Contingent Convertible Bonds & AT1 Instruments : Does Banking Law Facilitate That Contingent Convertible Bonds Effectively Provide Sufficient Additional Capital In A Crisis To Banks On A ‘Going-Concern’ Basis?

(1)

C

ONTINGENT

C

ONVERTIBLE

B

ONDS

& AT1 I

NSTRUMENTS

:

D

OES

B

ANKING

L

AW

F

ACILITATE

T

HAT

C

ONTINGENT

C

ONVERTIBLE

B

ONDS

E

FFECTIVELY

P

ROVIDE

S

UFFICIENT

A

DDITIONAL

C

APITAL

I

N

A C

RISIS

T

O

B

ANKS

O

N

A ‘G

OING

-C

ONCERN

’ B

ASIS

?

B Y

J

E N S

B

E N E D I K T

K

A L B H E N N

Master’s Thesis in European Private Law, University of Amsterdam 

Amsterdam Graduate School of Law

Author Jens Benedikt Kalbhenn Student nr.

Email

Date 2 January 2017

(2)

Abstract:

After the financial crisis that commenced in 2007/08, various analyses of the roots of the market failure determined as one of the main factor of the crisis a lack of sufficient core capital of banks. This broad analysis, however, did not address one significant factor in explaining the financial crisis, which is the quality of the regulatory capital. Most rescued banks did have sufficient core capital, yet in spite of this banks lacked capital on a going-concern basis, severely limiting the changes of a bank to recapitalise on the market. Banks had to be bailed out by the national governments with funds provided by taxpayers.

Among the new tools to address this issue is the admission of contingent convertible bonds (CoCos) into banks’ regulatory capital. CoCos were conceived as a hybrid between debt and equity. A subordinated debt obligation that in normal times does not share in profits but in a crisis automatically shares in loses, by being converted into equity or written down, designed as a contractual agreement between banks and bondholders in advance of a crisis.

The question of this thesis is, whether the current European banking law facilitates that CoCos are designed in a way that they can effectively provide banks with sufficient additional capital in a crisis on a ‘going-concern’ basis? To answer the question, CoCos will be approached from an evaluative point, after having established the background of CoCos, its concept and how it is regulated on a normative level. The evaluation of the European regulations, as it pertains to the issuance of Additional Tier 1 instruments is split into four subsections, focusing on different aspects discussed in this thesis: the trigger design, the coupon cancelling procedure and extension risks, the amount of AT1 instruments that can be included in the regulatory capital, and the impact of the decision to make write downs permissible.

This thesis will show that the answer to the research question is ambivalent. The base regulation is essentially constructive; the contractual freedoms of the bond issuer and the bondholder are not limited in a detrimental way. However, the CRD IV framework does contain statutory provisions that introduce a number of uncertainties affecting the legal relationship between bond issuer and bondholder, thereby impeding a widespread use of this instrument and thus limiting its utility.

(3)

The opening quotations showcase both extends on the academic debate about contingent convertible bonds (CoCos), and the chasm in opinions that exists on this issue. Unlike the opening quotation suggests, the basic concept of CoCos is quite transparent, it is a financial instrument that functions as a regular bond issue or loan in normal times and converts in a “crisis”, a contractual pre-defined 3

event, either into common equity or is written down, thus providing an ailing bank with financial relief. This gives CoCos elements of debt and equity securities, making them a hybrid of both. This is essentially the definition that the International Monetary Fund applies to CoCos, which defines them, as “dated bonds with principal and scheduled coupon payments that can automatically [be] converted into equity […] or written down when a predetermined trigger event occurs, enabling a fresh injection of capital into a distressed bank”. The IMF’s definition is more expansive than the 4

original concept of CoCos and extends it to include write downs. It should be pointed out, that a conversion/write down does not directly provide new capital, but it does reduce a bank’s the debt burden, and presumably restores its ability to recapitalise on the private market. 5

Thomas Hale, Music stops for buyers of bank coco debt, The Financial Times, 11 February 2016.

1

Cf. Hilary J. Allen, Cocos Can Drive Markets Cuckoo, 16 Lewis & Clark Law Review, Volume 125, 2012,

2

p. 127.

As opposed to a regulatory trigger, where the conversion/write down is dependent on a decision by a

3

regulatory authority. Although, most definitions include securities with a regulatory trigger in their definition of CoCos.

Ceyla Pazarbasioglu, Jianping Zhou, Vanessa Le Leslé & Michael Moore, Contingent Capital: Economic

4

Rationale and Design Features, IMF Staff Discussion Note No. SDN/11/01, 25 January 2011, p. 4. Mark J. Flannery, Maintaining Adequate Bank Capital, 1 April 2013, p. 19.

(5)

The advent of this type of security can be traced back to the broad discussion following the aftermath of the financial crisis 2007/08. One common train of thought was that one main factor 6

leading to the crisis was a lack of sufficient core capital. Consequently, financial regulators determined that banks should be required to finance their assets with a greater amount of equity; reducing their leverage ratio and thereby enhancing the financial stability by reducing banks’ risk-taking incentives and increasing banks’ buffers against losses. 7

However, this broad analysis does not include one additional, significant factor in explaining the financial crisis; the quality of the regulatory capital. Most struggling banks had sufficient core capital, and were “adequately capitalized”, at least according to the requirements of the Basel accord provisions at that time. In spite of being “adequately capitalized” the banks lacked capital 8 on a going-concern basis. This severely limits the chances of a bank to recapitalise on the market. 9

Consequently, the banks had to be bailed out by the national governments with funds provided by taxpayers, stepping in to provide necessary capital. 10

To limit the necessity for banks to be bailed out by the governments, the Basel Committee on Banking Supervision (BCBS) established the Third Basel Accord (Basel III). Basel III includes 11

additional measures to ensure the stability of the banking system. Among the new standards is the admission of CoCos as regulatory capital. Of the total regulatory capital, CoCos can be included in the Additional Tier 1 (AT1) Capital, functioning as going-concern capital, or, can be included as the Tier 2 (T2) Capital, functioning as gone-concern capital. To be admitted as such, Cocos must 12

comply with the criteria, set out in the relevant Basel III provisions. The Basel III standards, itself

See: Allen, Cocos Can Drive Markets Cuckoo, p. 126; John C. Coffee, Jr., Systematic Risk After

Dodd-6

Frank: Contingent Capital and the Need for Regulatory Strategies Beyond Oversight, Columbia Law Review, Volume 111, No. 4, May 2011, p. 826; Flannery, Stabilizing Large Financial Institutions with Contingent Capital Certificates, 6 October 2009, p. 12.

Natalya Martynova, Effect of bank capital requirements on economic growth: a survey, De Nederlandsche

7

Bank, Working Paper No. 467, March 2015, p. 2.

Flannery, Stabilizing Large Financial Institutions with Contingent Capital Certificates, p. 12.

8

Shanker Merchant, Contingent Convertible Bonds for Bank Regulatory Capital, 5 September 2011, p. 1.

9

BCBS, Proposal to ensure the loss absorbency of regulatory capital at the point of non-viability, August

10

2010, p. 1.

BCBS, Basel III: A global regulatory framework for more resilient banks and banking systems,December

11

2010 (reviewed June 2011).

