Global consistency in financial regulation: is the glass half full, half empty, or just more transparent?

Remarks by Wayne Byres, Secretary General, Basel Committee on Banking Supervision, at the RiskMinds Risk and Regulation Forum, Nice, France, 10 September 2013.

Thank you very much for the opportunity to speak this morning on what is a very important, and much discussed, topic.

I would usually describe bankers as 'glass half full' people, and regulators as their 'glass half empty' counterparts. To take an obvious example, where banks might first see opportunity for reward, regulators will first see exposure to risk. The two perspectives are, of course, just different sides of the same coin. But they are perspectives that define and create the healthy tension in the bank/regulator relationship.

When it comes to recent regulatory reforms, though, I find the situation strangely reversed - bankers are somewhat despondent, lamenting gaps and inconsistencies, while regulators are focusing on the good and steady progress that has been made. We are looking at the same objective facts, so when it comes to the consistency in the implementation of global regulatory reforms, should we see the glass as half full or half empty? And what is our benchmark when we say that things are getting better or worse? When we look at prudential and regulatory outcomes, how much of any problem is caused by regulators, and how much stems from banks themselves? I plan to offer some thoughts on these questions in my remarks this morning.

The steady path to increasing cross-border consistency in financial regulation

This year marks the 25th birthday of the first Basel Capital Accord. Since its delivery, there has been a continuous effort to improve and enhance global policy-making and coordination. In more recent years, the work of the Basel Committee and other international standard-setters has produced a large number of agreements on a wide range of issues. It is something of a pity that it took a global financial crisis to deliver, for example, the world's first global liquidity standard for banks, but we now have one where there was nothing previously, and that is a big step forward on the path to increased consistency in financial regulation.

The quest for consistency does not, however, imply the eventual development of a single global rulebook that can be applied without exception to all activities of all banks in all jurisdictions at all times. National differences in financial regulation continue to exist, and many of them are there for very good reasons. We therefore need to think carefully about our benchmark for any report card on global consistency - if we aspire to complete and absolute consistency in all cases, we will inevitably be disappointed.

For many this will be self-evident, but it is worth listing a few of the key benefits that consistent cross-border financial regulation can offer to society (and note that it is the benefits to society, and not to banks per se, for whom the benefits need to be measured). For example:

• most obviously, it promotes global financial stability;
• it supports cross-border trade, by promoting international financial markets to match those in goods and services;
• it improves the efficiency of capital allocation, by allowing capital to flow more readily across national boundaries to find its most productive use; and
• it facilitates more competitive financial markets, allowing consumers and investors access to the greatest range of financial services at the lowest cost.

Of course, there can also be reasons why complete harmonisation would be sub-optimal:

• national financial systems are at different stages of economic development, and may need differences in regulatory and supervisory approaches even if their financial stability objectives are the same;
• financial regulation does not exist in isolation, and needs to be blended with national tax, accounting and legal frameworks, amongst others, to achieve appropriate outcomes; and
• it would not provide room for macro-prudential (or micro-prudential) adjustments to regulatory requirements to deal with vulnerabilities and risks arising from the different stages of business and financial cycles in individual jurisdictions.

Furthermore, international standards such as those produced by the Basel Committee are typically minimum standards. It has never been the Committee's goal to try to constrain national authorities who want to go beyond the internationally agreed minimums for their own domestic reasons.

So there are clearly limits to the complete harmonisation of regulation. Over the past few decades, however, as banking and finance have become more global, the balance has shifted in favour of increased cross-border consistency to support the benefits that global financial markets can offer to society. And in a world in which, more recently, there has been legitimate concern about fragmentation, it is robust and consistently-applied international standards that provide the means by which fragmentation can be resisted.

That does not mean the task of international policy-making is getting any easier. The challenge can be no better demonstrated than by looking at recent correspondence to the Basel Committee. I recently received a letter from a prominent industry association, claiming that divergences in national regulation are on the increase, and that the Basel Committee, along with other international bodies, must do more to force convergence. Yet, as we publicly consult on regulatory proposals, a phrase that often appears in the submissions we receive is "one size doesn't fit all", usually accompanied by a request that Basel proposals grant more flexibility to account for domestic specificities. The more amusing requests manage to combine both - these boil down to 'I want consistency, as long as everyone does it my way.'

Let me assure you that within the Basel Committee, as well as in other international groups that I participate in, the commitment is undiminished to developing robust standards that can be implemented consistently across major markets. International policy-making is not easy, and proposals simply don't get off the ground without a strong buy-in from everyone at the table. The fact that all international groups continue to have active agendas should be taken as a clear sign that the momentum for internationally-consistent financial regulation remains strong.

