1

15 July 2015

Important to keep a risk-based approach for capital requirements Summary 

The undersigned associations are deeply concerned by the proposals from the Basel Committee on a revised standardised approach (SA) for credit risk and new capital floors based on the standardised approaches. We strongly recommend that a riskbased approach founded on internal models also in the future will be the basis for determining capital requirements for large banks.



Risk sensitivity is essential and internal models remain imperative in building awareness and understanding of risks both within banks and among supervisors. Internal models are also, given the availability of sufficient data, the best measure for reflecting the true risk of a bank’s portfolio of exposures, and therefore allocating adequate levels of capital against these risks.



We believe that reforming the internal ratings-based (IRB) approach is a far more fruitful method to achieve the objectives of the Basel Committee than the proposed new capital floors. The European Banking Authority has suggested a number of measures to address potential shortcomings in internal models and to make sure the internal models are implemented in a coherent way across jurisdictions and banks within Europe. We are more than willing to co-operate with the supervisors to further strengthen the IRB framework. In our view, this is the right way forward towards the goals of high confidence and increased comparability that the Basel Committee wants to achieve.



Looking at the proposal for a new standardised approach for credit risk as a standalone proposal (i.e. not as a reference point for disclosure and new capital floors) we can’t see any urgent need for change. Most European banks that today use the standardised approach are small or medium-sized and focus on retail lending. For these banks the proposal will add a lot of complexity and compliance costs, without necessarily strengthening the risk sensitivity or the measurement and control of risk.



If a revised standardised approach for credit risk is to be introduced we believe that it is essential to leave room for national flexibility and calibration, within a globally controlled framework. We strongly oppose a one-size-fits-all model that uses the

2

same risk drivers and the same calibration for banks all over the world. For instance, creditor protection is extremely strong in the Nordic countries and the Netherlands and foreclosure processes are efficient which is reflected in very low losses on mortgage lending. 

The proposed new framework from the Basel Committee will have large consequences not only for banks but also for the customers of banks and for the society as a whole. Therefore we will underline the importance of taking into consideration the total effects of potential changes, and the interplay with other ongoing risk weight-based regulatory proposals (particularly TLAC framework), in future work.

The proposed capital floors will be a game changer In our view, banks that use the internal ratings-based (IRB) approach generally have better knowledge of and control over their risk. The same goes for the supervisors who are required to scrutinize and approve internal models, provided they fulfil regulatory and supervisory requirements. This knowledge and control enhances the stability of the financial system. Therefore, the regulatory framework should continue to provide incentives to develop and use internal, risk sensitive methods to calculate capital requirements. The Basel Committee´s proposal of new capital floors, based on standardised approaches, will hit high quality IRB portfolios disproportionally. For several banks and many credit portfolios the proposed floors are likely to become the binding capital requirement, given a calibration of the floor in line with the current transitional Basel 1 floor. Hence, the incentives to develop and use internal methods for modelling and evaluating the risk of these portfolios would vanish. Potential consequences for banks and their customers Our estimates suggest that the proposal will dramatically increase the risk exposure amount for many Nordic and Dutch banks. This is in contrast to outcome of stress test performed by authorities where Nordic and Dutch banks seemed to be well capitalised. A removal of risk-based capital requirements will create incentives for the banks operating in low risk environments, such as in the Nordic countries and the Netherlands, to change their current behaviour. In particular the banks will get strong incentives to increase margins on their, from a credit risk perspective, best exposures and to securitise such exposures. This in turn is likely to result in smaller balance sheets in banks and more credit intermediation taking place outside the regulated banking sector, which would not be conductive for enhancing growth and employment in the economy. A higher portion of the society’s credit risk exposure being held by non-banks is likely to make the credit intermediation process less resilient in turbulent times and more procyclical, since non-banks may not have the same commitment to provide financing as the regulated banks have.