Stephanie Chan & Sweder van Wijnbergen, Coco Design, Risk Shifting and Financial Fragility, Tinbergen

12

(6)

having no ‘legal bite’ , were implemented on the European Union level by two principal 13

legislatives acts: the Capital Requirements Directive (CRD IV) and the Capital Requirements 14

Regulation (CRR). Another law relevant to CoCos is the Bank Recovery and Resolution Directive 15

(BRRD). 16

This gives rise to the question of this thesis: Does the European banking law, in particular the CRD IV framework (CDR IV and CRR) in conjuncture with the BRRD, does facilitate CoCos to be designed in a way that they can effectively provide banks with sufficient additional capital in a crisis on a ‘going-concern’ basis? In order to answer the question, I will approach my research question from an evaluative approach, after having established the background of CoCos, its concept and how it is regulated on the normative level by theEuropean legislation.

In this thesis I will show that the answer to this question is ambivalent. The base regulation is essentially constructive; the contractual freedoms of the bond issuer and the bondholder are not limited in a detrimental way. However, the CRD IV legislative package, does contain statutory provisions that introduce a number of uncertainties affecting the legal relationships between bond issuer and bondholder, thereby impeding a widespread use of this instrument and thus limiting its utility.

The thesis is divided into three chapters.

Chapter 1 focuses on the conceptual background of CoCos. This chapter contains a short description of the financial crisis and the demand this created for new and additional bail-in solutions. It also discusses the concept of CoCos, as it regards key design features. Furthermore, a comprehensive overview is given of how the distinct goals of CoCos determine the design of the

Chan & van Wijnbergen, Coco Design, p. 4.

13

Directive 2013/36/EU of the European Parliament and of the Council

14

of 26 June 2013on access to the activity of credit institutions and the prudential supervision of credit

institutions and investment firms, amending Directive 2002/87/EC and repealing Directives 2006/48/EC and 2006/49/EC.

Regulation (EU) No 575/2013 of the European Parliament and of the Council

15

of 26 June 2013on prudential requirements for credit institutions and investment firms and amending Regulation (EU) No 648/2012.

Directive 2014/59/EU of the European Parliament and of the Council

16

of 15 May 2014 establishing a framework for the recovery and resolution of credit institutions and investment firms and amending Council Directive 82/891/EEC, and Directives 2001/24/EC, 2002/47/EC, 2004/25/EC, 2005/56/EC, 2007/36/EC, 2011/35/EU, 2012/30/EU and 2013/36/EU, and Regulations (EU) No 1093/2010 and (EU) No 648/2012, of the European Parliament and of the Council.

(7)

legal terms and conditions of such instruments. It features a distinction between CoCos that aim to provide capital in a crisis on a ‘going-concern’ basis or a ‘gone-concern’ basis.

Chapter 2 addresses how CoCos are regulated by European law, it describes how the current regulation addresses CoCos. To answer the research question, the current European regulatory framework needs to be described. This includes the Basel III rules, although not European Union law, they serve as an underlying foundation for the CRR and CRD IV. The description of the supervisory framework, as it concerns CoCos, is focused on AT1 instruments. Accordingly, the BRRD is addressed only insofar as it pertains to AT1 instruments. The study of European law is limited to the above mentioned regulation and directives and executive acts — regulatory technical standards — based on them. Not addressed in this chapter, are company law aspects of CoCos, although both the European company law rules and the national company laws are relevant to 17 18

the issuing of CoCos, especially with respect to the mechanisms of the debt conversion into equity. Also not discussed are issues concerning the prospectus regime as it concerns CoCos. 19 20

Chapter 3 contains the evaluation of the European regulation as it pertains to the issuance of AT1 bonds and the research question. This section will make reference to circumstances surrounding specific AT1 bond issuances and specific market reactions.

This is followed by concluding remarks.

For example, since the conversion of CoCos into equity at the occurrence of a trigger event is an increase

17

of capital for the bank, Article 29 (1) of the Directive 2012/30/EU of 25 October 2012 is applicable, which provides that, any increase in capital of a public company must be decided by the general meeting.

In German company law § 221 (1) Aktiengesetz provides that if a creditor of a public company is entitled

18

to the conversion of stocks a resolution of the general meeting is mandatory.

For an overview of company law issues concerning CoCos, see: Andreas Cahn & Patrick Kenadjian,

19

Contingent Convertible Securities: From Theory to CRD IV, 2015 p. 29-55. For ethical concerns regarding CoCos and company law, about “One-share, one-vote”, see: Katie Bentel & Gabriel Walter, “Dual Class Shares”, Comparative Corporate Governance and Financial Regulation, Paper 2, 2016, p. 16.

Lizzie Meager, CoCos hit by new prospectus regime, International Financial Law Review, 9 May 2016,

20

(8)

CHAPTER 1: Conceptual Background and Design Features of CoCos

This chapter focuses on the conceptual background and design of CoCos. It is split into two sections. The first section contains a short description of the financial crisis and the demand it created for new and additional loss absorbing capital. It will feature a distinction between CoCos that aim to provide capital in a crisis on a ‘going-concern’ basis, or, a “gone-concern” basis. The second section will explain the different design elements, e.g. contractual elements of CoCos.

SECTION 1: The Financial Crisis and the Advent of CoCos

During the financial crisis that commenced in 2007/08, multiple governments had to bail out banks in their jurisdiction, after these banks experienced a sudden need of liquidity that the global markets were not willing to provide. Since private investors are commonly unwilling to provide external capital to banks in times of financial distress, such as in the extreme situations like the recent 21

financial crisis, governments ended up providing the necessary capital to prevent an insolvency of larger banks; because no one else was willing to do so. 22

Similarly to the unwillingness of the market to provide equity, the existing stockholders of a distressed bank were not willing to accept external capital, since this would have significantly 23

diluted their stock holdings in an unfavourable market condition and violated the interests of existing shareholders. This recapitalisation ‘gridlock’ forces a bank, that might otherwise still be 24

viable to go bankrupt or go into receivership. 25

Darrell Duffie, A Contractual Approach to Restructuring Financial Institutions, in: Ending Government

21

Bailouts as We Know Them, edited by G. Schultz, K. Scott & J. Taylor, Hoover Institute Press, 2010, p. 110; Coffee, Bail-Ins Versus Bail-Outs: Using Contingent Capital To Mitigate Systemic Risk, Columbia Law and Economics Working Paper No. 380, 22 October 2010, p. 22.

Stefan Avdjiev, Anastasia Kartasheva & Bilyana Bogdanova, CoCos: A Primer, BIS Quarterly Review,

22

September 2013, p. 43.

Cf. Charles W. Calomiris & Richard J. Herring; How to Design a Contingent Convertible Debt

23

Requirement That Helps Solve Our Too-Big-to-Fail Problem, Journal of Applied Corporate Finance, Volume 25, Number 2, 2013, p. 40. Describing that an executive of one of the banks that failed confessed that despite the need to replace lost equity, the price of his bank’s stock was “too low” to issue new shares.

Pazarbasioglu, Zhou, Le Leslé & Moore, Contingent Capital, p. 7; Markus K. Brunnermeier, Deciphering

24

the Liquidity and Credit Crunch 2007–2008, Journal of Economic Perspectives,Volume 23, Number 1,Winter 2009, p. 110.

Duffie, A Contractual Approach to Restructuring Financial Institutions, p. 116.

