Since the decision in Assénagon in 2012, there has been some doubt as to how much coercion can be applied in a consent solicitation that includes an “exit consent” and whether consents that seek to force detrimental economic change in the terms of bonds will stand up to scrutiny in the English courts. Market participants have generally taken a cautious approach to tender and exchange offers that include an exit consent, particularly those where the pay-out offered via the exit consent is below par. While the decision suggests that more extreme examples of this technique are likely to be unlawful, its findings are not as far-reaching as some market participants have tended to assume. And in light of the recent rapid rise in global interests rates combined with a succession of economic shocks that could incentivise issuers to attempt to squeeze-out non-tendering holders at below par, it is important that parties understand clearly where the red line was drawn in Assénagon and, beyond this, where uncertainty remains.
The exit consent technique sees an issuer invite bondholders to tender or exchange their bonds for cash or new bonds, on the condition that the bondholders agree to vote at a meeting of bondholders to pass a resolution (the ‘exit consent’) to amend the terms of the old bonds either to force an early redemption of the bonds at a set price or remove contractual protections afforded to those that remain in the bonds. In other words, the technique’s core features are the (A) ‘carrot’ of a cash redemption at an acceptable price or new bonds at an agreed exchange ratio and (B) ‘stick’ of impairment to the value of the old ones either with a lower cash payment or removing contractual protections that would permit a holder to accelerate remaining bonds.
The technique emerged in the US as a way to manage risks following the rise of ‘junk bonds’ in the 1980s. The technique has now become fairly popular in the US, especially in the wake of the 2008 financial crisis. Yet, even in the US, the use of exit consents2 has attracted litigation.3 Despite the associated litigation risk, exit consents have become an increasingly common tool used in liability management of bonds following the credit crunch,4 although there have been very few examples that have sought to force an exit at below par in that time. However, with the combined factors of higher interest rates and reduced risk appetite pushing up yields in the debt market, a significant number of fixed income debt instruments, issued before the pandemic in the low interest rate environment that followed the credit crunch, are seeing their values trade below par in the current environment, and the use of exit consents below par is likely to become a more commonplace tool in international capital markets. In that context, there is an increasingly urgent need for clarity in this area – in particular, the extent to which the viability of a ‘below par’ exit consent is limited by the “abuse principle”.
The abuse principle
It is an established5 proposition of English law that, where a majority is empowered to bind the minority, this power is not unlimited. The majority must exercise the power in good faith, for the benefit of the class as a whole (be they creditors, shareholders or some other class of stakeholder).6
Bond terms typically provide for amendment by resolution passed by a majority of bondholders, including reserved matters requiring a higher quorum / majority, which permit amendment of the economic terms, even where there are dissenting holders, so long as the requisite majority is reached. While all holders must accept, when they acquire an instrument, that the terms do permit such amendments (and, in fact, this is necessary to prevent contractual paralysis as it is unlikely 100% of holders will ever vote, or even be identified in a publicly traded instrument), this does not mean that such acceptance extends to the situation where the majority has exercised its power contrary to the abuse principle – as was the case in Assénagon.
Facts in Assénagon
Anglo Irish Bank Corporation Limited (the “Bank“) had issued bonds in 2007 with a nominal value of €750m due in 2017 (the “Notes“). The Notes were governed by English law and constituted by a trust deed which customarily permitted a majority of bondholders to modify any provisions of the Notes through an extraordinary resolution. Following the 2008 financial crisis, the Bank faced a liquidity crisis which led to its rescue by the Irish government. As part of the rescue plan, the Bank pursued a voluntary restructuring of its subordinated debt (including the Notes).
The Bank invited bondholders to exchange their existing Notes with new ones on more favourable terms but with an exchange ratio of 0.20 (i.e., 20 cents new notes for every one Euro of the (exchanged) Notes). The offer reflected the level at which the Notes were trading at the time with the added benefit that the new notes would be guaranteed by the Irish government.
