George Soros did not mince his words. "Basel II, which delegated authority for calculating risk to the financial institutions themselves, was an aberration and has to be abandoned," he told a congressional committee meeting in Washington, DC in November 2008. Strong words indeed, but were they fair? Given the years of hard work, the endless meetings, consultations and studies, not to mention the astronomical costs of implementation, calls for abandonment will no doubt annoy those involved in the Basel II project - some of whom will have committed a decade of their lives to it.
But no-one could observe the events of the past few months and not wonder how relevant the framework is to today's banking market. In particular, the decision by governments across the globe to inject capital into banks arguably calls into question the fundamentals of the Basel Capital Accord. In the UK, for instance, the Financial Services Authority (FSA), the Bank of England and the Treasury decided Royal Bank of Scotland (RBS) had to recapitalise to the tune of £20 billion - an amount that should push its tier-one capital ratio above 9%. But how was that figure reached, and in what way, if at all, did it relate to the internal models built by RBS to comply with Basel II? Only sketchy details have been made available on how the £20 billion figure was calculated, but outside observers note the intervention signals an abandonment of at least some parts of the Accord. As one risk manager, who was responsible for Basel II implementation at a UK bank, puts it: "The FSA appears to be walking away from Pillar II."
Pillar II covers the supervisory review process. It sets out specific oversight responsibilities for the board and senior management, and stresses the importance of internal control rather than external control by a regulator. The idea is that management conducts an internal assessment and sets targets for capital that are commensurate with the bank's particular risk profile. The supervisor then assesses this. So, it is up to the bank to set its own capital levels and then argue the case for keeping them. In other words, it is very deliberately not prescriptive.
This process appears to have been more or less ditched. Speaking at the annual British Bankers' Association banking supervision conference on October 28, Thomas Huertas, director of the banking sector at the FSA, described in broad terms how the capital levels for each UK institution was determined. Three factors were taken into account: ensuring the aggregate amount of capital held by financial institutions restored confidence in the banking system as a whole; making sure the amount of capital for each bank would sustain confidence in that institution; and ensuring each individual institution would have sufficient core tier-one capital, even after absorbing losses that might ensue from a severe recession.
The FSA used stress tests based on some standard assumptions, but with weightings tailored to specific institutions, Huertas explained. "The approach included a confidence premium and it should not be presumed this represents the FSA's view of the right long-term capital framework for deposit-taking institutions," he added. The FSA is due to publish a discussion paper, which will cover its views on appropriate long-term capital levels, in the first quarter of 2009.
In coming up with the current number, Basel experts believe the regulator used concepts and models developed under Pillar I (which covers minimum capital requirements). However, some speculate the supervisor started with a core tier-one ratio of around 9% in mind, and then applied stress tests severe enough to generate a capital number that matched this target. "People have been describing it as reverse stress testing," remarks Patrick Fell, a London-based Basel II consultant at PricewaterhouseCoopers (PwC). "My belief is they would have done the same kinds of assessments, calculations and stress tests that the banks do, come to an assessment capital number, got that and then said you need some more capital on top of it."
Regulators had to come up with new capital levels at a time of unprecedented stress, when confidence in the banking system had evaporated and market participants were wondering which financial institution would be next to collapse. Nonetheless, some feel the apparent scrapping of Pillar II in favour of more prescriptive minimum capital requirements is a return to the one-size-fits-all approach Basel II was meant to avoid.
"Pillar I is derived from a mathematical model that cannot be a good fit to all firms' balance sheets, especially if the assumptions are stressed," says Bruce Porteous, head of UK risk capital development at Standard Life in Edinburgh. "There appears to be a great deal of emphasis on Pillar I ratios now, whereas Pillar II should be a better measure of risk. A firm's own risk self-assessment, as discussed and agreed between the firm and supervisors, should be a better measure of risk because it is bespoke and involves good, detailed and wide-ranging conversations between firms and supervisors. Having spent a lot of time and money implementing Basel II, it would be a shame if the industry does not use and develop its risk-based tools in practice."
