How to cope with the too-big-to-fail problem?

Comments by Mr Stephen G Cecchetti, Economic Adviser and Head of Monetary and Economic Department of the BIS, prepared for the 10th Annual Conference of the International Association of Deposit Insurers, "Beyond the Crisis: The Need for a Strengthened Financial Stability Framework", Warsaw, Poland, 19 October 2011.

Introduction¹

It is a pleasure to be able to contribute to the 10th Annual Conference of the International Association of Deposit Insurers. I believe that "too big to fail" - the focus of this panel - is the single most important policy issue that has emerged from the crisis. In a market-based financial system, the right to succeed is the right to fail. The orderly entry and exit of firms, combined with an appropriate relationship between risk and return, means that risk-takers that stand to profit also stand to lose.

The too-big-to-fail problem and the associated moral hazard costs affect these core preconditions for competitive markets. That is why addressing the too-big-to-fail problem is of fundamental importance.

Before getting too far, let me pause to say that I will use the term "too big to fail" in a broad sense. Clearly, being too big is a major part of the problem, but it is not all just about size. Excessive interconnectedness of financial institutions, reliance on a single or few firms for the provision of key financial infrastructure, and complexity of operations and cross-border activity are all part of what I will refer to as "too big to fail". In combination, all these characteristics of a financial institution raise the impact of its failure on the financial system, and thereby give rise to the too-big-to-fail problem.

I should note also that there is a sense in which this session, "How to cope with the too-big-to-fail problem?", is mislabelled. We cannot and should not merely cope with institutions that are too big or too interconnected to fail. Rather, we should force these institutions to internalise the externality they are creating and to face head on the associated systemic risk. While firms are free to choose their business models, they should be compelled to pay for the externalities they create.

Why are additional measures needed for systemically important banks?

The rationale for putting in place specific policy measures for banks considered too big to fail is based on externalities which Basel III does not directly address. Basel III sets minimum requirements for the ratio of risk-weighted assets to common equity Tier 1 capital. It therefore meets the microprudential, institution-specific objective of addressing the traditional tendencies of managers to take on too much risk. Elements such as limited liability and deposit insurance give rise to such inappropriate risk-taking incentives. Basel III does not, however, capture the risk to others, or to the system as a whole, created by an individual institutional failure - though policies designed to target specific financial market externalities directly are difficult to implement, as the externalities themselves are difficult to observe and quantify.

In the light of such uncertainty and measurement problems, the objectives of regulatory policies developed to address the too-big-to-fail problem have been designed to:

(i) reduce the probability of failure of global systemically important banks (G-SIBs);

(ii) reduce the extent or impact of the failure of such G-SIBs; and

(iii) level the playing field by reducing the competitive advantages in funding markets that these institutions have.

The combination of capital surcharges, better resolution regimes, living wills, more robust financial market infrastructures and central clearing, and more intense supervision of systemically important financial institutions (SIFIs) together will contribute to achieving these objectives.

In the remainder of these brief remarks, I will expand on each of these three key objectives and describe the policy responses developed by the Basel Committee on Banking Supervision and the Financial Stability Board (FSB).

Reducing the probability of failure of G-SIBs

Reducing the probability of failure of G-SIBs is the cornerstone of the regulatory response to the too-big-to-fail problem. Raising the amount of going-concern capital for these institutions through the application of a capital surcharge will lower their probability of failure. This in turn will lower the ex ante expected impact of their failure.

The Basel Committee has developed a methodology to identify G-SIBs which brings together a number of indicators that proxy for the systemic importance of a bank. These are: size, interconnectedness, substitutability, global activity and complexity. Based on this methodology, G-SIBs are allocated into buckets according to their relative systemic importance. The proposal is to allocate banks to four buckets and apply a surcharge ranging from 1 to 2.5%. In addition, an initially empty bucket sits at the top, with a surcharge of 3.5% as a disincentive to a G-SIB becoming even more systemically important.

To see what this means, take the example of a bank that faces a 2% surcharge. Such a bank would face a 4½% minimum, plus a 2½% conservation buffer, plus the 2% surcharge, making a total of 9%. Taking into account Basel III's tougher definition of capital, the result is a substantial and necessary increase in minimum requirements.