The Bank required exchanging bondholders to commit irrevocably to voting in favour of a resolution (the exit consent) which entitled the Bank to redeem the Notes of non-exchanging holders for 0.00001% of their face value. In other words, any bondholders who did not participate in the exchange offer would lose almost the entire value of their existing Notes if the resolution was passed, since they would not receive the new notes but rather see their Notes redeemed at a fraction of their nominal value. The exchange offer was disclosed to all bondholders at the same time.7 The resolution was passed with 92.03% of the bondholders voting in favour of the resolution. The claimants (minority bondholders) challenged the resolution as an abuse of power by the majority bondholders which unfairly oppressed the minority bondholders who did not wish to accept the exchange offer.
The High Court agreed with the claimants8 and held that this particular exit exchange violated the abuse principle.
The Court found that the essential question before it was “can [it] be lawful for the majority to lend its aid to the coercion of a minority by voting for a resolution which expropriates the minority’s rights under their bonds for a nominal consideration[?]” – the Court’s answer: no. Given the fairly narrow scope of this question, any impression that Assénagon established a blanket prohibition on coercive exit consents in the context of debt instruments governed by English law would be misplaced.
However, it is true that Assénagon has left some uncertainty as to what amount of coercion will be permissible before an exit consent tips into unlawful abuse (see final section below), and this seems to have been the basis of many market participants deciding against below par exit consents due to a potentially misguided assessment as to the reach of the decision, which is worth reconsidering in the current economic environment. The Court appeared, in places, to base its reasoning on the inherent nature of an exit consent (not just this specific one) as a “coercive threat” or “negative inducement”, rather than an offer of a standalone benefit to induce bondholders into an exchange or new terms, such as a consent payment.9 This might seem to impugn exit consents generally. Yet, it was clearly the extremity of the coercion in Assénagon – the threat that the minority be left with (next to) nothing – rather than coercion per se, that meant the exit consent in that case was abusive.
Some have argued that this conclusion ignores the fact that those who voted for the exchange would have secured the benefits of: (i) real value for distressed securities on the verge of being downgraded in exchange for the new notes; and (ii) a government guarantee in respect of the new notes. However, the Court seemed to suggest that any benefits of new bonds, where a corresponding exit consent resolution threatens expropriation, will be irrelevant unless the minority receives, after the resolution has been passed, another opportunity to exchange the old bonds for the new (see further discussion under “Drag-along mechanisms” below).10 The resolution in Assénagon (to which no such second opportunity was attached) could not be characterised as for the good of the class as a whole, since it threatened to leave those who voted against it “out in the cold”11 with worthless Notes.12 Anyone considering the options would be motivated by the real ‘carrot’ of avoiding the ‘stick’.13 The “only purpose” of the exit consent, the Court found, was to “prey upon the apprehension of each member of the class… that he will, if he decides to vote against, be part of that expropriated minority if the scheme goes ahead”.14 This “form of coercion” (emphasis added) was “entirely at variance with the purposes for which majorities in a class are given power to bind minorities”.15
Status, reception and future of Assénagon
The decision in Assénagon was the first, and remains the only, English court decision to consider the legality of an exit consent. An appeal was lodged in Assénagon, but did not proceed. Therefore, it seems possible that a future decision (of an appellate court) may offer more clarity here or, even, move the dial on acceptable coercion vs. unlawful abuse.
Some commentators have criticised Assénagon for going too far in restricting the role of coercion, and argue (even extreme) coercion can play a proper part where contractual parties have agreed to amendment by the majority without placing express limits on the exercise of this power. The counter to this is, as Assénagon noted, English law has long-recognised a limitation of the majority’s power to bind the minority. The limitation may also be seen as in-step with English law’s recent tendency of increased openness to arguments for implied terms of reasonableness16 and good faith17 in the exercise of contractual discretion.
