1
## Making banks resolvable: the key to making resolution work
Speech given by
Andrew Gracie, Executive Director, Resolution Directorate
Risk Minds Conference, Amsterdam
4 December 2017
## Introduction
Resolution has come a long way since G20 Leaders put together the post-crisis financial reform agenda in summits in London and Pittsburgh in 2009. 1 In some ways, it represented the most notable gap, and significant change in the pre-crisis regulatory architecture. Nearly ten years on, huge progress has been made in establishing effective resolution arrangements and the commitment to ending too-big-to-fail (TBTF) is undimmed.
The immediate priority in this effort was to put in place the necessary legal frameworks. Agreement in 2011 of Financial Stability Board (FSB) Key Attributes for Effective Resolution Regimes provided the international standards to ensure a consistent approach to the design of resolution regimes across G20 jurisdictions. 2
The UK now has in place a comprehensive bank resolution regime that is compliant with international standards and will remain so after Brexit. Similarly, for all advanced economies, there are now resolution regimes that are largely compliant with the Key Attributes in all the jurisdictions that are home to global systemically important banks (G-SIBs).
But the Key Attributes were about effective resolution regimes rather than resolvable firms - they defined a tool-kit but not how to use it; and a process for resolution planning for G-SIBs but not what would make a firm resolvable. Powers without resolvability leaves resolution authorities vulnerable.
Indeed, the moral for me of recent failures and near-failures is not that resolution is misguided and will not work, but that it will not work if firms are not regulated and supervised in a way that makes them resolvable.
This is the story of the last five years. We have focussed on organising firms in such a way that authorities' resolution powers can be used without significant adverse consequences for the rest of the financial system or the wider economy. This moves us progressively to where we want to be against risk appetite. And though we are not yet where we finally need to be, what has been done has already yielded significant benefits. For example rating agencies have largely removed government support uplifts to bank' credit ratings. 3 Where UK firms have come under stress, our resolution arrangements have been one factor that has helped secure recovery.
Today I want to review where we are on the journey: what has been done and what is left to do. I want to focus in particular on three topics:
1 See FSB 'Declaration on Strengthening the Financial System - London Summit' (2009) http://www.fsb.org/wpcontent/uploads/london\_summit\_declaration\_on\_str\_financial\_system.pdf and FSB ' Leaders' Statement the Pittsburgh Summit' (2009) http://www.fsb.org/wp-content/uploads/g20\_leaders\_declaration\_pittsburgh\_2009.pdf
2 See FSB 'Key Attributes of Effective Resolution Regimes for Financial Institutions' (2014) http://www.fsb.org/wpcontent/uploads/r\_141015.pdf.
3 See S&P Global 'An Illustrative Rating Path for a Systemic Bank In A Bail-In Resolution' (2017) http://images.ratingsinfo.standardandpoors.com/Web/StandardPoorsRatings/Illustrative%20Rating%20Path.pdf.
2
- (i) internal total loss-absorbing capacity (TLAC) or minimum requirements for eligible liabilities and own funds (MREL) and the underpinning it provides for cross-border co-operation;
- (ii) bail-in mechanics - having required banks to maintain TLAC or MREL, we need, as resolution authorities, to be clear how we would use it; and
- (iii) disclosure - resolution needs to be credible as well as feasible. With credibility comes market discipline ex ante and less disruption in a resolution (lower probability of default and loss given default in other words).
All three areas build on the core work of resolvability that has already been done, illustrating the point that the process is incremental and resolvability is not binary but progressive.
## Resolvability of firms
Let me start then by reviewing what has already been done.
I will focus on the standards and guidance coming from the FSB. It is true that they have been drawn up for application to G-SIBs in Crisis Management Groups (CMGs). But in our view, they are equally applicable to domestic systemically important banks (D-SIBs) or other firms where bail-in is the preferred resolution strategy, especially where firms operate cross-border.
The convergence at international level on bail-in as the appropriate strategy for large banks is itself significant. This transcends whether firms are single point of entry or multiple point of entry. And whether the bail-in is effected by use of a bail-in tool in an operating bank or by application of a bridge bank tool at holding company level. It represents the realisations that if we are to end TBTF we must have bail-ins instead of bail-outs so losses can fall on investors not on taxpayers. And that the bail-in must enable continuity in a firm's operations to avoid interruption to critical functions and to buy time for an orderly reorganisation of the firm to the extent necessary to deal with the problem that first caused the failure.
