10 Years on: Has “Too big to Fail” been resolved.

Steve Webb, Senior Advisor

A raft of financial regulation has been put in place in the 10 years since the height of the financial crisis. This regulation was once pithily summed up by a client  (the COO of a global-markets division) as follows.

There are now three rules for banks:

1. Don’t get sick (have much more capital and liquidity and de-risk)

2. If you get sick, have a plan to get better; don’t die  (Recovery Planning)

3. If you do die, die quietly and alone (Resolution Planning)

This article summarises where we have got to on complying with rule 3.


This autumn (or fall for any Americans reading) marked the 10th anniversary of the critical period of the great financial crisis. The failure of Lehman Brothers, the global investment bank, is seen by many as the key event of the crisis.  In fact, in my view the events that took place shortly afterwards in the UK in October of 2008, in particular the bail-out and rescue of RBS by the UK government, were even more far reaching.  The decision by Gordon Brown and Alistair Darling to rescue RBS (and Lloyds, although arguably Lloyds could have survived anyway) is still criticised in some quarters.  At the point that it needed bailing-out RBS had grown to be the fifth largest bank in the world and the largest in the UK.  In my view, had the government allowed RBS to fail then the effects of the crisis would have been many times worse.  Not only would the UK economy have been massively disrupted by the failure of the largest domestic retail and SME bank, but other UK and global banks were so exposed to RBS, that some of them would have been dragged down too.  It truly was “too big to fail”.

This blog focusses on how regulators have addressed the “too big to fail” problem. That is, the problem that allowing a major bank to fail (like a normal commercial business can) causes widespread disruption to the wider economy.

The wider regulatory environment

Much of the regulation implemented since the financial crisis has been driven by lessons learned from it.  The G20 created the FSB (Financial Stability Board) in 2009 to coordinate efforts globally to enhance the security and stability of financial markets and create a more resilient banking sector.  This has led to most regulators globally introducing unprecedented levels of new regulation in their own locations.  Much of that effort has focussed on enhancing the risk processes in banks reinforcing market infrastructure (central clearing) and making banks more financially resilient to shocks. All of this is aimed at making banks less likely to fail.  However, despite these measures it is recognised that a zero-failure regime is not something that is desirable, as commercial banks are there to take measured risks in order to serve the wider economy. 

To manage the potential for failure, regulators globally have developed tools to manage the resolution of a bank (no matter how large and complex) safely and without recourse to taxpayer funds.  “Safely” in this context means that a bank failure may be resolved without immediate disruption of key business services, widespread or extreme market impact or significant risk of contagion.

The specific regulations relating to bank failure are generally referred to as Recovery and Resolution Planning or RRP.  Recovery and resolution are linked, but they are quite distinct both in execution and in intent:

  • Recovery plans are prepared by and executed by the management of a firm to take significant action to financially recover from a crisis.  This is likely to involve the disposal of certain material businesses of the firm in order to generate capital and liquidity, the run down of other businesses and the continuation of what remains as a going concern
  • Resolution plans deal with what happens when a firm goes beyond the point of viability and goes into resolution.  This relies upon measures put into place by the management of the bank that will be executed post resolution by regulators or their appointed administrators.

It is the latter, more complex challenge of “dying quietly and alone” that the rest of this article will focus upon.

The evolution of RRP regulation

As the very largest and standards most complex banks are global in nature, the need for regulators to work together has been well recognised in this space, with the creation of multi-regulator crisis colleges for the biggest global banks the GSIBs (Globally Systematically Important Banks).

The FSB has also developed global guidance for effective resolution.  For anyone interested in an overview of the intent of this regulation, the FSB paper Key Attributes of Effective Resolution Regimes for Financial Institutions is quite accessible and worth a read.  This describes a number of pillars that should be in place in order to have an effective regime, including a resolution authority and powers for that authority to use, to help drive safe resolution.

These attributes have been enacted globally in similar but subtly different ways.