Parties to future trust deeds (or other financing documents, e.g., syndicated facility agreements) may attempt to exclude the abuse principle18 in exchange, for instance, for a higher coupon, or define what level of coercion will not be challengeable by the minority. No English court has tested this specific type of clause. The decision in Redwood characterised the abuse principle as an implied term.19 This would suggest the principle is capable of exclusion. However, in Assénagon, the Court noted the principle was also derived from “general principles of law and equity” (as well as by implication), which may cast doubt as to whether or not exclusion will be effective here.20
The decision in Assénagon also demonstrates that the English courts are likely to pay close attention to the jurisprudence of the US courts as to when the line between legitimate coercion and abuse is trespassed. In its judgment, the Court cited a decision of the Delaware Chancery Court which had upheld an exit consent falling short of expropriation, from which it distinguished the offer in Assénagon.21 The Court also noted a previous decision considering the separate mechanism of consent payments.22 This decision had drawn on the fact that consent payments “had been a common feature of debt refinancing in the USA for some time” as part of its reasoning in concluding that technique was lawful.23 This may indicate that, to the extent the use of exit consents continues to become more common in the US, the English courts may be readier to accept their legality.
The decision in Assénagon (and the subsequent decision in Re Co-Operative Bank) offered some reassurance that using a drag-along mechanism may prevent an exit consent from transgressing into the realm of abuse:
- In Assénagon, the Court seemed to approve the claimants’ suggestion that “permitting (or even forcing)” those who did not vote for the exchange into the same exchange that the majority had accepted would “deprive the exit consent of its coercive effect”.24
- In Re Co-Operative Bank, the High Court sanctioned a scheme of arrangement which excluded retail noteholders, in circumstances where a substantial majority of such noteholders had voted in favour of a resolution which would have the effect of cancelling all retail noteholders’ subordinated notes in return for a cash payment. Although the main question before the Court did not turn on the findings in Assénagon,25 the Court considered these findings, in passing, to conclude that the noteholder resolution did not offend the abuse principle.26 Unlike in Assénagon, the retail noteholders faced no risk of being “left out in the cold”, i.e., there was no threat of expropriation.27 In fact, the cash payment the retail noteholders were due to receive was likely to provide them with an increased amount of cash (£4.50 for every £10 of debt) than the implied equity value of the notes under the scheme of £1.50 for every £10 of debt, offered to the non-retail noteholders.28
This approach is consistent with that taken in transactions where the exit outcome is broadly equivalent with the tender or exchange offer. In other words, the non-voting or dissenting holders receive the same economic terms as those that do participate, save potentially for the payment of a consent fee (although on some occasions this is also offered equally to all holders). However, as we explain below, a belt-and-braces approach of tacking-on a drag mechanism whenever an exit consent is used, may undermine the utility of the exercise.
How much coercion is too much coercion?
The decision in Re Co-Operative Bank does not suggest that dissenting or “dragged” noteholders must receive a more favourable cash return than assenting noteholders, in order to avoid offending the abuse principle. The Court’s reference in Assénagon to, and focus on, the particular “form” of coercion does seem to acknowledge that forms of coercion other than expropriation may be acceptable in this context. Based on this, one might assume that an exit or drag price for securities, equal to or some amount below the (fair) value returned to the majority may be acceptable as ‘for the good of the class as a whole’, and therefore, not fall foul of the abuse principle. Beyond this, there remains a real question as to how much of a reduction in value a majority resolution can impose on the minority, short of expropriation.
However, any impression that parties should seek to insulate themselves from litigation risk by always coupling an exit consent with a drag-along, seems to go too far and beyond the narrow question that the Court considered in Assénagon. As above, it is clear that some level of coercion is permitted. Equally, as a commercial matter, a coercive consent that permits a drag-along right after the vote is cast (if a holder may be left out in the cold) may render the ‘carrot’ of the consent process itself redundant if issuers are not able to ‘incentivise’ holders to vote in favour of a resolution, and in increasingly gloomy economic times, removing the ‘stick’ from the process may render the whole process ineffective. If everyone gets a ‘second chance’, then the requisite majority may rarely be reached, since there would be a significant incentive to vote ‘No’, just to see what happens.
