Stefan Ingves: The international regulatory agenda - necessary but not enough

Speech by Mr Stefan Ingves, Governor of the Sveriges Riksbank and Chairman of the Basel Committee on Banking Supervision, at the Affärs Världen Bank and Finans Outlook, Berns, Stockholm, 19 March 2014.

Regulations and regulatory frameworks for banks have been the subject of much discussion in recent years. As several of the regulatory frameworks drawn up after the crisis are now being introduced at the national level these discussions have also perhaps waxed rather than waned recently. New and amended frameworks are also the theme here today. The Basel Committee on Banking Supervision is, as many of you know, an international committee that produces global standards and regulations for banks. This summer, I will have been chairman of the Committee for three years. Perhaps the most important task facing the Committee when I took up the post of chairman was to finalise the Basel III regulatory framework, that is that part of the international reform agenda that followed in the wake of the financial crisis and that most closely relates to banks.

Today, I would like to tell you about the work of the Basel Committee. Nowadays, most regulatory frameworks are designed and drawn up internationally and my idea is to try to give you an insight into the international regulatory work. I would also like to present my view of what the international regulations entail for the Swedish financial system.

There are above all three specific points I want to make.

The first is that the regulatory frameworks that are now being finalised and that will be introduced in Sweden and globally will benefit everyone. Once they are introduced they will increase the safety margins in the global financial system. This will strengthen financial stability in Sweden and abroad.

My second point is that many of the international reforms, and particularly Basel III, are aimed at reducing just those risks and problems that Swedish authorities have identified in Sweden and in Swedish banks. They will thus directly contribute to increasing resilience in the Swedish banking system.

The third point I would like to emphasise is that the Basel regulations are minimum requirements. They will not be able to manage all of the risks irrespective of the country concerned. In my opinion, there are specific circumstances in Sweden, special structural vulnerabilities, that more than justify continuing to set stricter requirements for the Swedish banks.

New regulations will strengthen the banks and reduce the risk of crises

The origins of Basel III lie in the latest financial crisis, which revealed several weaknesses in the global financial system. Above all, it became clear that the banks had too little capital in relation to their risks and that the liquidity risk at the banks was too high. Moreover, the collapse of the investment bank Lehman Brothers demonstrated how difficult it is to manage very large institutions that experiences problems and to avoid the problems spreading to other banks and having negative repercussions on the financial markets.

The intention is that Basel III should capture the risks in the banking system better than the previous regulatory frameworks and thereby reduce the likelihood of new financial crises. Basel III is thus a reaction to the fact that previous regulations were inadequate. The Basel III regulatory framework was published for the first time in its entirety in December 2010. Since then, the Basel Committee has worked to finalise the regulations. Beginning in 2013, Basel III will gradually replace earlier Basel regulatory frameworks. In Sweden, Basel III will be incorporated through EU legislation.

More capital and larger liquidity buffers will increase resilience

Put simply, Basel III is founded on four measures that act as instruments for assessing various risks in the banks. On the basis of these measures, the framework then stipulates certain minimum requirements, for example how much capital a bank must hold.

The first measure is the risk-based capital measure, which is what we normally refer to when we speak about capital requirements. The risk-based capital measure was already a part of previous regulatory frameworks, but in Basel III it has been made much stricter. In addition to this there are three new measures: a complementary capital measure, the leverage ratio, and two liquidity measures. The latter are a measure of short-term liquidity - the Liquidity Coverage Ratio (LCR) - and a structural measure of liquidity risk, the Net Stable Funding Ratio (NSFR).

Today, I intend to focus on the leverage ratio and the NSFR. The reason for this is that the Basel Committee has made considerable progress on these measures this year. However, before I do this I would like to briefly describe the risk-based capital requirement and the LCR, as when I speak about Sweden and Swedish conditions later I will cover all of the measures in Basel III.