(9)

Other weaknesses of the banking regulation at that time were late responses to banks capital shortcomings due to the relying on regulatory discretion. Regulatory action came only when a 26

bank was experiencing severe funding troubles and market’s perception for that bank was that it was insolvent, making it virtual impossible to issue new shares and raise equity. 27

Two mechanisms to avoid the problems described above — inadequate regulatory capital and the necessity to rely on taxpayer provided funds, instead of finding a market solution — are the use of CoCos and regulatory bail-in of bank creditors (preferable bondholders instead of retail bank depositors). The difference between CoCos and the bail-in of bank creditors is that the first one is a market oriented instrument, based on a contractual agreement, while the latter is based on regulatory decisions and discretion. The two instruments should not be understood as alternatives, 28

but as complementary: CoCos as a first line of defence and bail-in as a secondary instrument, if the contractual conversion or write down of debt is not sufficient. One of the primary differences 29

between CoCos and bail-in of bank creditors is the quality of the capital. While both serve as loss absorbent capital, bail-ins only provide gone-concern capital, and Cocos also provide going-concern capital. Conceptually, bail-ins are only used at a point of non-viability, whose assessment requires an exercise of regulatory discretion. 30

Adequate regulatory capital can be distinguished between two approaches, a gone-concern approach, a regulatory approach prevalent before the crisis, and a going-concern approach. Both 31

approaches can be applied to CoCos. Gone-concern capital is debt designed to absorb losses when a financial institute is at a point of non-viability or insolvency, it can be thought of as a resolution

Flannery, Contingent Capital Instruments for Large Financial Institutions: A Review of the Literature,

26

Annual Review of Financial Economics 2014, , 11 November 2013, p. 6. Flannery, Contingent Capital Instruments: A Review of the Literature, p. 6.

27

Meta Zähres, Contingent Convertibles: Bankanleihen im Wandel, Deutsche Bank Research -EU-Monitor,

28

15 April 2011, p. 3.

Jianping Zhou, Virginia Rutledge, Wouter Bossu, Marc Dobler, Nadege Jassaud &Michael Moore, From

29

Bail-out to Bail-in: Mandatory Debt Restructuring of Systematic Financial Institutions, IMF Staff Discussion Note No. SDN/12/03, 24 April 2012, p. 6.

Simon Gleeson, Legal Aspects of Bank Bail-Ins, LSE Financial Markets Group Paper Series, Special

30

Paper 205, January 2015, p. 15.

Cf. The Financial Services Authority, A regulatory response to the global banking crisis, FSA Discussion

31

(10)

tool, usually at regulatory discretion. Going-concern capital operates well before resolution 32

mechanisms commence, given that (i) recapitalisation occurs when significant enterprise value still remains (especially if instruments are in place with a high trigger, see Section 2) and (ii) the conversion/write down mechanism applies automatically, when the trigger event occurs. 33

Another way to differentiate these two approaches of loss absorbent capital is the moment when they take effect. Going-concern capital can be thought of as an an ex ante measure, which is intended to reduce the chance that a bank defaults on its obligations, while gone-concern capital is an ex post measure meant to reduce the fallout and loss upon default, if a bank’s ex ante measures do not succeed in providing a sufficient level of equity for this bank. 34

When applying a going-concern approach to CoCos, the economic rationale for the inclusion of CoCos into banks regulatory capital includes the following arguments:

CoCos provide an automatic mechanism to increase the bank’s capital and to reduce the debt and debt to equity ratio. This is archived in two ways, firstly the commencement of a conversion/write down is independent of regulatory discretion, it sets in automatically after a trigger event occurs, without the necessity to rely on a discretionary decision, secondly it is equally independent of managerial discretion, since it avoids holdouts to issue new equity from executives hoping to recapitalise later at more favourable market conditions, witch is seen as less detrimental to the stock holding of existing shareholders.

From the bank’s perspective, CoCos might be preferable to equity since it is ordinarily, depending on pricing, but not without exceptions cheaper. Generally, interest payments on debt and bond 35

issues are tax deductible, while payments to stockholders, either directly through dividends or indirectly by share buybacks are not tax deductible. 36

Charles P. Himmelberg, Amanda Hindlian, Sandra Lawson & Louise Pitt, Contingent capital -

32

Possibilities, Problems and Opportunities, Goldman Sachs Global Markets Institute, March 2011, p. 3. Himmelberg, Hindlian, Lawson & Pitt, Contingent capital, p. 4; Yves Courtois, Doug McPhee & Jean

33

Florent Rerolle, Incentive effects of contingent capital, KPMG Global valuation Institute, Fourth Managerial Paper, 2013, p. 1.

Cahn & Kenadjian, Contingent Convertible Securities, p. 2.

34

Cf. Pazarbasioglu, Zhou, Le Leslé & Moore, Contingent Capital, p. 7.

35

Marc Rüdlinger, Contingent Convertible Bonds: An Empirical Analysis of Drivers and Announcement

36

(11)

An additional rationale behind CoCos is that they could help mitigate risk taking by managers, shareholders and bondholders. If the conversion of CoCos into equity leads to a potential wealth 37

transfer — depending on the conversion ratio (see Section 2), by converting the debt of previous debt holders into equity, thereby diluting the holdings of existing shareholders — then the shareholders will be incentivised to intercept in order to stop that the equity level reaches the trigger level, thus decreasing risk and leverage. This conceivably motivates shareholders to pressure the 38

management to take necessary action at an early stage. This argument is only valid if the conversion does not lead to a write down, which would reverse the logic behind this argument of shareholders incentive to pressure the management for decreasing debt. 39

With the use of CoCos, the costs of new capital is internalised. This is caused by the fact that 40

bondholders as well as equity holders are forced to participate in the recovery phase. When 41

successful, the need of public bail-outs is removed, and if the recovery is unsuccessful, a bail-out would occur only after the conversion of bondholders debt claims against the bank into capital, thus reducing the costs to the public.

A point of contention is whether the use of CoCos decreases or actually increases the systemic risk of a financial crisis. The argument that systemic risks decrease is that CoCos prevent or at least reduce ‘fire sales’ of assets by over-leveraged banks. Since fire sales have a detrimental effect on macro-financial stability, the use of CoCos would reduce systematic risks. 42

The argument of an increase in systematic risks is twofold; that the occurrence of the trigger has signalling effects and that the interconnectedness of banks holds inherent dangers in case of conversion / write down of bond issues. Proponents of the negative signalling effect argue that the conversion sends the negative signal to depositors about the expected return of deposits and thus

Zhou, Rutledge, Bossu, Dobler, Jassaud & Moore, From Bail-out to Bail-ins, p. 7.

37

Coffee, Systematic Risk After Dodd-Frank, p. 828; Coffee argues from the assumption that it was

38

shareholder pressure on management to significantly increase leverage and not the influence of executive compensation. It can, however, be argued that Coffee’s argument is focused on the rationale of the regulation specific to the United States of America, see: Bart P. M. Joosen, Bail In Mechanisms in the Bank Recovery and Resolution Directive, 19 October 2014, p. 5.

Coffee, Systematic Risk After Dodd-Frank:, p. 829; Coffee makes reference to one of the first meaningful

39

CoCo bond to be issued: the Rabobank issuance of 2010, which if conversion was triggered would lead the investors to lose 75% of their investment.

Boris Radnaev, A Model to Banks’ Tier 1 Capital with contingent Securities, London Business School,

40

Working Paper Series, August 2015, p. 2.

Zähres, Contingent Convertibles: Bankanleihen im Wandel, p. 3.

41

Flannery, Contingent Capital Instruments: A Review of the Literature, p. 4.

