Loss absorbing and recapitalisation capacity

Recommendation 3

Implement a framework for minimum loss absorbing and recapitalisation capacity in line with emerging international practice, sufficient to facilitate the orderly resolution of Australian authorised deposit-taking institutions and minimise taxpayer support.

Description

APRA should develop a loss absorbing and recapitalisation framework aligned with international standards: it should not generally seek to move outside international frameworks or ahead of global peers unless there are specific domestic circumstances to warrant this.

This framework should provide sufficient loss absorbing and recapitalisation capacity to facilitate the orderly resolution of ADIs. It should minimise negative effects on financial stability, ensure the continuity of critical functions and minimise the use of taxpayer funds.

Total loss absorbing and recapitalisation capacity should consist of an ADI’s equity as well as debt instruments on which losses can credibly be imposed in a resolution. This includes debt instruments that can be converted to equity or written off where specified triggers are met to recapitalise the ADI or its critical functions.

The Inquiry supports pursing such a framework, but cautions Australia to tread carefully in its development and implementation as this area is complex and evolving. The Inquiry recommends that the framework follow these guiding principles:

  • Clearly set out the instruments eligible for inclusion in a loss absorbing and recapitalisation capacity requirement.
  • Ensure clarity of the creditor hierarchy with clear layers of subordination between classes.
  • Ensure clarity of the mechanisms and triggers under which creditors will absorb losses.
  • Seek to ensure eligible instruments can be exposed to loss without adverse consequences for financial stability, including being held by investors who can credibly be exposed to loss.

The Inquiry intends that this framework would only include specific liabilities and not deposits. Deposits are protected by a guarantee under the FCS of up to $250,000 per account holder per ADI and by depositor preference. In Australia, deposits are not and should not be subject to bail-in.

In considering eligible instruments, the benefit of the lower cost of less subordinated instruments, such as a new layer between Tier 2 and senior unsecured debt in the creditor hierarchy, should be weighed against the ability to credibly write off or convert the instrument without causing financial instability. The clearer the mechanisms and triggers under which creditors will absorb losses are in advance, the more likely it is that this can be achieved. To this end, where losses are to be imposed through instruments being converted to equity or written off, issuing new contractual instruments has substantive advantages over broad statutory bail-in powers.

Objectives

  • Ensure Australian ADIs have sufficient loss absorbing and recapitalisation capacity in resolution to make it feasible to implement an orderly resolution.
  • Reduce perceptions that some banks are subject to an implicit Government guarantee to lessen market distortions created by this perception and improve competition in the banking sector.

Discussion

Problem the recommendation seeks to address

In a stable system, if financial institutions fail, they do so in an orderly fashion, without excessively disrupting the financial system, without interrupting the critical economic functions these institutions provide or exposing taxpayers to loss.52

The Inquiry believes three aspects of Australia’s framework for the orderly resolution of ADIs could be strengthened:

  1. Effective crisis management powers for authorities. Recommendation 5: Crisis management toolkit addresses this aspect.
  2. Effective pre-positioning and planning for the use of those powers. The Inquiry supports further work by authorities on this aspect.
  3. Sufficient loss absorbing and recapitalisation capacity, which is addressed by this recommendation.

Currently, Australia does not have requirements for loss absorbing and recapitalisation capacity. Introducing a loss absorbing and recapitalisation capacity framework creates credible alternatives to using taxpayer funds to resolve a bank and reduces perceptions of an implicit guarantee.

This is a focus of ongoing international policy work building on the experience of many national governments during the GFC, where significant taxpayer funds were put at risk to assist troubled banks as no other credible options were available to support financial stability. In many cases, even capital instrument investors were bailed out, despite these instruments being intended to absorb losses. As a core part of the G20 agenda to end the problems associated with some institutions being perceived as ‘too-big-to-fail’, the Financial Stability Board (FSB) is consulting on an international framework for loss absorbing and recapitalisation capacity for global systemically important banks (G-SIBs).53 Indications are that many countries will also adopt these standards for D-SIBs. As a small, open, capital-importing economy, Australia cannot stand outside international practice.

An orderly resolution can be achieved with Government support, but this puts taxpayer funds at risk and protects bank creditors from loss. If Australia introduces a framework requiring banks to have sufficient loss absorbing and recapitalisation capacity, losses or recapitalisation costs are more likely to be borne by a failed bank’s shareholders and creditors rather than taxpayers.