Let's begin with the risk-based capital measure, which measures a bank's capital in relation to its risk-weighted assets. This is the part of Basel III that was finalised first and it is already being applied in many parts of the world. The idea is that a bank should have sufficient capital to cover unexpected losses. Compared with previous Basel regulatory frameworks, Basel III entails higher requirements for how much common equity tier 1 capital the banks must have and stricter and more uniform regulations on what may be counted as Common Equity Tier 1 capital that is capital of the highest quality. Basel III also introduces two kinds of capital buffer as a further element of the risk-based capital requirement that could be said to lie "on top of" the minimum requirements. The buffers are intended to provide safety margins before the banks begin to use their minimum capital holdings. In the case of systemically-important banks there is also an extra capital requirement over and above the minimum requirements. How large this supplement is depends on how systemically important the bank is. For global systemically-important banks the extra requirement will begin to apply from next year.

For the last 12 months there has also been an agreement in the Basel Committee on the short-term liquidity measure, the LCR. In brief, the LCR means that the banks must have a large enough buffer of high liquid assets to be able to cover their cash outflows during a 30-day long stressed scenario. The aim is to strengthen the banks' resilience to short-term liquidity stress. The LCR will come into force no later than 1 January 2015, but some countries are already applying it. In Sweden, a stricter version of the LCR has been applied since 2013.

Leverage ratio complements the risk-based capital requirement

As I mentioned earlier, the Basel Committee has made a lot of progress on the leverage ratio this year. In January the Committee agreed on a common method for calculating the ratio. The leverage ratio can be described as measure of how much of a bank's assets the bank funds using its own capital. Like the risk-based capital measure, the leverage ratio aims to increase the ability of the banks to cover losses. However, unlike the risk-based capital measure, the leverage ratio does not take differences in risk between a bank's different assets into account.

The Basel Committee has been criticised by those who claim that introducing the leverage ratio creates a regulation that favours risk taking, as the measure deals with the banks in the same way irrespective of the risks they take. However, the leverage ratio is intended to complement the risk-based capital measure and to act as extra protection to ensure that the level of capital at the banks does not fall too low. It can, for example, limit how much the capital can fall in relation to the bank's total assets if the models the banks use to calculate risk weights fail to capture risk to a sufficient extent and the risk weights therefore become too low.

As yet, the Basel Committee's agreement on the leverage ratio does not specify any quantitative requirements. However, from the beginning of next year, large banks all over the world will need to publish their leverage ratios. The idea is that the Basel Committee will monitor developments when the banks begin to report their leverage ratios with a view to migrating to a minimum requirement from 2018. However, in most cases the risk-based capital requirement will continue to be the binding capital requirement, and not the leverage ratio. However, the opposite may of course apply in some cases. This is also reasonable because, when all is said and done, a requirement that can never be considered binding is not particularly meaningful.

In my opinion, the leverage ratio is a welcome complement to the risk-based capital measure. The Basel Committee's agreement means that we now have a common method for calculating the leverage ratio, which we did not have before. This is good. The agreement on the leverage ratio is therefore a major step forward.

The regulations for the structural liquidity measure, the NSFR, complete Basel III

The fourth measure in Basel III is thus the structural liquidity measure, the NSFR. The aim of the NSFR is to ensure that the banks fund their operations with sufficient long-term and stable funding so as to reduce the risk of being hit by funding and liquidity problems. The NSFR thus constitutes an important complement to the short-term LCR measure that aims to protect the banks when, in some sense, liquidity stress is already a fact.

The NSFR is the component of the Basel III framework that has not yet been finalised. Finalising the regulations for the NSFR is one of the Committee's most important priorities for 2014. This will in principle mark the completion of the work with the entire Basel III regulatory framework. As part of this work, the Committee recently published a revised proposal for the NSFR for stakeholders outside the Committee to comment on. In connection with this the Committee reiterated that it intends to introduce the NSFR as a quantitative requirement from 1 January 2018.

The changes made to the NSFR compared with the original proposal primarily aim to ensure that the measure better captures the funding structures that proved to be weak during the crisis. For example, the requirements have been tightened for operations that are typically associated with investment banks. Another change is that also funding with a maturity of between six and 12 months may to a certain extent be included as part of a bank's stable funding. However, the aim of the NSFR is still to ensure more stable funding at the banks.

Other new regulations will also increase the safety margins

In addition to the work with Basel III, the Basel Committee is devoting a lot of resources to completing the review of a number of other regulatory frameworks. These frameworks are not usually seen as a part of Basel III, but this work also began as a result of the financial crisis. For various reasons, these frameworks will perhaps have less direct impact in Sweden than Basel III. But they are important from the global point of view and will indirectly benefit Sweden by increasing the safety margins in the global financial system.

