19 August 2013

Bundesbank/Dombret: Systemic risk, too-big-to-fail and resolution regimes

While concepts have been developed to initiate regulatory reforms, the "too-big-to-fail" problem still remains unresolved. Market participants continue to anticipate that governments will rescue systemically important financial institutions – or SIFIs – in the event of their failure.

The resulting refinancing advantage is reflected in so-called rating "uplifts". Rating agencies usually calculate two different ratings for banks. One is a "stand-alone" rating that measures a bank’s genuine creditworthiness. The other is the "all-in" rating which includes the likelihood and extent of external support available for the bank’s debt. Although these uplift factors have recently shrunk to some degree, they unfortunately remain substantial.

However, in this regard we need to achieve two objectives at the same time. First, taxpayers should not have to foot the bill for bank failures and, second, systemic disruptions must be avoided. The experiences following the Lehman collapse five years ago show how important financial stability is – and how fragile. So, what can we do about the "too-big-to-fail" problem?

Solving the "too-big-to-fail" problem

I am convinced that overcoming the "too-big-to-fail" problem will require a multi-track approach. The main goal is to make SIFIs less likely to fail by increasing their loss-absorbing capacity. Basel III represents a landmark change in this respect. These rules are much more rigorous than any previous regulation, both in quantitative and qualitative terms. On top of this, global SIFIs face additional capital requirements, known as SIFI surcharges. Combined with other measures, such as more intensive supervision, these changes will enable banks to better cope with stress situations.

Basel III substantially raises capital requirements, and impact studies were carried out before the rules were finally endorsed. Now we should let Basel III take effect. Looking at the issue of complexity, it is true that risk measurement will never be perfect and that relying on internal models hampers comparability, among other things. However, simplicity can come at a cost, too, as it disregards the overarching need for risk sensitivity. Therefore, we need to strike a balance between risk sensitivity, simplicity and comparability. But the spirit of the Basel rules, particularly regarding risk sensitivity, should not be compromised, nor must their full implementation be called into question.

While it is necessary to increase banks’ resilience, that alone is not enough to solve the problem. There is a second broad goal which is now increasingly recognised: we have to ensure that SIFIs can be resolved without disrupting the financial markets.

But before discussing more specifically how to create effective resolution regimes, let me briefly touch on some suggestions for separating commercial from investment banking. Proponents of this approach believe that separating deposit-taking and lending from investment banking would prevent spill-over effects. The idea is to create a category of rather traditional banks whose customers would be protected by deposit insurance schemes. On the other hand, those banks engaged in riskier and more volatile business could not rely on deposits, nor would they be rescued at the taxpayers’ expense.

However, as the boundaries between various banking activities are fluid it is difficult to draw a clear line between them. Consequently, structural interventions in banks’ business models must be carefully designed. They might help to solve the too-big-to-fail problem, but they are by no means a magic bullet.

How to resolve a SIFI

Without any doubt, we need effective resolution regimes for financial institutions. Unfortunately, the crisis has revealed a significant lack of suitable resolution instruments, especially in a cross-border context. This is why the G20 leaders, back in 2011, endorsed the Financial Stability Board’s "Key Attributes of Effective Resolution Regimes for Financial Institutions" as a reference point for national resolution regimes. They set out core elements of national resolution regimes on a global level.

Jurisdictions around the globe are currently contemplating their preferred resolution strategy for systemically important banks. Two stylised models have recently been set out by the Financial Stability Board: a Single Point of Entry- and a Multiple Point of Entry-strategy. The question behind these strategies is whether resolution tools will be applied by a single authority at the top level of a failing bank or whether they will be applied in a coordinated manner by more than one authority at the level of regional or national units of the bank.

The Key Attributes are quite a step forward. Implementing them will gradually align national resolution regimes. I am hopeful that this will significantly curtail the ability of financial institutions to hold taxpayers to ransom.

In Europe, the implementation of resolution rules has been carried out through ECOFIN’s agreement of 27 June 2013. The Council adopted its general approach on the draft directive for the recovery and resolution of credit institutions and investment firms. I welcome the general thrust of the Recovery and Resolution Directive – or RRD. And I hope that the trilogue process with the European Parliament can be completed swiftly, as planned. Allow me to look in more detail at three key issues relating to the RRD.

1. The bail-in tool: this enables resolution authorities to write down the claims of shareholders and write down, or convert into equity, the claims of creditors of institutions which are failing or likely to fail. As predictability is crucial when it comes to allocating losses, I very much agree with the Council’s general approach stipulating a clear pecking order. Shareholders will be the first in line to bear losses, followed by clearly defined classes of creditors.
2. I welcome the minimum requirement for own funds and so-called eligible liabilities. This will ensure that each institution has sufficient loss-absorbing capacity based on its size, risk and business model.
3. I wish to comment on resolution funds. As a general rule, the draft RRD requires member states to set up resolution funds, which are to be funded ex-ante by banks. Within ten years they should reach a target level of 0.8 per cent of covered deposits. In my view, this mixing of public and private funds for resolution funding purposes is against the spirit of the RRD. Moreover, and to be frank, I wonder how this exemption can be applied in practice.

All in all, the draft RRD is quite close to striking a sound balance between the conflicting objectives of harmonisation and flexible rulemaking.

I firmly believe we should entrust a newly established European institution with resolution powers as laid out in the RRD. It must become a strong and independent body with full decision-making powers. If we agree this to be the objective, let’s find constructive ways to bridge an interim period until this can be realised.

Full speech

© Deutsche Bundesbank