The deadline for Basel II is fast approaching. Under the timetable set by the Switzerland-based Basel Committee on Banking Supervision, banks employing the standardised and foundation internal ratings-based (IRB) approaches will kick off under the new regime from January 2007. Those institutions implementing the more complex advanced IRB approach have an extra year, with the start date delayed to January 2008.
In South Africa, however, the supervisors have opted for a single implementation date for all approaches of January 1, 2008. And despite some early challenges in hiring Basel II specialists – hardly a phenomenon limited to South Africa – banks say they are more or less on track. According to a survey conducted by consultants Ernst & Young published in August, budget allocation within South African financial institutions is less of a concern now than it was one year ago. Banks appear to be adequately funding their Basel II operation and ensuring that such efforts are given appropriate importance internally.
South Africa's regulator, the South Africa Reserve Bank (Sarb), has also taken a step forward. It has expanded its Basel II team by migrating the bulk of its staff to Basel II from Basel I supervision. At the same time, discussions are under way through the Accord Implementation Forum (AIF) – a group that consists of representatives from the National Treasury, Sarb, the South African Institute of Chartered Accountants, the Banking Council South Africa and the country's banks – to try to resolve contentious issues arising from the framework, including the economic impact of Basel II on South Africa, the disclosure requirements contained in Pillar III, and how to address the various items of national discretion allowed by the Basel Committee.
However, perhaps the biggest change has been sparked by the takeover of one of South Africa's leading financial institutions, Absa, by Barclays in July. Following the acquisition, the bank is planning to adopt the advanced approach to credit risk rather than the foundation approach. And as a consequence, Absa's main rivals have decided to up their game. This necessarily increases the pressure on the banks in terms of allocating resources, but it also underscores their confidence in tackling the challenges ahead, say analysts.
"Certainly for the five largest banks [Absa, FirstRand, NedCor, Standard Bank and Investec], progress appears to be good, while the small local banks and foreign banks are largely waiting to see what Sarb will do," says Anton de Souza, lead financial services partner at Ernst & Young in South Africa. "To some extent, the long-standing requirement for higher than international standards of capital adequacy in South Africa has been beneficial in making sure banks are well placed for Basel II."
In its annual report for 2004, Sarb states that it should be possible for "every registered bank in South Africa to migrate to the base approaches of Basel II without major disruption or risk to individual banks or the banking system" by 2008, but declines to comment on the readiness of leading banks to adopt the more advanced approaches. It adds that, more generally, the groundwork done during 2004 has laid a solid foundation for the successful implementation of Basel II in South Africa.
However, challenges still remain. "There are still many questions left unanswered about what type of data can be used and how it will be reported to the registrar [regulator]," says de Souza. "To some extent, this is a chicken and egg situation because the registrar is reluctant to implement anything locally that would contradict international standards still to be clarified. Nevertheless, banks in South Africa say that they need more information as soon as possible."
Absa's decision to aim for the advanced IRB approach follows Barclays' acquisition of a majority 53.96% stake for nearly R30 billion ($4.6 billion) in July. "[The advanced approach] will bring benefits to the bank, especially on the wholesale side of operations where we will have an advantage in interbank business," notes André Blaauw, general manager of enterprise-wide risk management at Absa.
One of the main reasons for moving to the advanced IRB approach for credit risk has been the composition of its lending portfolio. "We have large collateral lending and there is no recognition for collateral under the foundation approach," explains Blaauw. "But that means we will have to build new models, which will obviously take time."
Unsurprisingly, Absa's main rivals have indicated that they also intend to move to the advanced IRB approach rather than foundation IRB in order to remain competitive with Absa. "The bar has been lifted in South Africa, and given the highly competitive market it is logical that other banks would follow suit," says Gert Kruger, head of economic capital and the Basel II implementation project at FirstRand Bank.
Kruger says FirstRand's decision was also based on the realisation that the benefits of the advanced approach have become clearer as a result of increased availability of information and guidance from the Basel Committee, academics and practitioners. "Furthermore, internal development has also allowed us to move forward and to be able to take the decisions to go advanced," he says.
Errol Kruger, registrar of banks at Sarb, is not convinced that the decision to go for the advanced IRB approach is "as cut and dried as it appears" at the major banks. "There is certainly the potential for banks to hype up what they are doing to spook the competition," he says. Although he is certain that the leading banks will honour their promises to adopt the advanced IRB approach, Kruger is unsure if this will be possible at the outset of Basel II implementation.