(12)

increases the likeliness of a bank run, spilling over to other banks if they have correlated assets, even if they have not issued CoCos themselves. 43

Another potential danger is the interconnectedness of banks. Banks may hold CoCos from other issuers, leading to the undesired state that banks become large equity holders in other banks.Such cross-holdings may increase systematic risks. This argument is based on the assumption that 44

banks would hold large amounts of CoCos. 45

SECTION 2: The Design of CoCos and Contractual Decisions

This section focuses on the two main elements of any CoCo bond, and how their design determines the effectiveness of CoCos as an element of loss absorbing regulatory capital: First, the design of the trigger event, when conversion or write down will take place, and second, how the conversion of debt into equity is arranged. Different decisions will impact the incentive effects on the stockholders, the possibilities of market manipulation, and the quality of the loss absorbency of the capital (going or gone-concern capital).

The Trigger

The trigger design has to contain three decisions. Should it be a contractual or a regulatory trigger? Which metric is used to determine if the trigger event has occurred? And ultimately at what stage should conversion into equity or write down of CoCos occur?

The rationale behind the choice between a contractual or regulatory trigger was explained in the previous section. The benefit of a contractual trigger is the automatism that triggers the conversion, while making the trigger subject to a discretionary regulatory decision leads to uncertainty and delays. The Basel Committee argues that the determination of the occurrence of a trigger events should be determined by the “relevant authority”, the Committee itself followed at that point a gone-concern approach. The benefits of a contractual, automatic trigger is that it is transparent, 46

Chan & van Wijnbergen, Cocos, Contagion and Systemic Risk, Tinbergen Institute Discussion Paper

43

14-110/VI/DSF 79, 30 September 2014, p. 7.

Jill Khoury, Contingent Convertible Capital: A New Capital Requirement to “Solve Too Big to Fail”, April

44

2016, p. 41; Karl-Philipp Wojcik, Bail-In In The Banking Union, Common Market Law Review, Volume 53, 2016, p. 130; Wojcik continues and emphasises that bail-ins through conversion would lead to more

contagion, thus increasing the systematic risks.

Although it should be noted that according to the BIS Quarterly Review, “The bulk of the demand for

45

CoCos has come from small investors, while institutional investors have been relatively restrained so far”; see: Avdjiev, Kartasheva & Bogdanova, CoCos: A Primer, p. 43.

Cf. BCBS, Proposal to ensure the loss absorbency of regulatory capital, p. 6. The Basel Committee argues

46

(13)

objective, faster — due to the lack of discretion — and reduces uncertainty, which would quell the 47

interest of investors, not wanting to give the regulatory authority the power to convert this debt claim into equity or write the claim down altogether. 48

Two frequently discussed metrics to determine the triggers are a book-value trigger or a market-49

value trigger. Both provide different up and downsides. A market-value trigger does not have the disadvantage of inconsistent accounting, and, if based on the ratio of the bank’s market capitalisation to its assets, is very transparent. A potential drawback appears to be its susceptibility to become a market manipulation target. However, this can be circumvented by basing the trigger 50

on a rolling average of the stock price. A book-value, or accounting-value trigger, is set 51

contractually as a ratio of the core capital book value against the risk-weighted assets. The transparency of a book-value trigger is dependent on the frequency and accuracy how these metrics are calculated and disclosed. 52

The final decision is when to convert, the rationale is that an early conversion would take place when the firm still has significant enterprise value (high trigger), whereas a late conversion would occur close to a point of non-viability (low trigger). Thus, providing either going-concern or gone-concern capital. The benefit of a low trigger, and vice versa the drawback of a high trigger, is that it would be more attractive as an investment due to its lower conversion or write down probability, and additionally lowering the risk premium, the coupon, banks would have to pay. A high trigger, 53

occurring when the health of the firm is considerable more stable, would raise the possibility that the bank recovers more quickly and reduces systematic risks. 54

Himmelberg, Hindlian, Lawson & Pitt, Contingent capital, p. 7.

47

Jan de Spiegeleer & Wim Schoutens, Pricing Contingent Convertibles: A Derivatives Approach, 12 June

48

2011, p. 5.

Cf. Coffee, Systematic Risk After Dodd-Frank:, p. 831, Coffee describes different bases for a trigger, for

49

example a rating downgrade.

Coffee, Systematic Risk After Dodd-Frank:, p. 841; Pazarbasioglu, Zhou, Le Leslé & Moore, Contingent

50

Capital, p. 9; Avdjiev, Kartasheva and Bogdanova, CoCos: A Primer, p. 45. Flannery, Contingent Capital Instruments: A Review of the Literature, p. 17.

51

Avdjiev, Kartasheva & Bogdanova, CoCos: A Primer, p. 44.

52

Zähres, Contingent Convertibles: Bankanleihen im Wandel, p. 7.

53

Cf. Khoury, Contingent Convertible Capital, p. 25.

(14)

The Conversion Ratio

A consequential decision is what happens after the trigger event occurs. Is the debt converted to equity in the bank or are parts or the whole debt written down? This determines the burden sharing between debt and equity holders of the bank. In this respect it should be noted that some banks 55

cannot issue debt that converts into equity because their legal form does not allow them to issue stocks to the public. 56

The conversion into equity can be arranged in different ways, the debt can convert into a fixed number of shares, or a fixed share value, with varying effects on the rights and obligations of bondholders and shareholders. For CoCos to work as an ex ante instrument — incentivising the 57

management of the bank to prevent a trigger event — the conversion has to lead to a dilution of the stocks, otherwise the shareholders would have less reasons to pressure the management to reduce leverage and risks that might lead to a conversion of CoCos. Thus, they are lowering the risk of 58

bank failures and of systematic risks.

All principal write down CoCos, as well as those bonds without a sufficient dilution effect on the existing capital, have substantially worse risk shifting incentives than a requirement for common equity would have. The desirable incentive effects of CoCos would simply be reversed, as they 59

would add value or at least not be detrimental to the shareholders of the bank. 60

It can be argued that without the dilutive effect a material rationale of CoCos vanishes, since “[t]o the common shareholder, contingent capital holds out the prospect of death by dilution, and it can be anticipated that shareholders would task management to undertake the necessary measures to avoid such dilution.” 61

Pazarbasioglu, Zhou, Le Leslé & Moore, Contingent Capital, p. 11.

55

Some banks particular in Germany and the Netherlands are not organised as a publicly held companies, let

56

alone admitted and traded on a stock exchange. In Germany this includes: Sparkassen, Volksbanken and Raiffeisenbanken and the remaining Landesbanken. In the Netherlands: the local Dutch Rabobanks and the Rabobank Nederland.

Merchant, Contingent Convertible Bonds, p. 3.

57

Cf. Coffee, Systematic Risk After Dodd-Frank, p. 831.; Coffee, Bail-Ins Versus Bail-Outs, p. 10; Calomiris

58

& Herring, Why and How to Design a Contingent Convertible Debt Requirement, April 2011, p. 6. Chan & van Wijnbergen, Coco Design, p. 6.

59

Courtois, McPhee & Rerolle, Incentive effects, p. 5.

60

Thomas Huertes, as quoted in: Calomiris & Herring; How to Design a Contingent Convertible Debt

61

(15)

CHAPTER 2: CoCos in the Third Basel Accord and Its European Implementation

The regulatory environment of CoCos will be expounded upon by showing how the Basel III standard and the main European regulations, implementing the Basel Committee’s reform, 62 63

address this type of regulatory capital.