This shared understanding at FSB, and the desire to meet the goal set by G20 Leaders to end TBTF and the risk of taxpayer bail-outs as soon as possible, has driven a focus on seeking changes at firms so that they can be stabilised and enter resolution safely without disruption to critical functions. Hence the focus on two dimensions:
- -Loss absorbency: firms need TLAC in the right amount (enough not only to absorb losses but to provide for recapitalisation so firms can continue to meet requirements for authorisation); in the right form (debt with residual maturity of a year and subordinated to operating liabilities to avoid breaches of the 'no creditor worse off than insolvency' (NCWO) safeguard or other challenges to the bail-in and at the same time to provide clarity to depositors that they are not likely to be bailed in); and in the right location (ensuring that resources are positioned within a group so that the key operating companies containing a firm's critical functions can be recapitalised
3
immediately in resolution).As such, international agreement of the FSB's TLAC standard in 2015 represented a major milestone in moving towards ending TBTF. 4
- -Continuity of critical functions: the second main dimension of stabilisation is to ensure continuity in a firm's operations in resolution. Part of this is legal - changing contractual arrangements so that entry into resolution does not result in widespread, disorderly termination, close-out or acceleration in financial contracts, provision of services or access to financial market infrastructures (FMIs). It is true that under the Key Attributes, resolution regimes include statutory powers to stay. But statutory stays may not be effective cross-border. And, anyway, repapering contracts helps to convey to counterparties that there will be continuity in resolution and so reduce incentives to break for the exit. FSB has published guidance in this space for operational continuity in resolution and continuity of access to FMIs. 5
Perhaps the most notable effort has been FSB's work with industry to agree a protocol to International Swaps and Derivatives Association (ISDA) and other master netting agreements that addresses close-out risk in over-the-counter (OTC), derivative and repo transactions. 6 Agreement in 2014 of the universal protocol amongst G-SIBs, and subsequently of jurisdictional modular protocols for individual jurisdictions to bring in buy-side and non-G-SIB counterparties, is another major landmark in ensuring big international firms are resolvable. 7 The protocol is built around the premise that entry into resolution should not be classed as an event of default as long as a firm continues to perform.
This underscores the importance of funding to making resolution credible. FSB set out principles on funding in resolution in 2016 and just last week published for consultation guidance for use in CMGs on the liquid resources and liquidity management capabilities firms require to be adequately resolvable and how liquidity in firms might be bolstered in adverse cases by public backstops. 8 CMGs will use the guidance to draw up resolution funding plans setting out how, in order to achieve resolvability, liquidity and collateral should be held in a group across legal entities, currencies and locations.
We have made good progress in regulating these various FSB standards in the UK, either through Bank of England Policy statements or Prudential Regulation Authority (PRA) rules. UK banks have been given indicative MREL requirement to meet by 2020; they will have operational continuity arrangements in place in 2019 (alongside the implementation of ring-fencing); and they are already trading on protocol terms with
4 See FSB 'Principles on Loss-Absorbing and Recapitalisation Capacity of G-SIBs in Resolution' (2015) http://www.fsb.org/wpcontent/uploads/TLAC-Principles-and-Term-Sheet-for-publication-final.pdf.
5 See FSB 'Guidance on Arrangements to Support Operational Continuity in Resolution' (2015) http://www.fsb.org/wpcontent/uploads/Guidance-on-Arrangements-to-Support-Operational-Continuity-in-Resolution.pdf and FSB 'Guidance on Continuity of Access to Financial Market Infrastructures (FMIs) for a Firm in Resolution' (2016) http://www.fsb.org/wp-content/uploads/Continuity-ofAccess-to-FMIs-Consultation-Document-FINAL.pdf.