For example, one element of this work concerns an updated framework for the capital requirements of the banks' trading books. A similar review is also underway for the capital adequacy framework for securitised loans and other assets. Many major international banks have large trading books and considerable exposures to securitised assets. It is important to introduce regulations for risks related to activities of this type so that they can be limited. In a globalised world this is also important to Sweden. In addition, the Basel Committee is also working with a regulatory framework to limit large exposures and a number of frameworks relating to the derivatives market. All of these are included in the reform package that the Basel Committee hopes to be able to finalise this year.

New focus: application of the regulatory frameworks and the calculation of risk weights

By now it is probably obvious to most of you that the Basel Committee and other standard setters have had a lot to do: the list of regulations is long. However, important as it is to design effective regulatory frameworks, it is just as important to ensure that they introduced in a consistent way around the globe. Only then will they be effective and credible. During my chairmanship, I have therefore considered it to be an important task for the Basel Committee to extend its efforts to monitor how various frameworks are implemented.

Today, the Basel Committee conducts this work at several levels. For example, the Committee is currently assessing how the capital adequacy regulations in Basel III have been incorporated into national regulatory frameworks. Six assessments have been carried out so far and three are in progress, including the assessment of the EU. This assessment will eventually be extended to also cover other parts of the Basel III framework. However, it is already clear that this is an effective way to ensure that the frameworks agreed on are actually put into place and to ensure that the regulations are used as intended.

The Basel Committee is also reviewing how the banks calculate their risk weights. Put simply, the risk weights determine how much capital the banks need to hold for a certain asset, for example a loan or a security. So far, the Committee's work has resulted in three studies that have compared how international banks calculate their risk weights for different portfolios. The latest study examined, among other things, the banks' risk weights for lending to companies, while the earlier studies focused on risk weights for the banks' trading operations. However, all three studies have revealed that the risk weights the banks calculate themselves using their own models vary to a relatively great extent without this always being justified by differences in the risk associated with the assets. This means that it is not possible to effectively compare the banks' risk weights. There is a risk that this could have a negative impact on confidence in the risk-based capital regulations.

The banks' risk weights must be credible

I would like to focus on risk weights for a moment as we are currently discussing these here in Sweden too, although the discussion here differs somewhat from the discussion abroad, for example in that it primarily concerns mortgages. I have long been clear that the risk weights for mortgages that the banks calculate using their own models are often too low. However, the causes of both the Swedish and the international problems can be traced to the same source. Internationally too, it has been concluded that the banks' own models are inadequate for ensuring that risk weights always reflect the risks in a good and reliable way. Discussions are therefore underway as to whether the regulatory frameworks, as well as the oversight and approval of the banks' internal models, must be reviewed so that the models cannot become tools for an unwarranted lowering of the capital requirements.

In my opinion, finding ways of ensuring that the risk-weight calculations are credible and that the risk weights of different banks can be compared is one of the most important tasks of the Basel Committee going forward. A number of measures are being considered and discussed; we will gradually begin to see the fruits of this work. This relates, for example, to reviewing the banks' choice of models for calculating risk weights and the degree of freedom for the banks to make their own assumptions in the models, as well as setting stricter requirements to ensure that the data used as input in the models is of sufficiently high quality. Another approach involves discussing and developing proposals for regulatory floors of different kinds, for example risk-weight floors like the one we have for mortgages in Sweden today or floors that cover the total capital requirement that is also in place today. Finally, it is a question of tightening up supervision and improving transparency concerning risk weights. It is probable that the assessment of the banks risk weights will be broadened and eventually become a permanent part of the undertakings of the Basel Committee.

As in the case of the leverage ratio, there are those that claim that these measures will reduce the banks' incentives to effectively manage their own risks. However, in my opinion it is question of avoiding a regulatory framework that gives the banks incentives to use their own models with the aim of reducing their risk weights and their capital requirements more than can be justified by the risks in their operations. It is important to find solutions to this problem as it will otherwise be difficult to restore the credibility of the frameworks and thus of the banks.