  • The US reflected this in the Dodd-Frank Act;
  • The EU has introduced the Bank Recovery and Resolution Directive (BRRD);
  • The UK has a number of policy and supervisory statements that have been introduced. The UK measures are well summarised in the so called “Purple Book” for those who might be interested;
  • Switzerland, where the government had to bail-out UBS, has introduced its own “too-big-to-fail” legislation;
  • Other regulators have adopted resolution planning rules to different degrees, though it is fair to say that Asian regulators are much less evolved in this space.

One of the most key areas of divergence between regulatory regimes lies in the approach to resolution strategy, or how the bank will be resolved:

  • In the EU and the UK, the ECB and the Bank of England have built their approach around asking for a great deal of data about how the critical functions of the bank are organised.  This is used by the resolution authorities to develop their own view of the resolution strategy. 
  • The US regulator places responsibility for development of resolution strategy on the banks.  Banks are asked to explain their strategy for a bank insolvency process and to evidence how that strategy would be operationalised.  This means there is a lot more focus on the outcomes of resolution and how the desired “safe resolution” will be achieved.  For the largest banks in the US the submission and feedback process has involved public disclosure of part of the plan (the public disclosure) and the US regulators response.  This has resulted in visibility of a number of iterations where firms are told what the regulator considers to be a barrier to their resolvability and the firm does something to address this.

It is fair to say that many observers, myself included, believe that the US approach has been more fruitful in terms of creating actionable plans for resolution.  It is little surprise therefore to find other regulators moving towards this approach, with the Bank of England announcing in the October 2017 Purple Book that it would introduce public disclosure of its assessment of firm’s resolvability measures starting in 2019.  The BoE states that it feels transparency will act as a further incentive for firms to remove remaining barriers to resolvability.

What does resolution of a bank involve ?

For the majority of banks, resolution would involve a two-step process of initial stabilisation through financial measures, followed by restructuring:

  • Step 1 Stabilisation would recapitalise the firm through bail-in of the group’s unsecured creditors, such as bond holders (many banks have issued specific bail-in debt) . The FSB has developed Total Loss Absorbing Capacity (TLAC) standards to ensure that firms have sufficient equity and medium term debt to finance a bail-in for most circumstances.
  • Step 2 Restructuring recognises that in most cases a bank that goes into resolution will not survive simply through financial measures. Recapitalisation is intended to buy time so that critical parts of the group with the greatest impact on the real economy, continue operating for a period and are resolved in a controlled process over time through a combination of sales, asset transfers, run-down of business and potentially bank administration. 

The firm’s structure also impacts how the resolution is executed (generally referred to as resolution strategy):

  • Single Point of Entry (SPE): approaches the resolution top down, attempting to maintain the overall group structure at the point of stabilisation and executing restructuring in a coordinated resolution of multiple group entities.  For most large, globally interconnected banks this is the preferred strategy.
  • Multiple Point of Entry (MPE): an MPE strategy would see multiple individual legal entities within a group enter resolution independently, applying bail-in and restructuring entity by entity. For most complex global banks, the degree of inter-connectivity and inter-dependence that exists between group entities makes this option very unattractive. 

Executing bank resolution

To state the obvious, regardless of what arrangements are put in place, the resolution of a large complex global bank is a significant undertaking that will take a lot of time and effort and carries significant risk. Some of the key areas to consider are as follows.

Executing Bail-In

The execution of bail-in:

  • Must recognise a defined creditor hierarchy, which can apply to all liabilities (unless explicitly excluded like insured deposits are in BRRD)
  • Requires rapid valuation of assets and liabilities to determine how much bail-in capital is required and the sources of this capital
  • must pass a “No creditor worse off”(NCWO) test to ensure that losses imposed on creditors will not be worse than they would have been had the bank entered a normal insolvency process
  • will require liaison with and between regulators globally who will want to understand how actions taken at a group level impact on their own local creditors and that these are not disadvantaged over those in other regions

It should be clear therefore that execution of bail-in will be challenging. The data and valuation models required are extensive and complex; regulators may be driven by local political concerns in a crisis and the NCWO test is complex (and has been made more challenging by the experience of Lehman in the UK where LBIE creditors received all of their money back plus statutory interest).

