Share the article
twitter-iconcopy-link-iconprint-icon
share-icon
ArchiveMarch 2 2001

Prudence makes perfect

Share the article
twitter-iconcopy-link-iconprint-icon
share-icon

As the Basel Committee makes its recommendations, Sir Howard Davies examines the ramifications for the overhaul of the 1988 Accord.

The proposals published by the Basel Committee on January 16 represent an important milestone in overhauling the capital standards for banks.

The industry has now been given a comprehensive statement of how supervisors intend to revise the 1988 Accord. The Financial Services Authority (FSA) has been heavily involved in the development of these proposals. So you would expect me to welcome them.

But I do so with a genuine sense of enthusiasm. We now have an opportunity to make a major advance in the area of prudential standards, which should make financial markets both safer and more efficient.

For those who come fresh to this subject, I would acknowledge that there is a “wood and trees” problem. The proposals are lengthy and detailed – an inevitable by-product of the Committee’s decision to break with a one-size-fits-all approach.

Banks will face a menu of options that take account of their differing levels of sophistication. But, despite all this complexity, the essential aims are simple: to devise a set of rules that create a better fit between regulatory and economic capital; and to create incentives for banks to assess their own risks, and need for capital, more accurately and systematically than before.

The new Accord will, it is thought, measure risks better than the old. Banks will have better incentives to engage in prudent lending, conduct proper risk assessment and hedge exposures.

But by introducing options that allow more sophisticated banks to use their own internal risk models, is there a risk of supervisors handing over too much control to banks? And does this mean that the new Accord will result in more volatility in bank capital standards?

The first point to make concerns the balance of responsibility between banks and their supervisors. The new Accord recognises explicitly what everyone already knows – that the primary responsibility for assuring that a bank has adequate capital resources lies with the bank itself and not its supervisors.

The supervisors’ role is to lay down the framework of principles and rules by which the assessment of capital adequacy is carried out, and then to assure themselves that the framework is adhered to – and take action if necessary.

This philosophy is enshrined in the three pillars of the new Accord:

Pillar 1 comprises a set of mandatory capital charges intended to yield a reasonable minimum capital requirement.

Pillar 2 describes the process by which banks, in conjunction with their supervisors, assess the buffer they need over and above the Pillar 1 requirement to support other risks.

Pillar 3 ensures that banks disclose sufficient information to allow the wider market to judge the credibility of their capital plans.

In Pillar 1, the Committee has stopped short of recognising credit or operational risk modelling. Some banks (and software suppliers) may protest that their models measure risks more accurately than ever before. That much is true, up to a point.

Supervisors have been willing to accept that value-at-risk models can play a role in determining capital standards for the trading book. But credit and operational risk modelling are still in their infancy and it remains to be seen how these models stand up if and when they are tested by a recession.

The Committee has, however, kept the door open, recognising that the robustness of such models will improve over time. Nevertheless, we have moved a long way from the current regime. In the more advanced options now being proposed for credit risk, banks will calculate the main drivers of risk (e.g. default probabilities and recovery rates) which are then converted into capital charges. In the more advanced option for operational risk, banks will use data from their own loss history as a driver of their operational risk charge.

So does this undermine the views on the balance of responsibilities? Probably not. First, banks will only be able to move to advanced options for credit and operational risk if they meet a set of challenging standards. Second, supervisors rather than banks will determine the function that maps credit inputs data into capital outcomes.

The same will be true for operational risk. Also, capital charges under the more advanced options will be calibrated so that the total amount of regulatory capital for an “average” bank will be similar to the amount of capital yielded by the standardised approach. There will be naturally a wide range of outcomes around this average, otherwise there would be no risk sensitivity.

Banks with a portfolio heavily weighted towards investment grade assets should expect their Pillar 1 capital charge to fall. Those who have a loan book tilted towards the other end of the spectrum should expect their Pillar 1 charges to go up. A number of commentators have suggested that bank capital could become more volatile as a consequence of the new rules.

Certainly, more sensitive charges for credit risk could result in Pillar 1 capital requirements increasing in downturns as banks take a more pessimistic view of the likelihood of default of various counterparties. Banks’ actual capital requirements will result from a combination of Pillars 1 & 2 and therefore any fluctuations in Pillar 1 capital may be masked under the total figure.

Given the importance of this issue, it should be examined closely during the consultation process. So there are good reasons for believing that we are in sight of an Accord which will combine a similar overall amount of capital in the system as a whole with a more sensible distribution of that capital and a wider application of modern risk management tools.

This ought to be a clear plus, particularly given that since the 1988 Accord there have been plenty of instances of bankers hitting on creative ways to lose money. There will no doubt be further debate during the next few months about the details of the package. But the initial response from UK banks has been quite positive. Indeed, perhaps the most widespread concern we have encountered so far is about implementation, not substance.

Banks in London are concerned that the lengthy European legislative process might delay the implementation of the new rules in Europe, giving American banks, in particular, a competitive advantage in the short term. We understand this concern – indeed, we share it. The current Basel rules are incorporated in European directives. So to allow an orderly move to the new capital regime, European law will need to be amended.

There is of course a significant overlap between the Basel Committee and the EU, which greatly assists the process of parallel development. But Brussels should not simply duplicate Basel. There are some specific European issues that arise from implementing these revisions.

The current directives, for example, apply to all banks and building societies and many investment firms, and not just to the internationally active banks that have been the main focus in Basel. In addition, the EU legislative process quite properly involves scrutiny by the Council of Ministers and the European Parliament as well as by supervisors and the Commission.

In practice, I think we have done what we can to get the EU process moving, without prejudicing the outcome of the consultation period. The Commission has already published its own paper explaining the process. So we are on track to deliver our implementation in 2004, in line with the expected US date.

But it would be fair to say that the European legislative system is not ideally suited to handle this type of rule-making. Financial directives are typically far too detailed, and therefore inflexible, and not easily adaptable to changing circumstances. In the UK, our new legislation gives the FSA the freedom (subject to consultation and due process) to change and update its rules from time to time, without constant recourse to Parliament.

There is no equivalent procedure in the EU. This point is made forcibly by Alexandre Lamfalussy, who has been asked by the European Council to review securities regulation.

He has recommended a new process, with framework directives amended from time to time by a committee of regulators. The same arguments apply in the banking arena. What we need is an arrangement that permits determination of the fixed portion of the legislation by the relevant EU bodies, including the European Parliament, but with the appropriate delegation to technicians.

If every technical clause has to go through a full legislative process, the European financial services industry may be unable to compete effectively with the rest of the world, without any real compensating public accountability gains. The politicians – both ministers and members of the European Parliament – say they understand the problem. But they have not yet come up with the solution.

It is vital that they do so in the next year. The revised Basel Accord will, I am sure, be beneficial for European banks and their customers.

So we must find a way of securing those advantages without unnecessary delay.

Sir Howard Davies is chairman of the Financial Services Authority.

Was this article helpful?

Thank you for your feedback!