6 See FSB 'Cross-Border Recognition of Resolution Action' (2014) http://www.fsb.org/wp-content/uploads/c\_140929.pdf.
7 See ISDA '2014 Resolution Stay Protocol' (2014) http://assets.isda.org/media/f253b540-25/958e4aed-pdf/, ISDA ' 2015 Universal Resolution Stay Protocol' (2015) http://assets.isda.org/media/ac6b533f-3/5a7c32f8-pdf/ and ISDA ' Resolution Stay Jurisdictional Modular Protocol UK (PRA RULE) Jurisdictional Module' (2016) http://assets.isda.org/media/f253b540-94/cd991d70-pdf/.
8 See FSB 'Guiding Principles on the Temporary Funding Needed to Support the Orderly Resolution of a Global Systemically Important Bank ('G-SIB')' (2016) http://www.fsb.org/wp-content/uploads/Guiding-principles-on-the-temporary-funding-needed-to-support-theorderly-resolution-of-a-global-systemically-important-bank-'G-SIB'.pdf and FSB ' Funding Strategy Elements of an Implementable Resolution Plan' (2017) http://www.fsb.org/wp-content/uploads/301117-2.pdf.
4
buy-side firms as well as other G-SIBs. To the point that resolvability is progressive, the major UK banks on average now have total loss absorbency of 23% measured against risk-weighted assets (RWAs) compared to an average end-state requirement of 28% (including buffers). 9
As described, there is more to do in some of these areas. But increasingly the emphasis is on implementation and, with it, assurance - how we supervise firms against these regulatory requirements and hold them to account that identified barriers to resolvability are removed and stay removed. In describing the progress towards end-state resolvability, I want to pick out three areas that will be a focus in the period ahead.
## (a) Internal MREL
The first is internal MREL. One lesson that the crisis brought home is that the distribution of resources within groups matters. While in life, firms might want to run themselves by business lines on a consolidated basis, failure and resolution occurs at legal entity level.
It is understandable then that since the crisis that there has been a discernible tendency for host authorities to hold onto more resources, capital and liquidity, or to force activity in branches into subsidiaries. The aim at FSB level with TLAC was to lean against this tendency towards fragmentation of international groups by providing comfort to authorities cross-border, not only that groups would have sufficient external TLAC to be resolvable but also that resources would be prepositioned as internal TLAC in material subsidiaries in host jurisdictions.
Previous international regulation of financial resources in groups has, like the Basel Accord, typically focussed more on the consolidated level. Now applying TLAC regulation at legal entity level has brought with it challenges.
One is timing: it is hard to roll out internal TLAC for a group until TLAC requirements are in place in all the jurisdictions relevant for the material subsidiaries. The Bank is in the middle of consulting on its internal MREL policy 10 but many other jurisdictions - notably the Banking Union, Switzerland and Japan - have still not set policy.
Another challenge is arithmetic: to the extent that there are financial dependencies between legal entities within a group the sum of solo RWAs is likely to be greater than consolidated RWAs. If a full TLAC requirement were applied to each subsidiary then the sum of the internal TLAC would exceed the consolidated TLAC requirement for the group as a whole.
9 The Bank of England published loss-absorbing capacity requirements for major UK banks in 2016 shortly after responding to the Treasury Committee's inquiry into capital. See
http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/treasury-committee/capital-andresolution/written/69208.pdf.
10 The Bank of England published a consultation paper on its approach to setting a minimum requirement for own funds and eligible liabilities (MREL) within groups on 2 October 2017. See https://www.bankofengland.co.uk/-/media/boe/files/financialstability/resolution/internal-mrel-consultation-october-2017.pdf?la=en&hash=33594C3FB3C7F1D129033AFE4E3A2BF20A4F9AA8 .
5
This was addressed in the TLAC term sheet in two ways:
- -by limiting internal TLAC only to subgroups and subsidiaries that are material to the group, accounting for more than 5% of RWAs or income. The logic there was that at that level failure of the material entity might trigger resolution of the group as a whole; problems for smaller entities, even those that might house functions that are critical to the host jurisdiction, should be more of a recovery matter.
- -by including a provision that internal MREL for subgroups should be scaled in single point of entry (SPE) groups to 75-90% of the requirement that would apply to the subgroup on a standalone basis. The logic was that given the interdependence in the business models of SPE firms, prepositioning in this range would be sufficient to secure cooperation and deter hosts from ring-fencing, and home authorities from cutting off, foreign subsidiaries. Prepositioning of resources in this way will provide concrete underpinning for cooperation that goes beyond the paper commitments to cooperate in Memorandum of Understandings (MOUs), honoured in the breach in previous cross-border failures.