Another factor which may reduce the appearance of abuse is if the inducement has been properly disclosed to all members before voting. However, based on Assénagon, it is clear this will not prevent a finding of abuse where the coercion used is tantamount to expropriation.
It is arguable that, in light of Assénagon, exit consents that threaten to strip the minority of key protections29 (e.g., security rights, cross-default rights) are also in danger of posing an abusive threat to the value of the minority’s holding. However, the Delaware exit consent, cited by the Court in Assénagon (see above), threatened removal of significant negotiated financial protections and deletion of all financial covenants for dissenting bondholders. Although the English Court pointed out that the Delaware Court had not analysed the exit consent through the ‘abuse principle’ lens (but in terms of a contractual duty of ‘good faith’), the question before the Delaware Court was materially similar: whether the exit consent was “wrongfully coercive”. The English Court seemed to distinguish the (lawful30) Delaware example from the unlawful one in Assénagon, and described the Delaware exit consent as “falling short of expropriation”.31 This might indicate that the English courts could, in fact, be poised to accept a fairly high level of coercion before deeming an exit consent is unlawful. Exit consents that threaten to remove other, less important, protections may be more easily distinguished from the abusive exit consent in Assénagon.
Considerations under US securities laws
Whenever a consent to amend the terms of a bond involves solicitation of votes from bondholders located in the US, the transaction parties should consider the potential application of US securities laws. In respect of a consent solicitation, the parties should consider whether the amendments proposed are so fundamental that they substantially affect the legal rights and obligations of the holders of the outstanding securities, thus resulting in a new investment decision for those holders. If it is determined that there is a new investment, then the new security that the investor will receive after the consent may be subject to the registration requirements of the US Securities Act of 1933, as amended (the “Securities Act“) for exchange offers. Whether or not a consent would constitute a fundamental change is an extremely fact-based test, with no bright line rule adopted by the courts or the US Securities and Exchange Commission, as to what amendments would or would not constitute the sale of a new security under the Securities Act. While it is generally accepted that modification of non-financial terms should not create a new security, amendments to terms such as the principal amount due at maturity (for example, by inserting a call provision that allows the issuer to redeem the bonds at below par) are more at risk of being treated as a new security. In such cases, parties should give some thought to both the exemption from registration under the Securities Act as well as potential liability under SEC Rule 10b-5.
The more significant jeopardy an exit consent threatens against the value of the minority’s holding, the bigger the litigation and reputational risk to parties involved. This does not mean that an exit consent that sees minority holders receive a different outcome to consenting holders will be considered coercive. Inevitably, the facts will be determinative, and where there is a material difference in outcomes for those that vote in favour compared to those that don’t, the risk increases. This also does not mean that the threshold for destroying the economics would be set at par – the economics may be well below par if the present value sits here due to the specific circumstances of the debt in question (for example, the coupon, tenor or financial stability or prospects of the issuer or the industry or country in which it operates). Market participants should note, however, that even if an English court were to rule that an exit consent at below par does not offend the abuse principle under English law, this will not remove the risk of potential limitations under the laws of other applicable jurisdictions. For instance, the deemed new security doctrine in the US described above may pose particular challenges.
Key questions to consider are:
- Are noteholders voting for the resolution to avoid suffering the outcome threatened against dissenters, or is the main motivation some standalone benefit for the whole class?
- If the former, then the transaction is in Assénagon territory, but the mere fact of reliance on a negative inducement will not be conclusive as to unlawfulness.
- Does the outcome threatened against dissenters significantly impair the value of the dissenting voter’s asset?
- If yes, then there is a high risk of falling foul of Assénagon.
Only time and successive consents will inform where the line sits, but while market participants test this, our expectation is that so long as there is not a material difference (determined on a case by case basis) between the outcome for holders who vote for or against a coercive consent, issuers should not shy away from making proposals to bondholders that may previously been considered to be ‘off limits’ following Assénagon.