The question is whether regulators have completely abandoned the self-assessment principles described under Pillar II or whether they will be reapplied in future. While Huertas hints the confidence premium currently applied to capital levels will at some stage be dropped, there is no suggestion banks will once again be allowed to use internal risk models to assess and set their own capital requirements. In fact, Huertas explicitly states this foundation stone of the Basel II framework has not worked.
"The broad assumption underlying the Basel Capital Accord - that regulators around the world could rely on firms' own risk models as the basis for capital requirements - has not turned out to be correct, at least for the trading book," he declared.
As such, some market participants feel Basel II, in its current form, is dead. With regulators seemingly moving towards a more prescriptive and rules-based approach, the heart has effectively been ripped out of Basel II, some argue. "I think it's all over. Basel II was a form of self-regulation from the banking sector, and clearly there is going to at least be a very serious attempt to depart from that," says Philippe Carrel, a Geneva-based compliance consultant at Thomson Reuters Risk Management.
Not everyone is so pessimistic. Some feel much of the current framework should be kept and argue Basel II was not a cause of the crisis. "It's not a matter of throwing it out and starting again; it's a matter of making the framework work in the current environment. Basel II, in its current form, is neither a cause of the current environment nor a major cure for it - it is just part of the operational way in which you assess capital for banks," says Andrew Cross, who has global responsibility for risk measurement and reporting, including counterparty risk, at Credit Suisse.
Others point out Basel II is sufficiently flexible to allow large parts of it to be adapted to the current environment. At the very least, the push to improve risk management models and processes within banks, encouraged by Basel II, should continue.
"I don't think Basel II is bust at all. It was never written in stone and there is much more you can do with the existing toolkit," says John Tattersall, a London-based PwC consultant. "Basel II was always designed to improve the quality of risk management and to give an incentive for banks to manage their risks better by reducing the capital they require. What's happened now is that capital reductions are out of the window and we're all going for capital increases. But in the process, there has been a significant improvement in the quality of risk management data and processes."
Despite the push towards more prescriptive measures by regulators, the Basel Committee has talked only of enhancing the Accord, as opposed to replacing it. "The Basel Committee's work programme is well advanced and provides practical responses to the financial stability concerns raised by policy-makers related to the banking sector," said Nout Wellink, president of the Netherlands Bank and chairman of the Basel Committee, speaking at a meeting in Beijing in November.
The Basel Committee is currently spearheading a number of initiatives that aim to work the lessons learned into the Accord. A big part of this is focused on improving the treatment of risks in the trading book, seen as a major failure during the financial crisis. Specifically, the majority of losses in bank trading books did not arise from actual defaults, but rather the widening of credit spreads, credit migrations and a drying up of liquidity - factors not covered under the Basel II framework.
On July 22, the Basel Committee released revised proposals for an incremental risk charge (IRC) for trading book positions, replacing the previous incremental default risk charge. The new rules embrace all positions except those whose valuations depend solely on commodity prices, foreign exchange rates or the term structure of default-free interest rates - in other words, it will include debt securities, equities, securitisations, collateralised debt obligations and other structured credit products. The IRC will also cover a much broader range of risks beyond default, including credit rating migration, spread widening and equity prices. Banks will be allowed to develop their own models, but the Basel Committee stipulates the model must measure all losses at the 99.9% confidence interval over a capital horizon of one year, taking into account the liquidity horizons applicable to individual trading positions or sets of positions (Risk September 2008, pages 26-281).
"The Basel paper proposing the IRC was a very clear recognition that the approach to credit risk in the trading book was not adequate in Basel II, whereas the approach to credit risk in the banking book probably was adequate," says Tattersall.
The Basel Committee also wants to address the issue of pro-cyclicality. A major criticism of Basel II has been that it allows banks to hold less capital when the economy is rosy, yet requires them to hold more during a downturn - potentially forcing banks to rein in lending and exacerbating the downward spiral.