Further, if the market believes that Government support is the only viable option, this creates the perception of an implicit guarantee and the potential for associated distortions. The Australian Government support provided during the GFC, although not at the same level as in some other jurisdictions, has reinforced perceptions of an implicit guarantee for some banks in Australia.

Perceptions of implicit guarantees have costs, creating a contingent liability for the Government and distortions in the market. They reduce market discipline and potentially confer funding advantages on the banks involved. Credit rating agencies explicitly factor in rating upgrades for banks they perceive to benefit from Government support, directly benefiting these banks.54 Reducing perceptions of implicit guarantees in Australia could therefore improve efficiency and competition in the banking sector.

Rationale

Australia’s prudential framework is not, and should not be, premised on the assumption that ADIs will never fail, nor that unsecured bank creditors will never be exposed to loss. Inevitably, failures can and will occur, the system will be exposed to crises and, at times, unsecured bank creditors will be exposed to loss.

A loss absorbing and recapitalisation capacity framework would help to implement an orderly resolution of a distressed ADI with minimum use of taxpayer funds. This would reduce perceptions of an implicit Government guarantee, thereby reducing the contingent liability of the Government and the associated market inefficiencies.

Options considered

The Inquiry considered two options:

  1. Recommended: Implement a framework for minimum loss absorbing and recapitalisation capacity in line with emerging international practice, sufficient to facilitate the orderly resolution of Australian ADIs and minimise taxpayer support.
  2. Make no change to current arrangements.

Option costs and benefits

The banking sector disputes the need for additional loss absorbing and recapitalisation capacity. However, banks generally acknowledge that such a framework is inevitable given the work underway to develop a set of international standards.

In this context, most of the major banks argue strongly that senior unsecured debt should not be subject to bail-in. They contend that, were such a bail-in ever used, it could have a significant destabilising effect on the financial system. To this point, they note that senior unsecured debt is a vital funding source and that a loss of investor confidence in that market could be damaging. Instead, banks prefer a loss absorbing and recapitalisation capacity requirement in the form of existing Tier 1 or Tier 2 capital, or a new layer of loss absorbing debt distinct from regular senior unsecured debt.

The banks also warn that a loss absorbing and recapitalisation framework would introduce costs, as bail-in debt would have higher spreads than existing debt, reflecting the additional risk. This could be exacerbated if the demand for these bail-in instruments is limited and spreads increased further to encourage greater holdings. Banks submit that changes in funding costs would be passed on to consumers, at least in part, which would raise the cost of credit and potentially affect GDP growth.

APRA notes that the global debate is moving beyond how to reduce the probability of bank failure, which is addressed by capital requirements, and now focusing on how to reduce the cost of failure. This will result in a global loss absorbing and recapitalisation capacity framework for G-SIBs to remove perceptions that such institutions are too-big-to-fail. Although Australia has no G-SIBs, when seeking funding in wholesale markets, the internationally active Australian banks must compete against banks that meet these global requirements. These competitors include other internationally active banks from jurisdictions that adopt these standards more broadly.

The RBA acknowledges that the risks associated with a bail-in of creditors need to be carefully considered, but notes that this does not necessarily preclude its inclusion in the suite of available resolution tools. It advocates for taking a conservative approach to implementing such features in Australia.

A very large number of submissions are concerned that introducing bail-in provisions in Australia could lead to depositors’ funds being bailed in to recapitalise a failed bank. The Inquiry strongly supports continuing the current Australian framework in which deposits are protected through an explicit guarantee under the FCS, supported by depositor preference. The Inquiry specifically does not recommend the bail-in of deposits.

The ultimate shape of the framework will influence the cost-benefit analysis. In assessing this, the most relevant factors are implicit guarantees, funding costs, lending rates, GDP and credit ratings.

Benefits

If banks have sufficient loss absorbing and recapitalisation capacity, a failed ADI is more likely to be resolved in a way that limits the effect of the failure on the broader economy, while minimising the use of taxpayer funds. This is a substantial benefit. As detailed in Recommendation 1: Capital levels, the costs of financial crises are wide ranging and severe. That recommendation focuses on reducing the probability of crises occurring in the first place, while this recommendation focuses on reducing the costs of crises that cannot be avoided. The magnitude of these avoided costs will depend on the specifics of the framework implemented. As an indicative measure, if the cost of financial crises is reduced by 10 per cent, it would provide an expected average benefit of 0.25–0.3 per cent of GDP per year ($4–$5 billion).55

By making it more credible to achieve a resolution with minimal use of taxpayer funds, this recommendation also reduces perceptions of an implicit Government guarantee. There are clear benefits to the economy in minimising perceptions of implicit guarantees, including reducing Government’s contingent liability and improving efficiency by removing market distortions, thereby making the banking sector more competitive.

