The original Basel Accord’s simplicity probably helped its introduction by national prudential regulators. But the insensitivity to variations in risk (both between and within risk categories) had the potential to increase the incentive for risk-taking behaviour. Hogan & Sharpe (1997a and 1997b) and Gup (2003: 74), for example, argue that attaching a risk weight of 100% to all commercial loans irrespective of counterparty allowed banks to pursue higher-risk (to achieve higher return) lending since this requires no more capital than less-risky lending but has greater upside income potential. Basel II addressed this shortcoming by enabling the use of a much wider range of credit-risk weights, by providing for the use of different approaches to determining risk weights and by extending the capital requirement to cover all risks banks face.
Basel II has three pillars.
The first deals with a bank’s core capital requirement (Pillar 1); the second allows for supervisor discretion to adjust this requirement to allow for additional risk and particular circumstances (Pillar 2); and the third fosters market discipline (Pillar 3).
Pillar 1: Capital requirements for core risks
Pillar 1 refines the calculation of regulatory capital in three important ways. First, it uses a more granular approach to credit-risk weights; second, it provides banks (subject to the regulator’s approval) with a choice of methods for calculating risk weights for certain types of risk; and third, it incorporates operating risk into the capital requirement.
The anatomy of Pillar 1 is represented in Figure 1, which shows the Basel II innovations in bold type to distinguish them from those of Basel I. Note the introduction of three possible approaches to the calculation of the capital requirement for credit risk under Basel II; the standardised (externally set) risk weights and two approaches that rely on internal ratings (the foundation internal ratings basis, FIRB, and the advanced internal ratings basis, AIRB). Observe the introduction of a capital requirement for operating risk also provides for three approaches to the calculation of the capital requirement. An introduction to each innovation follows.
The calculation of the capital requirement for credit risk starts by dividing a bank’s assets into five categories (corporate, sovereign, bank, retail and equity) within which there are sub-groups reflecting the different risk parameters for each asset type. The capital requirement for each represents an attempt to capture the average probability that a loan to each category of borrower would default, and the proportion of the loan that would be lost if default occurred.
Under the standardised approach, risk-weights are prescribed for each risk category, where the risk of each is rated by the borrower’s externally-determined credit-rating agency such as Standard and Poor’s (S&P). The value of the loans in each category is multiplied by the prescribed risk weight and the product is multiplied by 8 per cent to determine the minimum capital requirement. To illustrate, there are six credit-rating grades for corporate loans, where grade 1 covers loans rated AAA to AA– (on Standard and Poor’s long-term scale), grade 2 covers A+ to A– and so on. The standard risk weights vary from 20 per cent to 150 per cent for these grades (APRA, APS 112 and APG 112). While this is the ‘default’ approach, which can be viewed as an extension of Basel I, it represents a substantial advance. Basel I used just four risk weights, two of which (in Australia) covered the bulk of bank balance sheet assets. The standardised approach requires an improvement in risk-management systems to generate the data to satisfy Basel II’s more granular risk categories. Most ADIs are using this approach, which is expected to generate, on average, a modest reduction in regulatory capital (Egan 2007).
The more radical innovation is the provision for banks to use either of two internal rating approaches subject to the regulator’s (that is, APRA) approval. The foundation internal ratings-based approach (FIRB) uses internal estimates of the probability of loan defaults (PD) and feeds this into a more complex probability-based formula (that relies on the supervisor’s estimates of the other risk components) to determine the risk weight to be used to calculate the amount of capital to be held against the loan. The advanced internal ratings-based approach (AIRB) uses internal estimates of loss given default (LGD) and the other risk components (effective maturity and the exposure at default) in a prescribed formula to determine the risk weight and hence the capital charge against a loan. These approaches derive from the internal risk assessments banks (including Australia’s big banks) began undertaking in the 1990s and thus Basel II can be viewed as following industry practice.