It is now widely accepted the Accord needs to be modified to make it less pro-cyclical and to force banks to build strong capital buffers in good times that can be drawn upon in periods of stress. "We will explore promoting strong capital buffers above the minimum levels and how those can be used during a downturn to dampen shocks and encourage continued lending," Wellink said. "The committee is also assessing ways to strengthen prudential filters, promote through-the-cycle provisioning and contribute to efforts to strengthen accounting standards for financial instruments."
Liquidity risk management has been another area of focus. This was seen as another major failure during the crisis: banks presumed they would always be able to raise funding and woefully underestimated the capital they might need to satisfy contingent liabilities.
"If a year ago, a bank had a light report of 30 pages to submit to the FSA, then 20 pages of it would have been dedicated to capital adequacy and you would have been lucky if two or three pages were dedicated to liquidity. That weight of emphasis will have to change," says Selwyn Blair-Ford, a London-based consultant at FRSGlobal, which implements Basel II risk measurement systems.
The Basel Committee responded by publishing revised Principles for Sound Liquidity Risk Management and Supervisions in June 2008 - a document endorsed by global supervisors in September. The 17 principles focus heavily on governance, measurement and management of liquidity risks, public disclosure and the role of supervisors. Among the new proposals is the requirement that banks articulate a liquidity risk tolerance "appropriate for its business strategy and role within the financial system". Senior management should be responsible for establishing the strategy and maintaining a sufficient level of liquidity, with processes to be regularly reviewed and subject to board approval. Exactly how a firm goes about articulating liquidity risk tolerance is open to interpretation (Risk South Africa, Spring 2008, pages 30-332).
More work still needs to be done on the liquidity front. Wellink stresses the Basel Committee needs to "redouble our efforts to develop more consistent benchmarks for sound liquidity at global banks". This includes benchmarks for liquidity cushions, maturity mismatch, funding liquidity diversification and resilience to stress.
Aside from the headline initiatives, the Basel Committee is also reviewing a number of other parts of the framework. For example, it intends to re-examine the role of external ratings and their use in calculating capital for securitisation transactions, as well as assessing risk weights under the standardised approach of Pillar I. It also intends to review the treatment of counterparty credit risk, and it seems likely new leverage ratio benchmarks will be introduced as a way to manage the build-up of leverage at financial institutions.
With so much of the framework set to be changed, and with regulators and banks currently in emergency mode and therefore not following the current guidelines, some would argue the banking industry has already moved beyond Basel II. "I think Basel II will naturally phase out, to be replaced with a new set of regulations and an entirely new approach that, for example, provisions and allocates for worst-case scenarios as opposed to most probable scenarios," says Carrel.
Governments started to consider a new international framework to manage banking risks during the G-20 meeting in Washington, DC in November. The Basel Committee, through its Accord Implementation Group, will hope to play a leading part in future developments. But will it? Could the apparent failure of light-touch regulation - which lies at the core of Basel II - be enough to convince world leaders that what is needed instead is a tough, rules-based system, perhaps policed by a group such as the World Bank or the International Monetary Fund (IMF). Already, UK prime minister Gordon Brown has put forward the idea of giving the IMF some sort of global regulatory role.
In his address to the congressional committee, George Soros suggested replacing Basel II with a Basel III framework that embraces an interventionist regulatory philosophy. Under this new paradigm, regulators would accept responsibility directly for controlling asset bubbles, partly by intervening in a proactive way to adjust capital buffers to "counteract the prevailing mood of the markets". Whether we see the emergence of Basel III, an acceptance of a reworked Basel II, or some completely new regime, is anyone's guess. This will be decided, eventually, by the politicians, who are not due to meet again until April next year. Until then, the Basel Committee will have to fight to keep Basel II relevant.
The week on Risk.net, January 6–12Receive this by email