Basel III is also relevant to the Swedish banking system

Once Basel III is in place around the globe, the safety margins in the global financial system will be greater than previously. This will benefit everyone, even Sweden, partly by reducing the risk of global crises and partly by reducing the risk of crises spreading to Sweden and other countries through the links that exist on the financial markets.

At the same time, the problems and risks that the international regulatory initiatives aim to deal with correspond to those that Swedish authorities have identified in Sweden. The Basel III framework will therefore also have a direct positive impact on financial stability in Sweden. Let me give you a few concrete examples:

  • Stricter risk-based capital requirements will increase the ability of the Swedish banks to cover unexpected losses, in both their Swedish and foreign operations, thereby increasing the banks' resilience.
  • The leverage ratio will provide a complementary picture to the risk-based capital requirement and means that authorities and investors will be better able to assess a bank's capital situation. Having a quantitative requirement for the leverage ratio sets a limit for a how low the Swedish banks' capital can fall in relation to their total assets, irrespective of how risky their operations are considered to be. A leverage ratio requirement of, let us say, 3 per cent corresponds to a risk weight for mortgages of approximately 22 per cent for a major Swedish bank with a 12 per cent CET 1 capital requirement.1
  • Another example is the NSFR. The Riksbank has long pointed out that the Swedish banks should reduce their funding risks by extending the maturity of their funding so that it corresponds more closely to the maturity of the banks' assets. The aim of the NSFR is to get the banks to do just this.
  • Finally, we have the so-called countercyclical buffer, which is one of the new buffers introduced through the capital adequacy regulations in Basel III. This is intended to ensure that the banks build up extra shock absorbers in times when credit growth is particularly high so that they really take into account the risks that such an expansion entails for the financial system as a whole. This may also help to slow down an excessive growth in credit. Once this framework is in place it is in my opinion reasonable, given the course developments have taken in Sweden, that the countercyclical buffer is implemented and activated at a high level.

But the structure of the Swedish banking system calls for greater safety margins

Although the new regulatory framework will strengthen resilience in several ways there are structural vulnerabilities in the Swedish banking system that mean that the safety margins provided by the Basel III framework are not enough in all respects. As every country's financial system and banking system are unique and have their own special characteristics and vulnerabilities, the basis of the Basel frameworks is that the regulations are minimum requirements. Countries that believe it is necessary to set stricter requirements can and should do so.

In my opinion, the size and high degree of concentration of the Swedish banking sector, as well as the banks' extensive dependence on both foreign and domestic market funding, create vulnerabilities that justify complementing the internationally-agreed frameworks. The high and growing indebtedness of the Swedish households poses further risks to the financial system. All-in-all, this justifies Sweden taking action sooner and having stricter regulations in certain respects.

On the basis of several of these factors, the Riksbank has recommended, together with Finansinspektionen and the Ministry of Finance, that from 1 January next year the major Swedish banks should have CET 1 capital ratios of at least 12 per cent. This is significantly higher than the requirement in the Basel III framework. Several other countries with large banking systems have for similar reasons introduced or are discussing higher capital requirements for their banks. Two closely-related examples are Switzerland and the United Kingdom.

Another area where we need stricter requirements for Swedish banks is the area of short-term liquidity risk. At present, the largest Swedish institutions are required to have a liquidity coverage ratio of at least 100 per cent. The LCR requirement applies in total for all currencies together and for operations in euros and US dollars separately. However, separate quantitative requirements for individual currencies are not included as a quantitative minimum requirement in the Basel III framework. In my view, however, it is very important that the banks maintain a high level of resilience to liquidity stress by holding liquidity buffers in the currencies in which they have their largest cash flows. An LCR requirement per currency ensures that this is the case in euros and US dollars.

It is also reasonable that the banks report their LCRs in Swedish kronor to provide a more complete picture of their liquidity risks. The Riksbank therefore recommends the major banks to publish their LCRs in Swedish kronor.2 Unfortunately, only one bank, Swedbank, meets this recommendation today. There are good reasons for introducing some form of quantitative requirement for LCRs in Swedish kronor. The banks have low liquidity buffers in Swedish kronor at present, which means that they would have to rely on foreign markets and swap markets if they were to need liquidity in Swedish kronor. The Riksbank is of course always able to lend Swedish kronor, but in my view it is not reasonable that the banks can take liquidity risks on the assumption that the Riksbank will always be at hand if things do not go according to plan. The Riksbank should, potentially, be the lender of last resort, not the lender of first resort. Liquidity regulations in Swedish kronor would also give the banks incentives to extend the maturity of their funding. This is because, all else being equal, a bank that extends the maturity of its funding will reduce its short-term cash outflows and will thus have a higher LCR.