SECTION 1: Basel III

Under the Basel III rules, the Committee’s framework for the post-crisis reforms, CoCos are able 64

to be included within the total regulatory capital, provided that these CoCos comply with the relevant provisions of Basel III. 65

Basel III provides that the total regulatory capital of any bank will consist of “Tier 1 Capital (going-concern capital)”, composed of Common Equity Tier 1 (CET) and Additional Tier 1 (AT1), and Tier 2 Capital (gone-concern capital). With the CET having a ratio of at least 4,5% of Total Risk 66

Exposure Amount (TREA) at all times, T1 Capital of at least 6% of TREA and Total Capital of at 67

least 8%. 68

Additionally, all banks are required to have a capital conservation buffer of 2,5% above these minimal requirements. Furthermore, all banks may be subject to an individual countercyclical 69

buffer of up to 2,5%. 70

The Basel III regulation is certainly not directly legally binding in any jurisdiction, its authority is just of

62

an advisory nature. However, member states of the Basel accord had pledged to implement the Basel III provisions in their respective jurisdiction. Cf. Gregg Rozansky, The Loss Absorbency Requirement and “Contingent Capital” Under Basel III, EuroWatch 2011, Volume 23 No. 5, p. 3.

As early as 2010 did the Bank for International Settlement announce that the Basel Committee would

63

review the use of CoCos within its upcoming reform, Basel III, to be counted within the regulatory capital framework, BCBS, Proposal to ensure the loss absorbency of regulatory capital at the point of non-viability, August 2010.

BCBS, Basel III: A global regulatory framework for more resilient banks and banking systems,

64

(hereinafter: Basel III).

Jan de Spiegeleer, Stephan Höcht &Wim Schoutens, Are banks now safer? What can we learn from the

65

CoCo markets?, 20 August 2015, p. 1. Basel III, § 49.1.

66

The Base Committee uses the term Risk-Weighted Assets (RWA).

67

Basel III, § 50.

68

Basel III, § 129.

69

Basel III, §§ 136 et seq.

(16)

The use for going-concern CoCos is limited to AT1 Capital of up to 1,5% of TREA (assuming there is no incentive for banks to have an optional higher TREA ratio). Gone-concern CoCos are limited to 2% TREA within T2 Capital. The use of CoCos is therefore limited to 3,5 % of up to 13% TREA (including the conservation buffer and assuming an individual countercyclical buffer of 2,5%). To qualify as AT1 Capital CoCos have to fulfil certain requirements. The minimum trigger level (CET/TREA) has to be set at or above 5,125% and to include a trigger at a point of non-viability as well. The loss absorption, i.e. conversion, must be either through conversion to common shares or 71

write down of principal and accrued interest; the write down must have the effect of reducing the claim of the bond and reduce the amount paid when a call is exercised and reduce the coupon 72

payments on the bond (§ 55.11 of Basel III).

Moreover, CoCos have to fulfil the requirements laid down in § 55 Basel III. These requirements include that CoCos are perpetual — with no maturity date and no step-ups or other incentives to redeem the bond, (§ 55.4 of Basel III), constitute subordinated debt to depositors and to general creditors and to subordinated debt of the bank (§ 55.2 of Basel III), thus becoming subordinated to even other subordinated debt. Furthermore, any repayment of the principal can only occur after regulatory approval, and the terms of the bond agreement must state, that banks may not assume that approval will be granted (§ 55.6 of Basel III). The bank can first call the issue after a minimum of five years after having received regulatory approval (§ 55.5 of Basel III). Payment of the coupon is fully discretional for the bank, with cancellation of the payment not being designated an event of default (§ 55.7 of Basel III). Payment is limited to be paid out of the distributable items.

If CoCos do not fulfil the requirements to be included in AT1 Capital, they may be included in T2 Capital, provided that the requirements of § 57 of Basel III are met. With regard to the design of the trigger, a T2 bond only has to have a point of non-viability trigger. Overall the requirements are 73

less restrictive than the requirements for AT1 Capital. For instance, the bond does not have to be perpetual, it is not subordinate to subordinate debt of the bank, and the requirement of AT1 Capital, that non payment is no event of default, does not apply to T2 Capital.

Avdjiev, Kartasheva & Bogdanova, CoCos: A Primer, p. 48. It has to be noted, that Basel III in itself does

71

not require that the minimum level for the trigger is to be 5,125%. § 55.11 just provides for a “pre-specified trigger point”; cf. European Banking Authority, Buffer Convertible Capital Securities Common Term Sheet, 8 December 2011.

Call references here to a call option. The contractual option to redeem a bond prior to maturity, means that

72

the issuer can return the investor's principal and stop interest payments. Avdjiev, Kartasheva & Bogdanova, CoCos: A Primer, p. 46.

(17)

SECTION 2: European Regulation: CRD IV Framework and BRRD

The Basel III proposals were transposed into European Union law through separate legislative acts, the CRR, and the CRD IV. The CRR, which lays down eligibility criteria for AT1 instruments, was later supplemented with criteria concerning the AT1 instruments by the Commission Delegated Regulation No 241/2014 (Delegated Regulation No 241/2014), affecting the provisions AT1 bonds need to comply with. These legislative acts form the CRD IV framework. The CRD IV 74 75

framework cannot be read isolated from the BRRD, the directive concerning the recovery and resolution of banks. Although the BRRD is primarily concerned with regulatory bail-ins of creditors on a gone-concern basis at the point of non-viability, various provisions are relevant for CoCos, e.g. provisions concerning write downs and conversion of debt instruments. 76

The CRD IV Framework in General

The main Basel III provisions were implemented by the CRR. The European legislator decided to implement main provisions of Basel III in the form of a regulation instead of a directive, thus harmonising the rules within the European Union without the requirement to be implemented by the member states themselves. 77

The regulation itself cites legal certainty, the need for a level playing field within the Union, and a single set of regulations for market participants as a key element for the functioning of the internal market as reasons to legislate the topics by regulation. Another reason given, is to avoid market 78

distortions and regulatory arbitrage and to ensure maximum harmonisation with minimum requirements. The regulation was deemed necessary to remove obstacles to trade and competition 79

distortion resulting from divergences between national laws. 80

European Banking Authority (EBA), On the monitoring of Additional Tier 1 (AT1) instruments of EU

74

institutions, EBA Final Report, 29 May 2015, p. 2; Commission Delegated Regulation (EU) No 241/2014 of 7 January 2014 supplementing Regulation (EU) No 575/2013 of the European Parliament and of the Council with regard to regulatory technical standards for Own Funds requirements for institutions.

Cf. Cahn & Kenadjian, Contingent Convertible Securities, p. 18.

75

76

Article 288 Treaty on the Functioning of the European Union (2007).

77

CRR, recital 9.

78

Another way of interpreting this is that it was intended to prevent ‘gold plating’ by the United Kingdom

79

and those member states wishing to expand upon the CRD IV requirements. Cf. Cahn & Kenadjian, Contingent Convertible Securities, p. 18.

CRR, recital 11.

