The use of exit consents has been an important feature of many bond restructurings (see box “What are exit consents?”). The recent landmark decision in Assenagon Asset Management S.A. v Irish Bank Resolution Corporation Ltd (formerly Anglo Irish Bank Corporation (Anglo Irish)) casts doubt on the legality of coercive exit consents under English law ([2012] EWHC 2090).

The exchange offer

Between 2009 and 2010, Assenagon Asset Management S.A. (Assenagon) acquired €17 million of subordinated notes due in 2017 (2017 notes) issued by Anglo Irish. In the depths of the Irish financial crisis in January 2009, Anglo Irish was nationalised. The Irish government proposed a restructuring of Anglo Irish’s subordinated debt.

Anglo Irish launched a distressed exchange offer pursuant to which the bondholders could exchange their 2017 notes for new senior notes in an exchange ratio of 0.2 per each 2017 note and an “exit consent” commitment to vote in favour of a resolution introducing an amendment that would allow Anglo Irish to redeem all outstanding subordinated notes for the nominal amount of €0.01 per €1,000 in principal amount of the 2017 notes.

The resolution was passed at the bondholder meeting, allowing Anglo Irish to redeem all subordinated notes outstanding post-exchange for this token amount. Assenagon had not accepted the exchange offer, so received €170 for notes which had a face value of €17 million.

The challenge

Assenagon challenged the legal validity of the resolution on three grounds:

  • Pursuant to the terms of the bonds and the underlying trust deed (the trust deed), the resolution was not within the powers of the majority.
  • The bonds were held to the benefit of Anglo Irish at the time of the bondholder meeting and so had to be disregarded for voting purposes at the meeting. (This was, however, an unusual mechanical feature specific to the Anglo Irish exchange offer and not reflective of general market practice. Typically, the acceptance of bonds for purchase is expressly stated to be conditional on the passing of the resolution.)
  • The resolution constituted an abuse of the majority’s voting powers because it could not be inferred that the resolution benefited the bondholders as a class, so was unfair and oppressive of the minority.

The decision

The High Court granted the declaration sought by Assenagon. It ruled on the three challenges as follows:

  • After detailed consideration of the terms of the bonds and trust deed, it was concluded on balance that the complete extinguishment of bondholders’ rights was, in fact, within the powers conferred by the trust deed.
  • All the bondholders who voted in favour of the resolution held the notes beneficially for, or for the account of, Anglo Irish, from the date on which it accepted the exchange offer. The votes therefore had to be disregarded pursuant to a customary disentitlement provision in the trust deed to prevent a conflict of interests.
  • More critically, and with implications for the wider bond market, it was unlawful for the majority bondholders to aid the coercion of the minority by voting for a resolution which would destroy the minority’s economic rights under their bonds.

Serious implications

This decision has attracted huge interest among many financial market participants because of a number of serious and immediate implications.

In particular, the Eurozone authorities, sovereign wealth funds and banks will need to consider carefully the future use of any coercive element in English law exchange offers. This could rob them of a useful weapon in bank and sovereign bond restructurings at a time of enormous stress in the markets. Restructuring senior and subordinated bonds governed by English law will now be more difficult to achieve, and a greater level of negotiation and engagement will likely be needed.

This does not, of course, displace the possible use of coercive liability management techniques under domestic legislation in any jurisdiction. Anglo Irish could have relied on the threat of a Subordinated Liability Order issued under the Irish Credit Institutions (Stabilisation) Act 2010 to cram down the bondholders, for example. However, the fear of a challenge on constitutional grounds seemed to have prompted Anglo Irish to use a contractual coercive consent solicitation instead.

In terms of bond restructurings involving trustees, the bond trustee is likely to hesitate in allowing a notice calling a bondholder meeting or resolution to be circulated if it perceives that there is a coercive element in the exchange offer. Similarly, English counsel will be more cautious before approving structures that could give the slightest indication of coercion in bond exchanges.

This decision also demonstrates the real “protective” value in holding English law bonds (and will likely embolden the bondholders who are “holding out” on the Greek debt restructuring).

Considerable litigation may well follow. Investors whose bonds were redeemed and those who exchanged in fear of the coercive payment terms under the exit consent are likely to seek to assert that their bonds remain in existence and/or that they should be compensated.

On reflection

An English law exit consent in principle is fine, so long as it is not coercive of the minority position. In the present case, it was a significant factor that there was a substantial disparity between the value offered to those who accepted the exchange offer and voted in favour of the resolution, and those who did not. As Briggs J emphasised: “The exit consent is, quite simply, a coercive threat which the issuer invites the majority to levy against the minority, nothing more or less. Its only function is the intimidation of a potential minority based upon the fear of any individual member of the class that, by rejecting the exchange…he (or it) will be left out in the cold.”

The key will therefore be to establish whether there is in fact an acceptable level of coercion using the exit consent mechanism. This would fall well short of the relatively egregious nature of the coercion adopted by Anglo Irish, but still provide a real financial or legal incentive for bondholders to come into the exchange or tender offer.

Importantly, the court clearly rejected the argument that the problem of coercion is cured by providing clear and express disclosure on this issue, recognising that there seemed to be some misapprehension among market participants and legal practitioners.

This case illustrates that great care must be taken in drafting the provisions on bondholder powers in underlying legal documents to ensure that they are wide enough to allow bondholder rights to be extinguished. Perhaps more problematic will be ensuring that existing bond issues which are contemplating an exchange or tender offer are free from doubt on this particular point.

The case is subject to appeal and, until it is decided by the Court of Appeal, there will be a period of uncertainty and nervousness around this issue in the financial markets.

What are exit consents?

Exit consents are commonly used in bond exchange and tender offers as part of bond restructurings or so-called liability management exercises where a borrower wants to alter its debt maturity profi le, retire debt or replace expensive debt with cheaper debt.

The borrower invites bondholders to exchange or tender their bonds on condition that they commit to vote at the bondholder meeting in favour of extraordinary resolutions to amend the terms of the existing bonds. These resolutions typically include the insertion of an issuer call option to sweep up outstanding bonds and/ or strip out certain commercial or protective covenants such as enforcement rights or change of control protection.

The incentive to exchange or tender is that, ultimately, if the resolutions are passed, the minority bondholders who do not accept the exchange offer will be left with illiquid bonds of lesser fi nancial and/or legal value. Knowing that exit consents are on the table causes bondholders to think hard before adopting a hold out strategy.

This article first appeared in the September 2012 issue of PLC Magazine.