The Prisoner’s Dilemma is a renowned problem in game theory illustrating the unstable interaction of individual and group interests. Two suspects are interviewed independently about a crime; if neither talks the evidence is such that they will both get short sentences; if each implicates the other they will both get long sentences; but if one remains silent while being implicated by the other, he will receive an even longer sentence while his betrayer walks free. In such a situation, the interests of the pair are maximised by a cooperative decision to hold fast. But each is individually incentivised to break ranks, both for a chance of immediate freedom, and to minimise the damage if the other does the same.
This thought experiment has apparently provided inspiration to promoters of large-scale corporate debt restructurings. Key to the capitulation pressure which the prisoners face is their inability to communicate and thereby coordinate their actions. The same impairment affects the class of holders of a tranche of corporate debt, who are likely to be both numerous and unknown to one another. Furthermore, corporate bondholders, like the prisoners, have the ability to alter the position of their peers, through the majority voting mechanisms which are written into the constitutional documents of most corporate debt issuances. Two such techniques have recently been considered by the English courts: “Exit consents”, in Assénagon Asset Management SA v Irish Bank Resolution Corporation Ltd  EWHC 2090 (Ch),  1 All ER 495; and “Consent payments”, in Azevedo v Imcopa Importação, Exportação e Indústria de Óleos Ltda  EWCA Civ 364,  1 BCLC 72.
Schemes of this type attempt to achieve the desired variation of the terms of the debt by coupling an exchange of the old notes for new ones with a noteholders’ resolution which has the effect of destroying (or reducing) the value of the old notes. The twist is that each noteholder must decide whether to commit to the proposal (by accepting the exchange and agreeing to vote in favour of the proposal)
before the date of the resolution. Thus a noteholder who believes that the exchange is detrimental is faced with the risk that a majority3 of its peers will accept the offer and leave it with nothing. If the exchange is plainly a bad deal for the noteholders then presumably it can be safely rejected. But in a more ambiguous case, the exit consent procedure introduces a severe downside risk for a noteholder wishing to preserve its existing rights.
In Assénagon, the exit consent technique was used in the restructuring of the bailed-out Anglo Irish Bank following the credit crunch. The new notes were issued at 20 per cent of the par value of the old notes (but with more favourable terms such as a government guarantee). While on its face this looked like a bad deal for the noteholders, in the background lay an express threat from the Irish Government, as part of post-crisis “burden-sharing”, to legislate for a reduction in the noteholders’ entitlements if they did not agree to a voluntary restructuring. The resolution which formed the other limb of the exit consent scheme rendered the old notes redeemable for a payment of 0.001 per cent of par.
This technique is more straightforward: a resolution for the desired variation of the terms of the notes is proposed, together with a promise to make a payment to all those voting in favour of the resolution in the event that it passes. The “consent payment” is made outside the machinery of the notes, thus avoiding, so the court in Azevedo found, infringement of their pari passu terms.
In Azevedo, the consent payments had been used in the restructuring of a Brazilian soybean processing group of companies. Consent payments were made in respect of a series of resolutions for the postponement of interest.
Although the exit consent technique perhaps seems closer to the classic Prisoner’s Dilemma and more overtly exploitative of the noteholders’ position, it is important to note that a similar effect is created by consent payments. The offer of payment is not simply a partial compensation for what is lost by virtue of the variation, intended to alter the attractiveness of the proposed variation in the eyes of the noteholder.
If it were, then it could be offered to all noteholders irrespective of how they voted.
Rather, it creates a risk for the reluctant noteholder that if it resists the variation, it might both lose the vote and be denied the payment. To put it another way, if the noteholder thinks there is a chance the resolution will pass despite its opposition, it might as well vote in favour and at least claim the payment. Thus there is a coercive effect which interferes with each noteholder’s consideration of the merits of
the variation in itself.
Both exit consents and consent payments are well established in the English and US markets. In particular, exit consents were used in some of the restructurings arising from the financial crisis. However, neither had been tested in English courts until the above-mentioned cases in 2012, when, in quick succession, Hamblen J in the Commercial Court in Azevedo held that consent payments were valid, and Briggs J in the Chancery Division in Assénagon held that exit consents were not – or rather, the court so held in respect of the particular examples under consideration: the extent to which the reasoning might apply to all such schemes is considered further below.
Appeals in both cases were due to be heard together but the Assénagon appeal was withdrawn leaving the Court of Appeal to uphold the validity of consent payments without expressing a view on exit consents. This was unfortunate because as the preceding discussion illustrates they are at root very similar. The fact that the Assénagon case was decided on the basis of the principle of oppression of the minority whereas that was not even capable of being alleged in Azevedo sheds light on the general scope of that doctrine.
In Assénagon the court found two independent bases for invalidating the resolution at the heart of the exit consent. The first was that, in respect of those noteholders who had accepted the proposal, the contracts for exchange of the notes (being specifically enforceable) created a beneficial interest in the old notes on behalf of the issuing bank which, according to the express terms of the old notes, disenfranchised the holders of those notes from voting on the resolution. It was an unusual feature of this particular scheme that the exchange contracts were entered into prior to the meeting, so most exit consent schemes would be unlikely to fall at this hurdle.