(18)

Another result from choosing the legislative form of a regulation is its direct horizontal effect. If the provisions of a regulation are sufficiently clear, precise, and relevant to the situation of an individual litigant, they are capable of being relied upon and enforced by individuals before their national courts. Therefore, the provisions of the CRR — provided they are sufficiently clear, 81

precise, and relevant to the situation — can directly effect a contractual relationship, even the validity of an agreement, among its parties, e.g. the contractual relationship between a bank issuing CoCos and the bondholder. 82

The Specific Provisions in the CRD IV Framework Relevant to CoCos

The CRD IV legislative package adheres closely to the Basel III standards when it comes to matters regarding CoCos. Similarities include the distinction between more equity-like requirements for perpetual AT1 capital (going-concern) and the more debt-like requirements for dated T2 CoCos. 83

CoCos are the only type of instrument that may be included within AT1 capital under the CRD IV framework. Like Basel III, AT1 instruments can make up parts of the minimum risk-based capital, 84

or in the legislative language of the CRD IV framework ‘own funds’. It can make up to 1,5% of 85

the 6% minimum T1 Capital ratio and, since AT1 instruments can under certain circumstances count towards T2 Capital, up to 3,5% of the 8% minimum total capital ratio. 86

The requirements for capital instruments to qualify as AT1 capital are set out in Chapter 3, Section 1 of the CRR (Article 51-55). This includes the following provisions: The instrument is issued and paid up (Article 52 (1) (a) CRR) and the instrument ranks below T2 instruments in the event of

Paul Craig & Gráinne de Búrca, EU Law Text, Cases and Materials, sixth edition, 2015, Oxford, p. 184.

81

One example, of how the CRR influenced a national court decision, is the United Kingdom Supreme Court

82

judgment in BNY Mellon Corporate Trustee Services Limited (Appellant) v LBG [Lloyds Banking Group] Capital No 1 Plc and another (Respondents), [2016] UKSC 29. The relevant facts of the case can be summarised, as such: In November 2009 LBG issued an Enhanced Capital Note (ECN, a CoCo bond in all but name) with a conversion trigger set at 5% (1% higher than the 4% minimum requirement for Core Tier 1 capital at the time of the issuance) in order to fulfil core capital requirements. After the minimum

requirement for the heigh of the trigger was raised to 5,125%, the bond issue was ineligible to qualify as regulatory capital, as AT1 Capital, LBG called the bond. The Appellant, the security trustee representing the investors, sued against this decision of LBG but lost before the UK Supreme Court, since “[u]nder the Regulations passed in 2013, the ECNs cannot be taken into account so as to do the very job for which their convertibility was designed, namely to enable them to be converted before the regulatory minimum Tier 1 ratio is reached” (para. 45).

83

Bank of England, Financial Stability Report, Issue No. 35, June 2014, p. 33.

84

Article 3 (1) (119) CRR.

85

Article 92 (1), 92 (3) CRR.

(19)

insolvency (Article 52 (1) (d) CRR). Since an AT1 instrument is only junior to T2 instruments, it follows that AT1 capital instruments are senior to CET1 capital in the event of insolvency. 87

Furthermore, it has to be an unsecured instrument that is not subject to a guarantee that enhances the seniority of the claim (Article 52 (1) (e) CRR). The instruments are perpetual and there is no contractual incentive for the bank to redeem them (Article 52 (1) (g) CRR). The instrument may only be called, redeemed or repurchased five year after the issuance with prior permission of the competent authority (Article 52 (1) (i) and Article 77 (b) CRR). The competent authority shall 88

grant permission only in two cases: When the bank replaces the instrument with an instrument of equal or higher quality at sustainable terms or when the bank can demonstrate that its own funds exceed the capital requirements set out in Article 92 (1) CRR and its capital buffer requirements. 89

The payment of interest is limited to be paid out of the distributable items. When discussing 90

interest payments, the concept of the Maximum Distributable Amount (MDA) needs to be raised. Concerning the cancellation of interest payments, one must take into account that a bank that does not meet its Combined Buffer Requirement (CBR), Article 128 (6) CRD IV, is prohibited from making distributions on CET1 capital that would result in a breach of its CBR, Article 141 (1) CRD IV. If a bank fails to meet this requirement, then payment on AT1 instruments is prohibited as 91

well, Article 141 (2)(c) CRD IV, until the bank calculates the MDA and notifies the competent authority. As long as a bank fails to meet or exceed its CBR, it is prohibited from distributing 92

more than the MDA. According to Article 141 (3) CRD IV, this prohibition applies also to AT1 but not T2 instruments. Any distribution results in an automatic reduction of the MDA. The purpose of this limitations is to ensure that the distribution of profits does not jeopardise the capital position of

Joosen, Regulatory capital requirements and bail in mechanisms, in: Matthias Haentjens & Bob Wessels

87

(Editors), Research Handbook on Crisis Management in the Banking Sector, 2015, p. 205.

This includes among others: Call options combined with an increase in the credit spread if the call is not

88

exercised, call options combined with either a requirement or an investor option to convert the AT1

instrument into a CET1 instrument if the call is not exercised, call options combined with an increase of the redemption amount in the future; Article 20 (2) Delegated Regulation No 241/2014.

Article 78 (1) CRR.

89

Article 52 (1) (l) (i) CRR and Article 4 (1) (110) CRR define distribution as the payment of dividends or

90

interest in any form. Distributable items are defined as the amount of profits at the end of the last financial year plus any profits brought forward and reserves available for that purpose, Article 4 (1) (128) CRR.

Kamil Liberadzki & Marcin Liberadzki, Hybrid Securities: Structuring, Pricing and Risk Assessment,

91

2016, p. 38.

Competent authority is defined as a public authority or body officially recognised by national law, which is

92

empowered by national law to supervise banks as part of the member states supervisory system, Article 4 (1) (40) CRR.

(20)

a bank. Equally to the Basel III standards, the cancellation of payments does not constitute an 93

event of default, Article 52 (1) (l) (iv) CRR.

Most importantly, AT1 instruments are subject to a contractual provision providing for either a conversion or write down of the bond when a trigger event occurs. The trigger event requirements 94

are set out in Article 54 CRR. The trigger event occurs when the CET1 capital ratio of a bank, referred to in Article 92 (1) (a) CRR, falls below either 5,125% or below a higher discretionary level previous agreed to in the provisions of the bond issue. The minimum trigger level is set at 5,125%. This is 0,625 percentage points higher than the minimum CET1 capital level required by Article 92 (1) (a) CRR. Whether the minimum trigger level was chosen to represent a summation of the minimum CET1 ratio and the capital conservation buffer, equalling 0,625 % CET1 capital of the TREA required for the period from 1 January 2016 until 31 December 2016, is arguable, since the decision for setting the minimum trigger level at 5,125% is not discussed in the legislative materials. 95

When the provisions of the CoCo bond require a conversion, then the debt obligation is converted into CET1 capital. The conditions for capital instruments to qualify as CET1 capital are established in Article 28 CRR. Considering the fact that these instruments must be classified as equity, must rank below all other claims in insolvency and are prohibited from having a preferential distribution treatment provision, the debt is essentially converted into common shares. Article 54 CRR further stipulates that in case of a conversion both the holders of AT1 instruments and the competent authority are to be informed immediately. The AT1 instrument issuing bank is required to provide that its authorised share capital is at all times sufficient and that there are no procedural impediments to a conversion by virtue of its incorporation, statutes or contractual arrangements. Concerning write down provisions governing the instrument, the CRR stipulates that they are required to reduce the claim on both principal and interests of the instrument in insolvency or liquidation. Relevant for both write down and conversion is Article 54 (4) (a) CRR. It requires that the conversion fully restores the CET1 ratio to 5,125%. This requirement seems redundant, since a conversion is triggered automatically when the 5,125% trigger event occurs.