Lending interest rates

Making it more credible and feasible for creditors to bear losses would raise the costs of the relevant types of debt funding for affected ADIs by reducing perceptions of an implicit guarantee.

Higher ADI funding costs could result in small increases in loan prices for customers. Banks have acknowledged in submissions that the cost of other forms of regulatory capital would be less than the cost of increasing CET1 capital; the funding spread, and corresponding effect on lending interest rates, for subordinated debt is a fraction that of CET1 capital. Competitive pressure could see banks share some of the cost with investors through a lower ROE. Thus, the effect on loan interest rates is likely to be limited, even for a large increase in bail-in debt.

From an economy-wide view, reducing the implicit guarantee would offset at least part of the cost to banks by providing a corresponding benefit to taxpayers and Government, and reducing market inefficiencies. Greater volumes of new subordinated debt could also reduce the cost of existing subordinated debt on issue, since potential losses would be spread across a larger pool of claims. It should also reduce the cost of more senior debt, as losses become less likely to reach senior classes.

The Inquiry notes that markets for subordinated debt with conversion and write-off features are currently small and may require higher spreads to absorb large new issuance. This would particularly be the case if new requirements were implemented with short transitional arrangements.

GDP

The Inquiry expects the effect of higher lending rates on GDP to be minimal. An upper bound would be to assume that the full funding cost increase is passed through to loan interest rates, and that the RBA does not offset this through its setting of monetary policy. As discussed in Recommendation 1: Capital levels, the small expected effect on lending interest rates would lead to a correspondingly small effect on GDP.

However, a large part of the cost is offset by reductions in perceptions of an implicit guarantee. In addition, the RBA would likely consider the effect on GDP when formulating monetary policy.56

Credit ratings

The net effect on credit ratings is unclear. Debt designed to more easily expose creditors to loss through write-off or conversion features is likely to be rated lower than debt without these features. However, it is not clear whether banks’ credit ratings, which are based on the risk of loss to senior unsecured debt, would change as a result of introducing a loss absorbing and recapitalisation framework.

If the loss absorbing and recapitalisation framework increases ADIs’ subordinated debt, there would be a larger buffer before senior unsecured debt takes losses. It may therefore make senior debt safer. However, introducing the framework may be taken as a signal of a lower likelihood of Government support for banks, especially since this is an intended outcome of the framework. Currently, the major banks receive a two-notch credit rating upgrade on the basis of expected Government support.57 Credit rating agencies may reconsider this upgrade in light of credible mechanisms to impose loss.

Conclusion

The Inquiry judges that there is a net benefit of a loss absorbing and recapitalisation capacity framework.

Loss absorbing and recapitalisation capacity on its own does not guarantee a successful resolution nor eliminate all perceptions of implicit guarantees. It must be part of a broader resolution framework that includes strong crisis management tools for regulators, as outlined in Recommendation 5: Crisis management toolkit.

The extent of the net benefit will be influenced by the ultimate shape of the framework, including the quantum, the composition and the time given for transition. Generally, costs will be higher the larger the capacity required, the more subordinated the eligible instruments and the shorter the period required to build the capacity. Benefits will be greater where loss absorbing and recapitalisation capacity is high and clear, and where creditor hierarchy and triggers are transparent. However, if designed carefully and according to the articulated principles, the framework can attain net benefits.

Implementation considerations

Sufficiency

To minimise the need for taxpayer support, ADIs need sufficient capacity to absorb losses and, in some cases, provide the recapitalisation necessary to implement their resolution strategy.

This may require enough capacity to fully recapitalise the institution. International work proposes that G-SIBs need a range of 16–20 per cent of risk-weighted assets and twice the Basel leverage requirement.58 A similar quantum may be appropriate for internationally active Australian ADIs.

For smaller banks, an orderly resolution may be possible through activating the FCS or through a merger or acquisition at the point of resolution. In this case, the loss absorbing and recapitalisation capacity sufficient to implement the resolution plan is likely to be lower.