If we look simply at what Basel III has done to promote greater cross-border consistency, we have made some major steps forward:

• Beyond the primary objective of strengthening bank capital, Basel III also included important measures to improve the consistency of bank capital measurement (eg by standardising deductions made from banks' capital bases) and disclosure.
• Before the crisis, some jurisdictions operated with leverage ratios, some did not, and accounting differences clouded what reporting there was of bank leverage. Now we are working towards a consistent definition, and minimum requirement, against which bank leverage can be monitored and compared. As with capital, these will be accompanied by standardised minimum disclosure requirements.
• Almost all jurisdictions had some form of short-term liquidity requirement, and some also had a structural funding measure. The only consistent thing was that all were different. In the Liquidity Coverage Ratio and the Net Stable Funding Ratio, we will have agreed standard measures and disclosures that will apply to banks in all key markets.

Beyond Basel III, the regulatory community has also agreed, or is working towards:

• an internationally agreed framework for dealing with global systemically important banks, an agreed set of principles for domestic SIBs, and an emerging framework for systemically important insurers;
• principles for financial market infrastructure, designed to promote the safe and efficient operation of FMIs;
• international agreement on the key attributes for resolution frameworks;
• standardised margin requirements for non-cleared OTC derivatives;
• internationally agreed minimum haircuts for repos;
• consistent large exposure limits; and
• improved disclosure requirements, designed to provide more meaningful and comparable information to key stakeholders.

All of these initiatives are greatly strengthening the global consistency of financial regulation.

Greater focus on consistent implementation of global standards

I don't think there is any disagreement that international policymakers have been busy. If there is, I hope that list dispels the idea. But - and I hear this regularly - that may be all well and good: what really counts is whether global agreements are being faithfully implemented in national regulations.

That message is well understood. Indeed, it is why, for example, every communiqué from the G20 Leaders since 2009 has stressed the importance of implementing internationally agreed reforms; more recently, the G20 Leaders have regularly endorsed the full, timely and consistent implementation of Basel III. And it is why the post-crisis role of standard-setting bodies is changing. That is, standard setters are tasked not only with forging agreement on international standards, but also with actively monitoring their implementation - a very important new mandate.

In the case of the Basel Committee, this is being done in a number of ways through its Regulatory Consistency Assessment Programme (RCAP).

First, there are our simple semi-annual reports showing how jurisdictions are tracking relative to the agreed implementation timetable for Basel III.[1] Even this simple device has had positive effect; I know of a number of examples where the imminent release of another Basel Committee 'report card' has caused national regulators to make sure they moved forward with the next step of their rule-making process so as to be seen to be making progress.

More substantially, the Committee has established a programme of peer assessments of national regulations against the Basel standards.  At present, this involves a comprehensive and detailed assessment of the risk-based capital requirements; in due course, it will expand to cover leverage, the LCR, NSFR and SIB requirements as well. These assessments, as anyone who has read one will know, are essentially a line-by-line assessment of the national regulations, and where a difference is found that is potentially less stringent than the minimum Basel standard, an assessment of the actual or potential materiality. Again, these assessments are having a real and positive impact: not only does the prospect of a subsequent peer review mean that jurisdictions have often sought views on the consistency of their draft regulations before implementing them, but the peer reviews of final regulations completed thus far have all led to national authorities making changes to their local requirements to correct unintended divergences from the Basel standards.

The third leg of the work is operating at the level of individual banks, looking at the consistency of bank-level capital calculations (specifically, of risk-weighted assets determined using internal models). I will say more about this shortly.

I hesitate to call this new monitoring initiative a policing role, since we have no enforcement power beyond the power of peer pressure and public disclosure. But it is nevertheless a strong demonstration of the commitment to international consistency and, where this cannot be perfectly achieved, to much greater transparency. In particular, if an individual jurisdiction departs from internationally-agreed standards, the nature and materiality of that divergence should be well understood. What we doing can be thought of as encouraging greater 'truth in advertising' - if a bank does not operate under regulations that are consistent with Basel standards, any difference should be much more transparent when it reports a 'Basel ratio'. That way markets can provide something of a policing role, where necessary offsetting regulatory differences in their assessment of banks' financial ratios.