B. Mesnard & M. Magnus, Briefing: What to do with profits when banks are undercapitalized: Maximum

93

Distributable Amount, CoCo bonds and volatile markets, European Parliament Economic Governance Support Unit, 18 March 2016, p. 1.

Article 52 (1) (n) CRR.

94

Joosen, Regulatory capital requirements, p. 218.

(21)

The procedures and the timing for the determination when a trigger event has occurred are set out in Article 22 of the Delegated Regulation No 241/2014. When the bank has established a CET1 ratio below the level that activates either conversion or write down, its management body is required to determine the occurrence of the trigger event without delay and shall be under the irrevocable obligation to either convert the instrument into equity or write down the debt obligation.

It is notable that the CRR provision permits the inclusion of a multitude of non-compulsory trigger events. This is a clear novelty compared to the Basel III regime. Although there is seemingly no 96

explanation for this either in the CRR or its background information, it is not an original concept to issue CoCos with multiple trigger events. Coffee suggests that from a political perspective, it 97

makes sense to prefer multiple incremental conversions to a single major conversion, in order to reduce market shocks and diminish the impact of market’s anticipatory reactions to each incremental conversion. 98

The BRRD and CoCos

From a conceptual perspective, AT1 instruments and the BRRD — dealing primarily with resolution and recovery planning and regulatory ‘bail-in’ powers, after a national authority has determined that the bank is failing or likely to fail — should only be examined together to discuss 99

inherent differences in the underlying concepts. This concerns in particular gone versus going-concern capital. In practice, the BRRD contains various provisions that effect CoCos.

Under Basel III, AT1 instruments need, in addition to a going-concern trigger, a trigger at the point of non-viability of the bank. This requirement cannot be found in the provisions of the CRD IV 100

‘[I]nstitutions may specify in the provisions governing the instrument one or more trigger events in

96

addition to that referred to in point (a)’, Article 54 (1) (b) CRR.

Cf. De Spiegeleer & Schoutens, Multiple Trigger CoCos: Contingent Debt Without Death Spiral Risk,

97

Financial Markets, Institutions & Instruments, Volume 22, Issue 2, May 2013, p. 140.

Coffee, Systematic Risk After Dodd-Frank, p. 830. This addresses to some degree a critic addressed by

98

Allen, that because trigger events are low-probability and conversion into equity is a fundamental and irreversible change to CoCos, the perception that a trigger event is likely to occur, will incentive

stockholders, bondholders and third parties to engage in panic selling and shorting activities; cf. Allen, Cocos Can Drive Markets Cuckoo, p. 156.

BRRD, recital 6.

99

Basel III, § 49. 2 making reference to an earlier proposal, that ‘[a]s set out in the Committee’s August

100

2010 consultative document, Proposal to ensure the loss absorbency of regulatory capital at the point of non-viability, and as stated in the Committee’s 19 October 2010 and 1 December 2010 press releases, the

Committee is finalising additional entry criteria for Additional Tier 1 and Tier 2 capital. Once finalised, the additional criteria will be added to this regulatory framework.’

(22)

framework, but in the BRRD. Article 55 BRRD requires member states that they require banks to 101

‘include a contractual term by which the creditor or party to the agreement creating the liability recognises that liability may be subject to the write down and conversion powers and agrees to be bound by any reduction of the principle or outstanding amount due, conversion or cancellation that is effected by the exercise of those powers by a resolution authority, […]’. The provisions 102

determining the time and procedure of the ‘bail-in’ of AT1 instruments can be found throughout the BRRD, since the directive uses a system of cross references.

Only recital 81 deals expressly with the conversion or write down of AT1 instruments and defines the point of non-viability stating: Member States should ensure that Additional Tier 1 and Tier 2 capital instruments fully absorb losses at the point of non-viability of the issuing institution. Accordingly, resolution authorities should be required to write down those instruments in full, or to convert them to Common Equity Tier 1 instruments, at the point of non-viability and before any resolution action is taken. For that purpose, the point of non-viability should be understood as the point at which the relevant authority determines that the institution meets the conditions for resolution or the point at which the authority decides that the institution would cease to be viable if those capital instruments were not written down or converted. The fact that the instruments are to be written down or converted by authorities in the circumstances required by this Directive should be recognised in the terms governing the instrument, and in any prospectus or offering documents published or provided in connection with the instruments.

It should be noted that the the provisions regarding the resolution tools of the national authorities, especially ‘bail-in tool’ (Chapter IV, Section 5 BRRD) apply to AT1 instrument as well, regardless if a point of non-viability has been reached.

Without elaborating details of the bail-in mechanism, including conversion and write down of debt, and its procedure (e.g. Article 48 regarding the sequence of write downs and conversion), a few remarks are necessary.

Although recital 45 of the CRR declares, that ‘[i]n line with the decision of the BCBS, as endorsed by the

101

GHOS on 10 January 2011, all additional Tier 1 and Tier 2 instruments of an institution should be capable of being fully and permanently written down or converted fully into Common Equity Tier 1 capital at the point of non-viability of the institution.’

By requiring banks to provide for a contractual recognition between itself and bondholders of the

102

statutory ‘bail-in’ power of the resolution authority, the BRRD has provided for a ‘hybrid’ trigger. Being both of contractual and statutory law nature, thus eliminating the need to determine if the ‘bail-in’ power of the BRRD have direct effect. Cf. Joosen, Regulatory capital requirements, p. 226.

(23)

While a cursory reading of the directive might lead to the understanding that the competent authority may only apply the resolution tools, including bail-ins, in case a bank is either failing or likely to fail (recital 6), thus providing financial relief on a gone-concern basis, the directive itself provides for resolution objectives that are going-concern objectives. Article 43 (2) BRRD states that the resolution authority may apply the bail-in tool to meet the resolution objectives specified in Article 31, one of these objective is to ensure the continuity of ‘critical functions’, defined in Article 2 (35) BRRD. This definition is supplemented by a delegated regulation. The delegated 103

regulation requires that ‘critical functions’ must meet two conditions: Firstly they are being provided to third parties, and secondly a failures would give rise to contagion or undermine the general confidence of market participants due to the systemic relevance of the function for third parties and the systemic relevance of the bank in providing this function.

These critical functions can include deposit taking, lending and loan services, payment, clearing, custody and settlement services, wholesale funding markets activities, and investments activities (recital 4). The continuity of those functions is in essence a going-concern objective. 104

Another direct reference to going-concern is made in recital 68: “In order to ensure that resolution authorities have the necessary flexibility to allocate losses to creditors in a range of circumstances, it is appropriate that those authorities be able to apply the bail-in tool both where the objective is to resolve the failing institution as a going-concern if there is a realistic prospect that the institution’s viability may be restored, (…).”

With these two examples, preservation of ‘critical function’ and the objective of recital 68, a conceptual clear-cut distinction between the resolution of a failed bank (gone-concern) using the statutory tools of the BRRD and the recapitalisation of a still vital bank (going-concern) with contractual provisions of CoCos is dispelled. 105

Commission Delegated Regulation(EU) 2016/778 of 2 February 2016 supplementing Directive 2014/59/

103

EU of the European Parliament and of the Council with regard to the circumstances and conditions under which the payment of extraordinary ex post contributions may be partially or entirely deferred, and on the criteria for the determination of the activities, services and operations with regard to critical functions, and for the determination of the business lines and associated services with regard to core business lines.

Cf. Joosen, Regulatory capital requirements, p. 224; with regard to payment systems.