The Inquiry recommends considering a graduated approach across the banking sector when developing the loss absorbing and recapitalisation capacity framework for Australia. This approach should take into account the likely resolution strategy for an ADI.

Eligible instruments

The framework should consider a broad range of equity and debt instruments.

Equity instruments have the advantage of being well understood by investors. These instruments have a long history of automatically absorbing loss without causing systemic disruption. However, they are more expensive than debt funding and may not be available in resolution, having already been depleted. Experiences overseas suggest that ADIs only tend to enter resolution after significant losses have been incurred and there is little or no equity value left.59 That is, equity instruments may not be available to assist in recapitalising a distressed institution.

Requiring eligible debt instruments would give the regulator greater confidence that the loss absorbing and recapitalisation capacity will be available in resolution. These instruments are not depleted until a trigger has occurred, so — once triggered — they can act to replenish capital. This gives the regulator greater certainty about the resources that will be available when conducting their resolution planning. Debt instruments are also typically less expensive than equity instruments.

Additional Tier 1 and Tier 2 capital instruments with conversion and write-off features, which already exist in the Basel framework, can provide loss absorbing and recapitalisation capacity. Investors already hold these instruments. As these conversion features are relatively new, instances of instruments being converted into equity or written off are very limited. If constructed carefully, a new layer of contractual instrument in the creditor hierarchy between Tier 2 and unsecured senior debt would have similar benefits to Tier 2, at a lower cost. In substance, it should be no less credible than a Tier 2 instrument.

Addressing challenges

Stakeholder submissions, and a wide range of policy research and commentary, note a number of major difficulties in implementing a bail-in regime that can be credibly activated in a crisis.

Most concerning is the possibility that activating a bail-in for creditors of one bank may actually worsen the crisis. This could occur if converting one bank’s creditors caused creditors of other banks to reassess the likelihood that they will take a loss, resulting in investors withdrawing funds (or refusing to roll over debt) to other banks in the system. This contagion could cause acute liquidity problems and distress in other banks, exacerbating the crisis. Also, if banks were unable to access international funding markets, it could take longer for them to resume lending to the economy once the crisis is over, potentially prolonging an economic downturn.

Addressing these challenges is critical to developing a viable loss absorbing and recapitalisation capacity framework. Although such difficulties give reasons to be cautious, in the Inquiry’s view these can be addressed, especially given the considerable work underway on these issues globally. In developing its own framework, Australia should take account of this international work to create a system that, where possible, overcomes the problems associated with bail-in by being credible, predictable, in line with international practice, and having an appropriate transition period.

Other considerations

To keep any costs to a minimum, an appropriate implementation period should be allowed where the framework imposes a significant quantum.

Developing a successful loss absorbing recapitalisation framework depends on a large number of other important aspects, which this Final Report will not discuss in detail. These aspects include:

  • Possible need for legislative change; for example, to ensure certainty of the creditor hierarchy.
  • Considering whether requirements form part of Pillar 1 or Pillar 2 requirements.
  • Ensuring the legal basis for exposing creditors to loss is sound and the framework adequately accounts for where an ADI is part of a group or operates across borders.

52 Bank of England 2014, The Bank of England’s approach to resolution, Bank of England, London, page 7.

53 Financial Stability Board (FSB) 2014, Adequacy of loss absorbing capacity of global systemically important banks in resolution, FSB, Basel.

54 For example, see Standard & Poor’s 2013, Australia’s developing crisis-management framework for banks could moderate the Government support factored into ratings, Standard & Poor’s.

55 Based on the expected average cost of a financial crisis of 2½–3 per cent of GDP ($40–$50 billion) per year, as outlined in Recommendation 1: Capital levels.

56 For example, Battellino, R 2009, Some comments on bank funding, remarks to the 22nd Australasian Finance and Banking Conference, 16 December, Sydney; Hansard 2009, Reference: Reserve Bank of Australia annual report 2008, House of Representatives Standing Committee on Economics, 20 February, Canberra.

57 Standard & Poor’s 2013, Australia’s developing crisis-management framework for banks could moderate the Government support factored into ratings, Standard & Poor’s.

58 Financial Stability Board 2014, Adequacy of loss absorbing capacity of global systemically important banks in resolution, FSB, Basel, page 6.

59 Gracie, A 2014, Making resolution work in Europe and beyond — the case for gone concern loss absorbing capacity, speech given at the Bruegel breakfast panel event, 17 July, Brussels.