We can also view risk-weight floors for mortgages as a complement to the existing regulatory frameworks. The leverage ratio deals to some extent with the problem that the risk weights perhaps do not always reflect the actual risk in the banks' portfolios, but it does not do so precisely enough. International measures may eventually be proposed, but until then and as long as household indebtedness constitutes a significant risk we need a sufficiently high risk-weight floor for Swedish mortgages. At present, we have a risk-weight floor of 15 per cent in Sweden. As long as household indebtedness increases, there are good reasons for raising this floor to at least 25 per cent.

Finally, in the case of the leverage ratio and the NSFR it is too early to say whether the minimum requirements in Basel III are enough for Sweden or not. As yet, the Basel Committee has not decided on a quantitative requirement for the leverage ratio and it is therefore not possible to determine today how a Swedish requirement should relate to the international requirement. With regard to the NSFR, the Riksbank has recommended the major Swedish banks to extend the maturity of their funding in order to be able to reach the NSFR once it is in place. The revised proposal for the NSFR means that the reference point for this recommendation has moved somewhat. However, irrespective of the technical design of the NSFR, the large Swedish banks need to extend the maturity of their funding. What the appropriate levels for the Swedish banks' leverage ratios and NSFRs are, and thus whether the Riksbank needs to work for the introduction of higher levels than those that will apply internationally, are questions that we will need to consider carefully and return to in the period ahead.

The Basel regulations provide a good foundation, but more may be needed

I hope that I have now succeeded in giving you a better insight into the work of the Basel Committee and my view of its importance to Sweden.

The Committee has recently taken important steps towards finalising Basel III and hopefully a complete regulatory framework will be available at the end of 2014. I also hope that the work that has begun on a number of other important frameworks in the wake of the financial crisis will be finalised this year. This will entail a change in the focus of the Committee's work. Going forward, the Committee will place more emphasis on monitoring and assessing the application of the frameworks to ensure that the regulations are effective.

The international regulatory work is extensive and will affect us directly when the regulations are introduced here in Sweden. Having a global regulatory framework that increases the safety margins in the global financial system will benefit everyone and is therefore important. It is also important to remember that the international discussion on regulations and regulatory frameworks has a lot in common with the discussion underway here in Sweden. In the main, the measures proposed in Basel III aim to reduce risks that also exist in our Swedish banks and that we have been pointing out for a long time. The international regulatory work therefore serves Swedish interests - in other words, it is not a question of introducing regulations that have mainly been drawn up to manage problems that only concern other countries. Moreover, discussions are underway in international forums that are similar to those underway here in Sweden on many other policy issues, but perhaps not as much progress has been made abroad as in Sweden on considering what measures can be introduced. One example of this is the discussion of floors for risk weights.

Finally: as things look today there are, in my opinion, specific circumstances in Sweden, especially structural vulnerabilities, that justify continuing to complement the internationally-agreed requirements with stricter requirements in certain areas. One of the greatest risks I see as far as Sweden is concerned is the mortgage market. The measures needed to come to terms with this have not been in focus here today, but this is certainly an issue I will return to. Today, I have instead talked about international regulations and about Sweden's need to go further in some respects. Those I have mentioned today are higher capital requirements, LCRs per currency, risk-weight floors for mortgages and the countercyclical capital buffer. In the case of the leverage ratio and the NSFR, we need to examine more closely whether the Riksbank should work for stricter requirements in Sweden than the upcoming international minimum requirements.


1 Having 12 per cent CET 1 capital here entails, according to the regulations, a primary capital requirement of at least 13.5 per cent of risk-weighted assets. As 13.5 per cent of 22 per cent equals 3 per cent, the leverage ratio requirement, put simply, corresponds to a risk weight of 22 per cent for mortgages. The calculation is based on a leverage ratio defined with primary capital in the numerator.

2 Financial Stability Report, November 2013. Sveriges Riksbank.