104

Cf. Emilios Avgouleas & Charles Goodhart, Critical Reflections on Bank Bail-ins, Journal of Financial

105

(24)

CHAPTER 3: Evaluation of CoCos

While AT1 instruments are appreciated by most banks for allowing funding and satisfying their regulatory capital requirements at a lesser cost than equity, the questions of their significance in a 106

crisis for a bank on a going-concern basis remains. This chapter will contain an evaluation of the European regulation, as it pertains to the issuance of AT1 instruments. To answer the question if the current European regulation indeed facilitates that CoCos are to be designed in a way that they can effectively provide banks with sufficient additional capital in a crisis on a ‘going-concern’ basis, a scrutinisation of the legal provisions apropos the conceptual background of CoCos is necessary, and a scrutinisation of how the provisions have affected the actual contractual design of AT1 bonds. For that matter, the chapter is split into four subsections, focusing on different aspects discussed in the previous chapters: The design of the trigger (i), the procedure of coupon cancelling and extension risks (ii), the amount of AT1 instruments to be included as regulatory capital (iii), and finally the impact of the decision to make write downs permissible (iv). This chapter will take market reactions regarding specific issuances into account. It will give particular emphasis on how the regulation relates to the topic of legal certainty.

1. Do the CRR provisions for AT1 bonds feature a trigger to function as going-concern capital The 5,125% minimum level of CET1 ratio, constituted by the CRR for the contractual trigger, is a low value trigger. Since AT1 instruments should function as going-concern, one might consider 107

that a trigger set at 5,125% functions more as gone-concern capital.

This low level might especially be considered treacherous when one considers the BRRD, especially regarding the ‘PONV’ trigger specifically with regard to the discretionary regulatory power to bail-in AT1 instruments on a going-concern basis. With respect to the ‘PONV’ trigger, the directive reiterates the conditions set out in recital 81, that the trigger event occurs when a bank is failing or likely to fail. Concerning the determination, whether a bank is to be deemed failing/108

likely to fail, the BRRD specifies that this supposed when objective elements occur, ‘to support a

Angelos Delivorias, Contingent convertible securities Is a storm brewing?, European Parliament Briefing

106

May 2016. This is not consensus, Deutsche Bank CEO Cryan stated, that “AT1 is a lost instrument. It is incredibly expensive. […] never wanted to issue it”, implying that the bank is unlikely to issue more CoCo bonds going forward. FT reporters, Deutsche co-CEO says CoCos are “bad product”, The Financial Times,16 March 2016.

Pazarbasioglu, Zhou, Leslé, & Moore, Contingent Capital, p. 10.

107

Article 59 (4) (a) BRRD.

(25)

determination that the group, in the near future, will infringe its consolidated prudential requirements in a way that would justify action by the competent authority including but not limited to because the group has incurred or is likely to incur losses that will deplete all or a significant amount of its own funds.’ (Article 59 (6)). The article does not clarify what is deemed a significant amount or if a remaining high level of capital negates the impression that the bank is failing.

Bearing in mind the level of regulatory capital that banks have to fulfil once the CRD IV framework is fully implemented (up to 18 % of TREA for banks deemed systemically important or 13 % for ordinary banks), it is possibility that the regulatory authority deems a bank to be failing before any contractual trigger of an AT1 instrument is activated. 109

Looking back on the crisis of 2008, that gave rise to the concept of CoCos, reveals that most banks that were bailed-out had capital ratios well above 8% and consequently reached a PONV trigger before a 5,125% trigger event would have been in near sight. 110

One recent case that highlights the uncertainty about the deployment of a regulatory trigger before the bank reaches a 5,125% capital level is the example of the Deutsche Bank. Although the minimum trigger level provided by the CRR only sets the floor and not the ceiling, thus far all AT1 capital instruments issued by the Deutsche Bank provide only for a trigger event of 5,125%. The 111

crisis for the Deutsche Bank was sparked by the threat of a 14 billion dollar settlement between the U.S. Department of Justice and the bank, once described by the IMF as the world’s most systemically risky bank. 112

Even though the CET1 capital ratio remained at 10,8% and according to an analysis to trigger a contractual write down on the AT1 securities, a fine of 16 billion euro would be required, well above the settlement proposal, the price of the bank’s AT1 instruments showed the concern that 113

Cf. Cahn & Kenadjian, Contingent Convertible Securities, p. 56.

109

110

Cf. https://www.db.com/ir/en/capital-instruments.htm#tab_additional-tier-1 (last revisited at: 

111

30 December 2016).

Patrick Jenkins, Systemic fears over Deutsche Bank do not add up, The Financial Times, 30 September

112

2016. Jenkins refutes the notion of systemic risk by this bank.

Hale, Deutsche Bank’s riskiest bonds rattled by $14 bn claim, The Financial Times, 16 September 2016.

(26)

the securities might reach a trigger event, falling to a valuation of 69,97 cents on the euro. An 114

additional sign of uncertainty was that the German government deemed it necessary to publish a statement affirming that Germany was not preparing a rescue plan in case the bank failed to raise capital. 115

It can reasonably be argued that the cause for the uncertainty regarding Deutsche Bank’s AT1 bonds was less about the low trigger but rather the bail-in tool of the BRRD, and that a higher contractual trigger might not have had an impact in the value drop of these AT1 bonds. But the argument, remains that the current minimum trigger is set too low, given that most significant banks have a CET1 ratio well above 10%; a contractual trigger set at 5,125% becomes subordinate, since in a crisis the national regulatory authority will have stepped in. A higher mandatory contractual trigger would avoid the threat of regulatory action prior to an implementation of the mechanical trigger. Thus impeding the functioning of a contractual going-concern trigger.

One explanation why the European legislator opted for a low trigger requirement could be a concession to the banks. Low trigger AT1 bonds command a lower yield premium than high-trigger bonds, the yield to maturity spread for high-trigger issuances over subordinated debt is 3,6%, compared to a 2,5% spread for low-trigger issuances; considerably lowering the financing costs for maintaining the capital requirements, the reason for this is that the conversion risk is lower for low-trigger issuances. Another explanation might simply be that the priority of the legislator was not 116

to reduce systemic risks of banks ex ante, but to limit the necessity of a public bail-out by increasing bail-in options. 117

2. Coupon Cancellation and Ability to Call AT1 Instruments

Two elements of AT1 instruments, not included in early concepts of CoCos, are the cancellation of coupon payments and the extension risk (due to their perpetual character without maturity) imposed on issuer and instrument holder. These two elements can be directly placed within those characteristics of CoCos that correspond to the equity side of this instrument.

FT reporters, Deutsche Bank’s coco bonds slide to record low, The Financial Times, 30 September 2016.

114

Zeit online, Bundesregierung bereitet Notfallplan für Deutsche Bank vor, Die Zeit, 28 September 2016.

115

Avdjiev, Kartasheva & Bogdanova, CoCos: A Primer, p. 53. Defining low trigger as below 6% and high

116

trigger as above 6%.

Cf. European Securities and Market Authority, Statement: Potential Risks Associated with investing in

117

Contingent Convertible Bonds & AT1 Instruments : Does Banking Law Facilitate That Contingent Convertible Bonds Effectively Provide Sufficient Additional Capital In A Crisis To Banks On A ‘Going-Concern’ Basis?

C

C

B

& AT1 I

:

D

B

L

F

T

C

C

B

E

P

S

A

C

I

A C

T

B

O

A ‘G

-C

’ B

?

J

B

K

Table of Contents