3

Against this background, it is not reasonable that the Basel Committee agrees on regulatory changes that will have a large impact on both banks and their customers in many jurisdictions. It is important to have a very thorough analysis and public discussion, in which national authorities are involved, not only regarding risk weights and capital requirements but also on how established business models for banks may change and how the customers will be affected. One important conclusion is that the timeline of the process has to be extended significantly to allow a wider discussion about the proposals merits and shortcomings. Moreover, a thorough assessment of the impact not only for banks but for the society as a whole is needed. Finally, interplay with other ongoing risk weight-based regulatory proposals (particularly TLAC framework) should be evaluated. Capital floors will introduce a new, unnecessary layer of complexity The large number of different capital requirements in Basel 3 and CRR/CRD4 give regulators sufficient tools to adequately supervise banks and a lot of possibilities to impose the level of capital that they believe is necessary, both on individual banks and on the banking sector as a whole. Moreover, Pillar 2 is intended to ensure that banks have appropriate risk controls and adequate capital with respect to all risks faced by the institution and not just the risks accounted for under Pillar 1. An assessment of model risk is included in this process. To introduce an additional layer of complexity in the form of a new standardised approach for credit risk and new capital floors is in our view neither necessary nor useful. On the contrary, we believe that new floors can be counterproductive to the purpose described by the Basel Committee, especially if they are introduced before the other parts of the capital framework have been implemented and the effects have been analysed. Since the Basel proposal in itself have very significant consequences both for SA banks and IRB banks, and for their customers, we can see a large risk for unintended consequences if the Basel Committee continues with the proposal before the total effects of new regulations are well understood and analysed. Important to keep a risk-based approach for capital requirements Against this background we believe that reforming the internal ratings-based approach, and maintain a risk-based and internal models based approach, is a far more fruitful method to achieve the objectives of the Basel Committee than the proposed new capital floors. The European Banking Authority has suggested a number of measures to address potential shortcomings in internal models and to make sure that the internal risk-based approaches are implemented in a similar way across jurisdictions and banks within Europe. In our view, this is a much more useful and less destructive way of achieving the goals of high confidence and increased comparability that the Basel Committee wants to achieve through the capital floors. Therefore we are more than willing to put effort in further strengthening the IRB framework and co-operate with the authorities in this respect.

4

No urgent need to revise the standardised approach for credit risk Furthermore, we cannot see any urgent need to change the current standardised approach. Most banks − in the Nordic countries, the Netherlands and in EU as whole − that use the standardised approach are small or medium-sized and focus on retail lending. For these banks the proposal will add a lot of complexity and compliance costs, without necessarily strengthening the risk sensitivity or the measurement and control of risk. Significant banks with respect to size, internal organisation as well as the nature, scale and complexity of activities are according to EU legislation to be encouraged to implement risk sensitive capital requirements. SA is to be used primarily by smaller, less complex banks and the approach is therefore to be maintained sufficiently simple and thus by necessity less risk sensitive. According to our view developing a standardised approach for credit risks with the purpose to use it as a reference point for risk disclosure and risk weight floors will, as long as the Basel Committee holds on to the current “one size fits all” approach, be an impossible task. A minimum requirement on such a reference point model is that there is a world-wide unanimous correspondence between the risk weights provided by the model and the actual risk of loss. It is our view that this will not be possible to achieve with the kind of simple models the Basel Committee proposes. In the figure below the development of non-performing mortgage loans during 2001-2010 in a sample of important jurisdictions are shown. It clearly illustrates the challenges associated with finding a simple world-wide uniform relationship according to which one or two common risk drivers, e.g. LTV and DSC, are sufficient to describe the (expected and unexpected) level of loss in a bank’s mortgage loan portfolio.

Source: IMF, Global Financial Stability Report, April 2011

5

If a revised standardised approach for credit risk is to be introduced we believe that it is essential to leave room for national flexibility and calibration, within a global controlled framework. We strongly oppose a one-size-fits-all model that uses the same risk drivers and the same calibration for banks all over the world. For instance, creditor protection is extremely strong in the Nordic countries and the Netherlands and foreclosure processes are efficient which is reflected in very low losses on mortgage lending.

Johan Hansing Swedish Bankers’ Association

Otto ter Haar Dutch Banking Association

Louise Caroline Mogensen Danish Bankers’ Association

Piia-Noora Kauppi Federation of Finnish Financial Services

For questions, please contact: Johan Hansing, Swedish Bankers’ Association, [email protected], Michael Friis, Danish Bankers’ Association, [email protected], Reima Letto, Federation of Finnish Financial Services, [email protected] Otto ter Haar, Dutch Banking Association, [email protected]