The second basis was more fundamental. Briggs J cited various common law formulations of the principle of minority oppression in the exercise of powers conferred on majorities, including the following proposition, in the specific context of debenture holders, from British America Nickel Corporation v MJ O’Brien Ltd  AC 369 at 371: “[Such powers] must be exercised subject to a general principle … that the power given must be exercised for the purpose of benefiting the class as a whole, and not merely individual members only. Subject to this, the power may be unrestricted.”(Emphasis added.)
Analysing the purpose of the resolution within the exit consent process, Briggs J found that it failed this test: “[The resolution’s] only function is the intimidation of a potential minority, based upon the fear of any individual member of the class that … he (or it) will be left out in the cold … This form of coercion is in my judgment entirely at variance with the purposes for which majorities in a class are given power to bind minorities, and it is no answer for them to say that it is the issuer which has required or invited them to do so.”
Significantly, this was not a decision based on any perceived unfairness in the underlying variation of terms offered to the noteholders – Briggs J was ready to accept that the offer may have been a beneficial one in light of the threat of state intervention – it was purely the element of coercion that was oppressive.
In Azevedo too, it was not contended that anything about the variation was inherently unfair. Rather the claim was brought on the basis, inter alia, that the consent payments were in the nature of a bribe. This argument was rejected both at first instance and by the Court of Appeal on the broad basis, which would seem to apply to all similar schemes, that such payments were not unlawful wherever two conditions pertained: the payment was available to all members of the class; and the basis of the payment was made clear to all members of the class.
Contrasting the two decisions
Lloyd LJ, giving the only reasoned judgment of the Court of Appeal, thought that Assénagon was too far away from Azevedo on the facts to be of any particular assistance. Standing back however, it is not intuitively obvious why one of these techniques should be lawful and the other not. They both represent the exploitation by the issuer of the disunity of the noteholders in order to induce approval of a variation.
The potential unfairness, which is the coercion, exists in both cases. Indeed in Azevedo, Lloyd LJ acknowledged the coercive effect of consent payments and referred to an academic paper which provides a technical game-theoretic proof of that effect. That paper expressly encompassed both consent payments and exit consents.
There may appear to be a difference in the severity of the two techniques: in Azevedo the postponement of interest was relatively minor and the consent payments commensurately modest, whereas in Assénagon the variation was more dramatic. But the ratios of the cases suggest that the decisions would be the same if a large consent payment was offered to make noteholders think twice about rejecting a more severe reduction in value (thus creating a greater potential detriment to any objecting noteholder), or if a less drastic version of the exit consent was used.
On the other hand, in both cases it is, in reality, the issuer who is exploiting the noteholders rather than any subset of the noteholders exploiting a minority of their peers; and the exploitation is applied equally and openly to all noteholders. That was enough for it to be condoned in Azevedo, but not in Assénagon.
The divergent conclusions arise only from the mechanism by which the exploitation is created. Where the coercion is effected by a noteholder resolution (as in exit consents), it is considered to off end against the minority oppression principle. Where the coercion is applied outside the resolution (as in consent payments) there is no transaction between the noteholders inter se on which that principle can bite. Arguably, this is a case of form over substance; the mischief is fundamentally the same.
MINORITY OPPRESSION MORE GENERALLY
In a 2012 article, one of the present authors compared the application of minority oppression principles in the bondholder and shareholder contexts. It was noted that the modern commercial approach to the duty of creditors may be less strict than in the case of shareholders, following the decision in Redwood Master Fund Ltd v TD Bank Europe Ltd  EWHC 2703 (Ch),  1 BCLC 149 in which Rimer J held that the implied duty between providers of a syndicated loan was limited to exercising the power “in good faith for the purpose for which it was conferred”, as opposed to the duty to act in the interests of the class as a whole which governs shareholder votes at class meetings.
In Assénagon and Azevedo, the court was happy to draw propositions from minority oppression cases across different contexts and it was agreed that there was a single fundamental principle with a common source.
Having referred to bondholder cases, Briggs J referred to: “a more general and enduring expression of the generality of the principle” in Greenhalgh v Arderne Cinemas  2 All ER 1120, which was a case concerning shareholders in which the principle was expressed in terms of a vote having to be for the benefit of the company as a whole.
Briggs J did however note that the extent and content of the principle was context dependent and subject to the intention of the relevant parties.
In this connection it is significant that the creditors who were held to be subject to a less onerous duty in Redwood were members of a syndicate who, as Lloyd LJ pointed out in Azevedo, would be able to collaborate in order to maximise their bargaining power.
Th e same may or may not be true of shareholders, depending on the company. It is clear from the Court of Appeal’s judgment in Azevedo that a consent payments scenario in respect of shareholders would be lawful provided that the payments were offered openly and equally.
Th e novel question raised by Assénagon is whether the minority oppression principle may be invoked in shareholder or other non-bondholder contexts to invalidate transactions in which majority voting is a tool for oppression in the hands of a third party. The court has shown (at first instance) that it is willing to offer the protection of this equitable doctrine to those who find their minority position taken advantage of, even if it is not the majority who are doing the taking.
The relaxed attitude of the Court of Appeal in Azevedo to bondholders who found themselves similarly exploited however, indicates that such protection may be open to challenge.
Stephen Moverley Smith QC and Harry Sharpe practise at XXIV Old Buildings, Lincoln’s
Inn, London, WC2A 3UP. Email: firstname.lastname@example.org, email@example.com.