Some jurisdictions have announced their intention to require even higher capital requirements. This is in line with the fact that Basel III sets a minimum and that some countries' banking systems are very large relative to the rest of their economy. That is, in some places, banks are not only too big to fail, they are also too big to save.

Reducing the impact of failure

The simplest way to reduce the impact of a firm's failure is to reduce its systemic importance directly (eg by placing limits on the firm's size or business functions).

Restrictions on the activities that banks can undertake have been proposed in some countries. For example, the Vickers Report in the United Kingdom proposes ring-fencing traditional retail banking business activities. The Volcker Rule in the United States proposes restrictions on proprietary trading by banks and limits on owning and investing in hedge funds.

Such proposals aim to separate fundamental, essential banking services from more speculative investment activity. The aim is to reduce the impact of the failure of certain banks and the potential cross-subsidisation from safe retail banking business to riskier wholesale banking functions and investments. In the same way that an airline company should restrict its operations in oil and currency futures markets to hedging its profits from flying people around the world, the idea behind such proposals is that a bank should confine itself to activities that serve its customers' needs and not use deposits as a source of funds for its proprietary trading operations.

At the international level, efforts to reduce the impact of a G-SIB's failure have focused on improving recovery and resolution regimes and promoting bail-in within resolution. These measures target the problem that certain firms are difficult to resolve or place into resolution. This applies in particular to large, complex cross-border firms.

While I think it is fair to say that we remain a long way from achieving a global cross-border resolution regime, a number of jurisdictions are carrying out reforms of their national resolution regimes to enhance their resolution powers. This process has been facilitated by the FSB's release of the Key attributes of effective resolution regimes, which sets out new international standards for the resolution of distressed financial institutions. These measures are complementary to, and not substitutes for, higher loss absorption capacity.

Levelling the playing field by reducing too-big-to-fail funding advantages

Finally, let me turn to level-playing-field considerations. A number of studies have attempted to quantify the funding advantages enjoyed by banks that are perceived as too big to fail (see, for example, Ueda and Weder di Mauro (2011)²). The conclusion is that these advantages are significant, with a funding subsidy of as much as 60 basis points during normal times and more during crisis periods. After all, if an elderly relative asked you where to deposit their savings, wouldn't you tell them to put their deposit in an institution that you thought the government would be likely to support in a crisis?

The policy responses I have discussed - the capital surcharge, restrictions on certain business activities, and improvements in recovery and resolution regimes - all help to reduce this subsidy. Rather than think of a systemic capital surcharge as disadvantaging big banks, think of it as making things fairer for small banks.

In addition, the Basel III framework now requires all regulatory capital to fully absorb losses at the point of non-viability before taxpayers are exposed to loss. This can be achieved either through contractual means or via a statutory resolution regime. It seeks to address the problem that, during the financial crisis, Tier 2 capital instruments (mainly subordinated debt), and in some cases Tier 1 instruments, did not absorb losses incurred by certain large internationally active banks. The work on resolution and bail-in would extend gone-concern loss-absorbing capacity to other parts of the capital structure.

Concluding thoughts

Prior to the crisis, numerous academic studies and banking textbooks discussed the too-big-to-fail problem and moral hazard more generally. However, I am sure that, even for those who have written about these issues for many years, the true depth and seriousness of the concerns were only revealed during the recent financial crisis. I remain somewhat surprised to hear the occasional voices which claim that the too-big-to-fail problem is overstated.

It is imperative that we not only cope with the too-big-to-fail problem, but that we also deal with it effectively. The capital surcharge for global systemically important banks introduced by the Basel Committee is a significant step in the right direction. The same is true of the progress on improving recovery and resolution planning.

Finally, the Macroeconomic Assessment Group, which I chair, has issued its final report on the economic impact of requiring additional loss absorbency for global systemically important banks. The results show that the transitional costs of higher capital requirements for global systemically important banks are very small, and that the long-term economic benefits are very large.

So, let me end where I started: too big to fail is too big to exist.

¹ Comments prepared for the 10th Annual Conference of the International Association of Deposit Insurers, "Beyond the Crisis: The Need for a Strengthened Financial Stability Framework". I would like to thank Neil Esho for his help in preparing these remarks. The views expressed here are those of the author and do not necessarily reflect those of the BIS.

² K Ueda and B Weder di Mauro, "Quantifying the value of the subsidy for systemically important financial institutions", mimeo, 2011.