrowers, would avoid using this loophole.
Ukraine has proven that view wrong. In its recent debt reprofiling, it indicated it would redesignate the vote as needed to obtain relief. While this proved unnecessary given widespread creditor support, we wonder whether investors fully appreciate just how easy redesignation could be for most debtors. Unlike Argentina, Ukraine did not need to be especially creative in reading its bond contracts. Indeed, as we read its bond contracts, it might not have needed anyone’s permission — even a bare majority — to redesignate as it pleased. In the consent solicitation for its debt reprofiling, Ukraine told creditors it had the power, in its “sole discretion,” to redesignate their votes as it needed. And indeed, that seems to be the most plausible reading of the contracts. To be fair, nothing explicitly gives the country a post-vote power to redesignate. However, section 12(a)(iv)(J) requires Ukraine to give creditors notice of redesignation, specifying:
. . . if applicable, the conditions under which a multiple series aggregation will be deemed to have been satisfied if it is approved as to some but not all of the affected series of debt securities.
On its face, this is simply a notice provision. But it seems to imply that Ukraine can exclude a subset of its bonds from the restructuring after the vote has taken place. The provision explicitly contemplates that some series of bonds will be included in the vote (hence, “affected”), will not approve the modification, and will then be excluded so that the two-limb voting thresholds “will be deemed to have been satisfied.” What else could this mean, other than that the issuer can redesignate?
Given the underlying purpose of the two-limb CAC — again, ensuring broad-based support while protecting against discrimination — we find it odd that redesignation should be so easy. Giving the debtor this level of discretion over how to count the votes can, in theory, gut the two-limb voting mechanism. It is hard for us to imagine that investors understood this when they agreed to the inclusion of these CACs as a standard part of international bond documentation. Then again, perhaps we shouldn’t be surprised. When times are good, who reads bond contracts?
What’s the upshot of all this? The Ukraine deal passed smoothly, as far as we can tell. Investors didn’t revolt against the sole discretion redesignation power that Ukraine claimed for itself. But the Ukraine deal only required a stretching of maturities that already had widespread creditor support. The same provision exists in some bonds issued by Sri Lanka and Pakistan, who may not have comparable support.
The question is, what will happen with the next set of restructurings, when the sovereign debtors ask for real haircuts? Will it be too late?