These complexities have been recognised by most firms, who have put in place tools and procedures to rapidly carry out the required valuations and have structured internal debt in such a way as to reassure regulators that bail-in capital would “flow through”.  Regulators have demanded action for example BRRD requires that firms can provide valuations data at short notice and can test this in dry run exercises. 

Operational Continuity

Another key pillar for stabilising a bank, post resolution, is the ability to continue to run all of the critical operations.  This area has been a focus for specific regulatory developments under the Operational Continuity in Resolution (OCIR) heading.  The FSB produced guidance on the topic and the Bank of England has probably been the most forward thinking in terms of introducing OCIR rules (SS 9/16 , which must be implemented by January 2019).

OCIR is one of the most far reaching aspects of RRP for banks and in summary requires:

  • Mapping of all critical operational services to the critical economic functions (i.e. the business services that the bank provides to customers) that they support.  This should allow regulators to understand what needs to be kept running and what the operational implications of restructuring the business are.  This has led most firms to develop comprehensive service catalogues that list critical operations and map them to critical economic (business) functions.
  • Formal management of service arrangements with all third parties and intra group service providers such that service levels are clearly defined and service metrics understood. This is so that SLAs can continue to be enforced post resolution and can be converted to new Transitionary Service Arrangements (TSAs) to support structural changes if required.
  • Securing contractual arrangements such that suppliers of critical services cannot terminate due to resolution (assuming they continue to be paid). This has required banks to include “resolution resilient” language into contracts and agree this with suppliers.
  • Understanding charges for services and introducing “arm’s length” charging.  Firms need granular transparent pricing data to understand operating costs in resolution and to provide financial data as input to restructuring.  The Bank of England has also introduced concepts around “arm’s length” commercial arrangements.
  • Being able to continue to fund operational services.  Firms need to set aside liquidity that will be used to continue to pay for operational services after a resolution event (6 months cost of critical services for BoE rules).
  • Maintain access to operational assets that critical services require. This includes staff, premises, applications, licences (for software), IP. Securing continued access to such assets requires the data about the assets to be linked in some manner to the critical service catalogues and plans to retain suppliers, staff, IP etc.
  • Governance arrangements and prevention of preferential treatment.  Firms are asked to consider how governance arrangements will avoid conflicts of interest post resolution. In complex groups with multiple entities and shared service models, this has led to examination of dual-hat arrangements, secondments and matrix reporting lines with pre-emptive changes to arrangements and specific post resolution actions being identified.

OCIR is one of the most challenging areas for firms as it impacts heavily on the BAU operating model. In fact, getting OCIR preparation right has a number of potential efficiency benefits for firms (but that will be the topic of a future blog).


It is difficult to predict in advance what restructuring actions a regulator/administrator would take following a resolution and stabilisation event.  The actions taken will depend upon the specifics of the firm, the nature of the crisis that led up to resolution and the situation in the market (i.e. is the market operating normally or is the crisis systemic).

The following actions are likely to be considered:

  • Maintain and preserve the core economic functions of the bank (likely to be considered for Retail/SME banking, Mortgages, Payment Services) through:

                 o Continuing as standalone concern

                 o Sale to another bank

                 o Bridge bank (followed by sale or asset transfer)

  • Sell off discrete elements of the bank, that are sufficiently self-contained businesses and are a going concern financially (likely to apply to stand-alone businesses e.g. Asset Management). Sale will require consideration of:

                o State of wider market 

                o Ability to execute quickly

                 o Interdependencies with the rest of the group

                 o Legal entity structure

  • Wind-down (solvent) of business is likely to apply to trading business and large corporate lending portfolios (note that both BoE and the ECB have asked selected banks to model the solvent wind-down of their trading portfolios). Key considerations include:

                o Tenor of trading book (the natural maturity profile of the portfolio)