Regardless of whether the five leading banks will be ready for advanced IRB by January 2008 or not, the soul-searching that has accompanied Barclays' takeover of Absa has focused minds on the challenges posed by Basel II overall. While Kruger at FirstRand says that the advanced IRB approach has no overly onerous technological implications compared with the foundation approach, it does require enhanced methodologies, improved robustness of data and more detailed information, specifically with regard to recovery data. Under the advanced approach, figures for loss-given defaults and exposure-at-default must come from a bank's estimates rather than from the local supervisor.
Ernst & Young's report shows that tier-one banks cite data gathering as the number one potential difficulty in implementing the advanced approach. As Absa's Blaauw notes, much of the information required for the advanced IRB approach is simply not available in South Africa. "The requirement is for seven years of data and we don't have anything like that. We are lucky in that we will be able to leverage the work that Barclays has already done elsewhere and we will be able to adapt and recalibrate models they have used elsewhere but it will prove a challenge for the industry in South Africa overall."
Fill the gaps
FirstRand's Kruger says the answer for most banks lies in procuring international data to fill the gaps. "As long as you can create a reasonable relationship between the data sets, it can prove helpful," he says. "But of course, there are always going to be issues about data of loss given default given the importance of local legal environments and other rules."
Other concerns revolve around the verification of models. "The requirements for model verification will be important to their ease of use," says Blaauw. "To date, it has not been clarified whether verification will be internal or external. Who would be qualified to verify a model should external verification be required? Rating agencies? Consultants? We need to know."
These issues are currently being discussed by the AIF, and a decision is expected by December. Other important factors also being considered by the AIF include the home-host issue (how supervisory authorities should co-ordinate their roles) items of national discretion and the application process for advanced and foundation IRB, including issues such as external documentation.
According to Sarb's Kruger, the AIF is the best way – indeed, the only efficient way – to address such issues: "Everyone wants clarification, but on issues such as the verification of models, the market just has to accept that we don't have 500 people available to analyse these problems and answers will be provided through the AIF when they are available."
All market observers note that the timetable for implementation is tight: the AIF's subcommittees were due to present their suggestions to the steering committee at the time Risk went to press, and working groups will then finalise the proposals that need to go before parliament for voting by the middle of next year in order to become legislation in time for implementation.
That gives the industry and Sarb less than nine months to resolve a host of thorny problems. Kruger at Sarb is sanguine and appears relaxed about the timetable: "We feel that we moved our staff to Basel II at exactly the right time – right now – because that is when the banks have also made their decisive moves," he says. "We also need to remember that further guidance could be on the way [from the Basel Committee] about a number of issues. We're not here to reinvent the wheel."
And despite their protestations, the leading banks are sufficiently well resourced, disciplined and organised to be ready for Basel II implementation – even if some aspects of it have yet to be decided. As FirstRand's Kruger notes: "If we don't get answers on many of these issues such as requirements for model validation, it won't be a huge problem because we've already started to pre-empt them. We made a decision, on models, for example, to go for external verification, which is the most arduous likely scenario. Any other outcome will simply be a bonus for us." n
The capital adequacy question
One of the main issues concerning all banks about Basel II is that of capital adequacy in Pillar I (minimum capital requirements) and II (supervisory review process). South Africa has had a capital adequacy level of 10% for some years – principally as a way of bolstering banks in case there was capital flight – compared with a requirement of 8% under the 1988 Basel Accord.
Banks are generally willing to concede that higher levels of capital adequacy than international minimums are necessary – most hold higher levels of capital than they are required to and, at the end of 2004, the average for the entire market was 13.5%, according to Sarb. However, many believe the Basel II regime could prove problematic to their business model without a rethink.
"The characteristics of the South African market mean that banks will already have to hold high levels of capital because of the types of loans they make," says André Blaauw, general manager of enterprise-wide risk management at Absa. Most South African banks have loans with a loan-to-value of 100%, unlike developed world banks. "We feel that a 10% level could result in a type of double counting that would be unhelpful for the banking industry's international competitiveness," adds Blaauw.
Errol Kruger, registrar of banks at Sarb, says capital adequacy is non-negotiable. "To a large extent, Basel II is about following the market, and for major international banks the benchmark is 12%," he says. "Moving from 10% to a lower level would be a mistake and would send all the wrong signals to the international market: 10% reflects the fact that there has to be a cost to be paid for being an emerging market."
Kruger says he is open to discussions about how Pillar II should work – and he acknowledges suggestions that part of the capital adequacy currently part of Pillar I could be factored in to calculations for Pillar II. But he also reiterates the sanctity of the 10% figure and notes that "banks get payback" as a result of maintaining capital adequacy of 10% by being able to access capital at more advantageous levels.