The advent of this monitoring work makes me wonder whether regulations are really getting more disparate, as many claim. Maybe it is possible to count more instances where jurisdiction X differs from internationally-agreed standards but, then again, there are now more standards to be complied with. I have already heard grumbling about marginally inconsistent implementation of the LCR, but this needs to be considered against a pre-2010 world in which there was no standardisation of liquidity requirements whatsoever.

That is not meant to sound blasé about the need to pursue greater consistency. Since regulation ultimately remain in the hands of national authorities, domestic perspectives will tend to outweigh international ones without due care and attention. Continual effort to emphasise the need for international consistency, where it can be achieved, is important if the benefits to society that I spoke about earlier are to be maximised.

Different differences

When looking at the international consistency of financial regulation, differences essentially arise in four circumstances.  These variations derive from:

• considerations related to the systemic importance of individual institutions - examples include the so-called 'Swiss finish' to Basel III, and recent proposals with respect to higher leverage ratios for the very largest US banks;
• macro-prudential considerations - examples include the capital buffer (Switzerland) and higher risk weights (eg Hong Kong SAR, Sweden) for residential mortgage lending;
• super equivalent/gold plating of international standards - examples include the many jurisdictions (particularly in Asia) that have set higher minimum capital ratios than called for in the Basel framework; and
• shortcomings in implementation - for example where the standards have not been implemented (or are slow to be implemented), or where they have been modified in a way that makes them weaker than the internationally agreed minimums.

Here is where the apparent disconnect between the industry's 'glass half empty' view and the regulator's 'glass half full' view becomes understandable. Regulators focus primarily on the last category - gaps in the framework that mean minimum standards are not being met, creating risks to both financial stability and the level playing field. By contrast, we tend to hear much more from industry participants on the competitive and operational problems caused by the other three sources of variation.

In the current environment, there is a particular concern about a range of initiatives that could be grouped under the broad heading of 'ring-fencing'. These potentially risk a degree of fragmentation in global markets, jeopardising the benefits that these markets bring. To the extent these initiatives go beyond the minimum standard prescribed by international standard setters - which in banking tend to focus on consolidated supervision of a global group - they create challenging issues for the regulatory community to debate. Standard setters such as the Basel Committee are tasked with agreeing what must be done, at a minimum. They are not structured to preclude or prevent domestic initiatives that go beyond this. Yet domestic regulatory initiatives such as ring-fencing, like other policies, can have spillover effects across national borders. Where these spillovers are significant, they provide a strong signal where increased international cooperation is needed, as domestic initiatives of this type will typically be in response to a perception that the international regulatory framework is lacking in some shape or form.

The role of model-driven differences

In defending the efforts of regulators to develop more consistent financial regulation, I am also the first to acknowledge that more could be done. But recent work by the Committee shows that whatever regulatory differences exist, they may well be outweighed by the differences in regulatory outcomes driven by the use of internal models.

I am sure that many of you will be aware of the studies produced by the Committee on the consistency of risk-weighted asset calculations in the banking and trading books. These focus on the extent to which differences in the internal models used by banks to determine their regulatory capital are producing material differences that do not reflect differences in underlying risk. This was done by using hypothetical portfolio exercises (HPEs), which allowed the Committee to observe banks' estimated risk weights and capital requirements for identical portfolios.

The studies are available on the Basel Committee's website, and contain far more material than I can present today. Let me just focus on the headline messages:

• there are material differences from internal modelling practices that do not reflect underlying differences in risk; and
• public disclosures are insufficient to allow investors and other interested parties to adequately understand the impact of these modelling choices.

The chart below comes from the study on risk weighted assets in the banking book. It shows the hypothetical impact on bank capital ratios if a bank's actual risk weights for wholesale credit portfolios (sovereigns, banks and corporates) were replaced by the median risk weights used by the bank's peers.

Clearly, differences in risk-weights can have a material impact on reported capital ratios. Even after allowing for the removal of an outlier bank at the bottom end, the range is wide.  While some degree of variability in risk-weighted assets derived from internal models is expected - and indeed desirable - the current situation looks less than ideal, particularly if the model-driven differences are not transparent to the users of bank capital disclosures.

Let me give you another, more micro, example of the problem. When looking at the risk weights applied to the four largest G-SIBs (that is, the G-SIBs currently in the top bucket of the G-SIB framework), the Basel Committee's study found a very wide degree of divergence in the risk weights being used by banks. If we drop off the bottom and top quartiles of the distribution and focus just on the inter-quartile range, we still find that the top of the range is around twice the bottom of the range. In other words, even if we forget half the sample and just focus on the banks that most agree with one another about the appropriate risk weights for each of these G-SIBs, some are still using risk weights half/twice that of their competitors. And this is for the largest, best known banks, ones that provide the market with substantial disclosures about their activities and financial health!