Even with these restrictions there may still be a sum of the parts problem in which the sum of TLAC requirements at solo level pushes up external requirements. If internal TLAC requirements are high they are also likely to stand in the way of holding a surplus at the top of the group that could provide flexibility in covering losses if they are concentrated in a particular part of a group.
Thus in our internal MREL policy our intention is first to consider setting internal MREL generally for the firm at the bottom of the 75-90% range. 11 But to set a higher requirement if we have doubts about: the home resolution strategy, the availability of a surplus at the top of the group or a lack of reciprocity in MREL-setting in other jurisdictions (that is others all set at the 90% end of the range). We expect to say more about surplus MREL and the form it should be held in to be readily available when we come to finalise our internal MREL policy next year.
Another important objective for us in setting policy on internal MREL is to avoid any distortion in the sequence of loss absorption between operating companies (OpCo) and the group holding company (HoldCo). We require UK banks to issue external MREL out of their HoldCos, structurally subordinated to liabilities at the OpCo. The issuance proceeds are then onlent to the OpCo as internal MREL broadly mirroring their external form, for example: Tier 2 could be onlent as Tier 2, AT1 as AT1 and senior debt out of HoldCo would be downstreamed as an internal debt instrument sitting senior to capital instruments but junior to operating liabilities in the OpCo.
We need to be sure that as losses are incurred in the operating company they are absorbed in an order that follows the OpCo creditor hierarchy and does not bypass certain instruments. The aim with a single point of
11 Our starting point for calibrating internal MREL for ring-fenced bank sub-groups would be 90%, with the possibility that this could be scaled down if there are sufficient readily deployable resources.
6
entry bail-in strategy, as the name suggests, is only to put one entity at the top of the group into resolution, preferably a holding company which is only used for issuing capital instruments to the market and otherwise has no or limited operating liabilities.
Write down or conversion of internal MREL instruments via contractual triggers will then allow operating companies lower down the group to be recapitalised automatically without putting them into resolution. But legacy non-equity capital instruments in OpCos without the requisite contractual triggers could cut across this. We do not want these to be spared bearing losses for want of a trigger. And so, we are clear that such instruments should not count as MREL beyond 31 December 2021 and if not matured by then we will work with firms to remove them if they represent a barrier to the resolution strategy.
This is part of a larger effort to add to the credibility of resolution by being clear to debt investors and other stakeholders how a bail-in will work in practice. We want debt investors to be able to price risk effectively by being clear about where they rank in the creditor hierarchy and to give them confidence that we can stick to the hierarchy in applying losses and that we will not be picking winners and losers within a class. Counting as MREL senior debt issued out of operating companies without subordination looks like a recipe for undermining that confidence.
## (b) Bail-in execution
We also want to be clear as a resolution authority how we will conduct the resolution in a way that preserves value and distributes it fairly, ensuring no creditor worse off protections are met. To that end, we have thought hard about the valuation capabilities we need firms to have so that we can value losses and recapitalisation needs in an effective and timely way. We have just consulted on a set of principle level requirements and expect to finalise policy in this space in the next few months. 12
We believe that the bail-in valuation for a large cross-border firm is not something that can be fixed over a weekend. Valuation on that timetable will inevitably tend to overshoot on the estimation of losses, not least because a significant driver of the valuation will be the reorganisation that follows the bail-in and the restructuring costs and disposal valuations associated with that.
Rather, it is our intention to take several months for the bail-in valuation to ensure that through an independent valuation process, losses as far as possible are bottomed out and tie back to the reorganisation plan for the surviving business. It will be critical to determine the net asset value that had been generated in the bail-in and the liability holders it should go to as compensation.
You may ask, what happens in the intervening period? From the perspective of the debt investor, our intention on entry to resolution is to take control of the shares in issue and immobilise all the other external
12 The Bank of England published a consultation paper on valuation capabilities to support resolvability on 17 August 2017. See https://www.bankofengland.co.uk/-/media/boe/files/financial-stability/resolution/boes-proposed-policy-on-valuation-capabilities-tosupport-resolvability.pdf?la=en&hash=4044F91DF1DDE7A131EA3186F66F304380553306 .