                o Complexity of products and ability to sell part of portfolio

                o Transfer/return of client assets

                o Ability to novate trading positions

  • Bank administration.  This is the fall-back option and is unlikely to be used for major banks in all but the most extreme circumstances.  Insolvency would follow defined rules for the best interest of all creditors and involves:

               o Freezing normal commercial operations

               o Return of client assets

               o Collection of liabilities 

               o Establish creditor positions 

               o Manage creditor claims

As noted earlier it is impossible to have a defined plan for restructuring in advance of a resolution occurring (aside from UK structural changes around ring-fencing that have pre-positioned the retail SME functions of banks ready for transfer/maintenance).  The key concern of regulators therefore, is to have “optionality” around the actions that they can take to split up the bank in different ways and use different options for different parts of the business.  Key to achieving this, is having sufficiently granular data to understand the impact of change.

Why is this difficult?

To those who have not spent much time working inside the global banks it may appear that none of the above should prove too challenging.  So, what are the key challenges that banks face in complying with the rules?

  • The global model.  Banks have spent a lot of time organising themselves globally to be as efficient as possible at sharing costs, pooling liquidity and managing assets across business lines, borders and legal entities.  They are organised in this way to optimise costs in a business as usual (BAU) going concern scenario.  For this reason, systems have not been set up to provide data broken down country by country, entity by entity and business by business as may result from the restructuring required in resolution.
  • Inter-connectedness.  Most complex organisations have multiple dependencies between business, legal entities and regions.  It is one thing having the data to explain these dependencies, but another to unpick them quickly if post resolution restructuring required it.  This has led some (mostly US firms) to simplify their legal entity structures.
  • Data and MI.  The vast amount of data required to support valuations, operational cost analysis and wind-down analysis represents a serious challenge to all firms.  Most have begun to address it but many have far to go.
  • Integrated operations.  Many firms have their operational functions embedded in the business and see it as a cost centre without any formal approach to service levels of service metrics.  This stands in the way of OCIR requirements as it will not facilitate the operational change management required of restructuring.

In conclusion

In conclusion is “Too Big to Fail” really a thing of the past? 

Mark Carney doesn’t think so, or at least didn’t as of a year ago (September 2017) when he  said that he thought some banks were still not in a position where they could be allowed to fail. 

This is not entirely surprising as in many cases the legislation is still evolving, for example, in Europe, regulators are playing catch-up to the UK, US and Switzerland and even in these “trail blazer” countries new legislation is being introduced (for example in the UK OCIR rules are only in force from next year and the introduction of resolvability assessments foresees further changes).

At the same time, I have heard first hand from many of my old PwC colleagues the lessons that they learned from the administration of Lehman, and in particular how the total lack of critical data hampered the early weeks and months of their efforts.  In my view RRP legislation has driven us to a much better situation than existed back then, and in particular:

  • TLAC and bail-in rules would allow most banks in most circumstances to be refinanced to buy crucial time.
  •  Several iterations of public disclosures and regulator response in the US has driven banks to take ex-ante actions that reduce the complexity of resolution  (e.g. legal entity simplification)
  • OCIR rules have forced firms to look at operating models and enhance documentation, procedures and systems such that critical operations would be much easier to secure in the event of a crisis
  • Solvent Wind Down modelling has given banks and regulators insight into what it would take to exit a complex trading business and the cost and time this would involve.  This could inform options in any real crisis and help avoid the value destruction associated with ISDA default

In conclusion, I feel that trying to answer this as a binary question is unhelpful. 

It may be better to see resolving Too Big to Fail as a journey.  The very end of that path would be that even in a global financial crisis that paralyses the market and places all firms at risk, regulators were confident to be able to resolve more than one major bank. This would be a resolution regime that was ready for the most extreme “black swan” events, and I don’t believe that we are there yet (whether we ever want to get there is another interesting question – how failure-proof do we want to be?).

Today, though we have better tools, the challenges are more clearly understood and most banks have better data that could be used to manage in a crisis. We are not at the end of the road, but we have traveled a long way  from where we started in 2008.