I raise this point because, in looking at the problems of inconsistency in regulation, it is important to look not just at instances of different sets of regulations, but also at the inconsistency introduced by the use of models within the regulatory framework. Unlike impacts that stem from differences in the regulations that apply to individual jurisdictions (and are therefore at least consistently applied to all banks within that jurisdiction), modelling differences do not respect national borders. Indeed, the Committee's studies show that there can be substantial differences even within national boundaries. And unlike differences in national rules - which can be understood by anyone willing to read of individual regulations carefully - differences in modelling approaches are far more opaque, making the impact much more difficult to assess and quantify.

For that reason, the Committee is currently looking at a range of measures to reduce undesirable practice-based variations in RWAs, and to improve the comparability of regulatory capital calculations by banks. These includes:

• enhanced disclosures by banks to foster greater market discipline and prevent misperceptions as to the level and causes of RWA variations. In addition, use of standardised definitions and templates could support greater consistency and comparability of disclosures. These thoughts are consistent with the recommendations of the Enhanced Disclosure Task Force report of 2012;
• additional guidance on aspects of the Basel framework. Examples include adjustment of risk parameters for conservatism or cyclical effects, and use of external data, particularly for low default portfolios. It may be appropriate for the Committee to provide additional guidance to reduce or eliminate undesirable variation attributable to such differences;
• harmonisation of national implementation requirements. Examples where additional clarity could be provided include capital floor adjustments, partial use of the standardised approach, definition of default, treatment of defaulted exposures, exemptions from the one-year maturity floor, and requirements related to the estimation of IRB parameters. Many of these drivers could be addressed through clarification of the framework, through efforts to harmonise national implementation requirements, or through review of the continued relevance of various aspects of national discretion; and
• constraints on model parameters could also be considered. For example, supervisory benchmarks for risk parameters could be created from the data collected through the Committee's work. Creation of such benchmarks could fill a valuable niche, for example, for low-default IRB portfolios, creating reference points for supervisors and banks. Other alternatives could include more explicit constraints, such as the creation of floors for certain parameters (such as LGD), or even fixed values of such parameters.

 As for when you will see the results of the Committee's thinking on these issues, it will vary. I expect that the Committee's response will entail a series of incremental steps, rather than a single 'big bang' policy response. This is because of the multi-faceted nature of the issue, and because some issues touch on policy work already underway. For example, the fundamental review of the trading book has been generating new proposals for market risk capital requirements; these are being refined with the findings of the recent study on trading book risk weights in mind. Similarly, the Committee recently established a working group to review the Pillar 3 framework; this will factor into its work the findings on the adequacy of existing disclosure. The staggered approach allows the Committee to respond to a number of the issues more promptly than if we waited for every piece of the puzzle to be fully specified, thereby generating continual, incremental improvements to the consistency of bank capital requirements as quickly as they can be achieved.

Concluding remarks

There are many reasons why there are differences in the national implementation of global standards. Some of these make good sense, others should be avoided if possible. We should not be aiming for a single, monolithic global rule book: there are very good reasons why complete and absolute harmonisation across all banks, all jurisdictions and all times is not a good idea if we want to maximise financial stability. If we were to seek complete harmonisation everywhere and at all times as the goal, our glass would always be seen as (at least) half empty.

But my personal view is that the glass is more than half full. The advent of global liquidity, leverage and funding standards, for example, where previously these had been left entirely to national discretion, is a major step towards much greater consistency in financial regulation. Even if some minor differences of implementation remain within the standards, this is far better than the previous situation where there were differences in the existence of standards. Furthermore, the Basel Committee, along with other standard-setters, has embarked on a major new initiative to monitor the implementation of its standards. This has already had a positive effect of narrowing differences, and provides much greater visibility of the nature and materiality of national differences that remain. Whether you adopt a glass half full or half empty perspective, at the very least the glass is now much more transparent.

If we are truly concerned about levelling the playing field internationally, then we cannot focus only on the shortcomings in the national implementation process. Another, potentially bigger problem - at least if we measure the impact of differences on the reported capital ratios of banks - may exist in the differences that arise from the use of internal models within the regulatory framework. Therefore all of us - regulators and industry - need to continue to work hard to ensure that the consistency we are all seeking is achieved, both in standards and in practice.

Thank you.