7
MREL instruments at the relevant international central securities depository (ICSDs) and central securities depository (CSDs). We will at the same time issue onto the bondholders accounts at the ICSDs a tradeable certificate of entitlement secured by the shares in issue. There will be as many classes of certificate as there are classes of claim. Once the valuation is complete and we have announced the terms of exchange, holders will present the certificates in exchange for whatever equity they are entitled to as compensation. The resulting shareholders will then vote in a new board and the firm will return to private sector control.
Issuing certificates of entitlement in this way will provide a mechanism for debtholders that do not want to, or due to their mandates, are not able to hold equity to trade out of their positions. And this will have an ancillary benefit for us of providing some sort of shadow market valuation of the firm.
But what about other liability holders - depositors, market counterparties, trade creditors etc. - during this interval?
Our aim with the announcement of the resolution is to send a strong signal to them that their claims are safe and that the operating companies they are transacting with will continue to perform.
Full conversion of internal MREL in the key operating companies around the group and a strong sense that the firm will have access to liquidity will contribute to that. To this end, the Bank has established new arrangements to clarify, first, that a firm in resolution would continue to have access to the ordinary central bank facilities in the Sterling Monetary Framework, subject to meeting the necessary eligibility criteria. And second, those arrangements will be supplemented where necessary by a flexible Resolution Liquidity Framework, designed to provide liquidity, in sterling or foreign currency, in the necessary scale, for a sufficient period of time, and secured against a wide range of collateral to allow the firm to make the transition to market-based funding. 13 Sending a strong message in this way at the outset of the resolution will maximise the chances of stabilising the firm and if liability holders know their claims will be refinanced, reduce the risk of a further run.
This need for clarity is understood at international level. FSB has just published a consultation paper on bail-in execution addressing a number of issues I have described including valuation and exchange mechanics. 14 Not least there are important home-host dependencies that need to be addressed. After all, in a UK G-SIB resolution we will need to bail-in debt on both sides of the Atlantic, in the Depository Trust and Clearing Corporation (DTCC) as well as Clearstream and Euroclear. This includes working through with market regulators the securities law issues that will arise through a bail-in, ensuring that disclosure and listing requirements continue to be satisfied.
13 See Box 2 - the Bank's approach to providing liquidity in resolution in https://www.bankofengland.co.uk/-/media/boe/files/financialstability/resolution/boes-approach-to-resolution.pdf?la=en&hash=8213BE00D67C4CADB948D51FEBD164E136A70BE6 14 See FSB ' Principles on Bail-In Execution' (2017) http://www.fsb.org/wp-content/uploads/P301117-1.pdf
8
## (c) Disclosure
The third and final area I want to address is disclosure. A recurrent theme of my remarks has been the importance of ex ante disclosure to making resolution credible and to realising the benefits of increased market discipline. We want debt investors to have the information they need to price risk and so support the proposals in the TLAC term sheet and Basel's Pillar 3 Framework that require banks to disclose the rank ordering of their liability structure at legal entity level.
As I have described, we want debtholders and other stakeholders to understand how we will use out resolution powers and so have published a document setting out our approach to resolution - otherwise known as the Purple Book. 15 The latest edition in October indicates how far we have come since we first published the Purple Book in 2014. But we have further to go.
It is our intention - which we indicated to the Treasury Select Committee in March - to publish summaries of the resolution plans for the major UK banks and our assessment of their resolvability. We will do this after the start of 2019 when ring-fencing, the first level of TLAC requirements and operational continuity in resolution arrangements are in place.
As an institution we want to be open and accountable. As a resolution authority we need to be for it to work.
There is a famous Chinese proverb that notes a journey of a thousand miles begins with a single step. On the resolution journey there have been a lot of steps - in fact we could all be forgiven for feeling footsore but the thousand mile marker is coming into view.
15 See the Bank of England's approach to resolution https://www.bankofengland.co.uk/-/media/boe/files/financialstability/resolution/boes-approach-to-resolution.pdf?la=en&hash=8213BE00D67C4CADB948D51FEBD164E136A70BE6 .
9