Basel II does not change the two methods that can be used for assessing the capital requirement for market risk introduced in 1996. However, it introduced a capital requirement for operational risk exposures. Operational risk refers to the risk that losses may result from a lack of verification and control processes (such as the loss of Є4.9 billion at Société Généralé due to a trader’s ability to circumvent operating systems that was revealed in January 2008). Three approaches for assessing operating risk are available; two that are relatively simple (the basic indicator and several standardised approaches) and the third advanced measurement approach (AMA) could be used by banks ‘with advanced operational risk measuring and modelling capabilities’. Under the standardised approach, an ADI divides its activities into three categories — retail banking, commercial banking and all other activities, which have different capital requirements — and the sum of these requirements sets the ADI’s operational risk-capital requirement. The capital requirement for retail and commercial banking is based on an ADI’s gross outstanding loans and advances (as an indicator of its operating risk exposure) whereas for the third category the capital requirement is based on the ADI’s gross income from these activities (APRA, APS114). To be accredited to use the advanced approach banks must have ‘an operational risk management framework that is sufficiently robust to facilitate quantitative estimates of the ADI’s ORRC (operational risk regulatory capital) that are sound, relevant and verifiable’ in relation to the ‘complexity of the ADI’s business’ (APRA, APS115: para. 21).
Three banks were accredited to use the advanced methods from January 2008 and a fourth (NAB) was given approval to use the foundation IRB approach. Three other banks have applied to move to an IRB approach during 2008 and are operating under Basel I in the meantime (RBA 2008a: 67). The advanced approaches are expected to reward banks with modest reductions in regulatory capital (for lower credit-risk exposures); although 10 per cent will be the maximum reduction in 2008 and 2009 (under Pillar 2 provisions) while the banks are demonstrating the performance of their risk-management models (Egan 2007).
Pillar 2: The Supervisory Review Process
Pillar 2 has two aspects. The first requires banks to assess their overall risk profile (in addition to the risks specified under Pillar 1) and to calculate any further capital that should be held against this additional risk. The additional risks potentially identified under Pillar 2 include credit concentration risk, liquidity risk, reputation and model risk. Consequently, Pillar 2 could be expected to add to the amount of capital held by banks (and offset the lower credit-risk capital requirement).
The second aspect of Pillar 2 is its inclusion of a ‘supervisory review process’. This allows supervisors to evaluate each bank’s overall risk profile and to mandate a higher prudential capital ratio where this is judged to be prudent (APRA, APS 110). APRA’s decision to increase NAB’s capital requirement following its foreign exchange losses (in January 2004) illustrates this process.
Pillar 3: Market Discipline
Pillar 3 requires disclosure of information regarding the calculation of bank capital positions and risk-management processes designed to strengthen the capacity of security markets to respond to changes in bank risk profiles. The idea is that banks which the market judges to have increased their risk profiles without adequate capital will have their securities sold down in debt and equity markets. The additional costs that this will impose on financing bank operations will provide an incentive for management to modify either the bank’s risk profile or its capital base. This dimension of Basel II is thus designed to complement Pillars 1 and 2 by providing additional discipline on bank risk-taking behaviour.
APRA’s prudential information disclosure requirements are most detailed for the Australian-owned ADIs that use the advanced risk-management approaches because of their use of internally-generated risk ratings. They are required to report quantitative risk-management information on a semi-annual basis and qualitative risk-management information on an annual basis, as well as reporting basic capital-adequacy information on a quarterly basis. The reporting requirements are less detailed for the ADIs that use the standard risk weights and for overseas-owned ADIs, assuming their home regulator’s prudential information disclosure requirements are equivalent to APRA’s (APRA June 2007 and APS330).
Securitisation
In this and the next sub-section Basel II’s approach to securitisation (BCBS 2004b: 120–43) and credit risk mitigation (BCBS 2004b: 31–51) within Pillar 1 are introduced because of the role these processes played in the crisis triggered by the US sub-prime loan debacle (which is discussed in section 5). The process of moving assets (these principally have been housing loans) off the balance sheet via securitisation (a variation of the ‘originate-to-distribute’ model used by Australia’s loan originators) has been an important feature of bank asset-liability management. In Australia it has been used especially by the regional banks.
The assets are securitised through the services of a special-purpose vehicle (SPV) established by a bank. The SPV arranges the issue of asset-backed securities (mortgage-backed securities, MBSs, where housing loans are involved) to investors and pays the bank for the loans with the proceeds. The bank avoids a capital requirement for the securitised assets provided the arrangement ensures it is no longer exposed to any risks associated with the assets, such as the risks that would arise if the originating bank agreed to any explicit credit enhancement of the securities or from implicit liquidity or solvency support for the SPV which could result in the securitised assets being brought back onto the originator’s balance sheet.
Credit-risk mitigation
An important dimension of Basel II is its treatment of credit-risk mitigation techniques such as the use of collateral, guarantees (by a third party) and other credit-risk reduction measures such as credit derivatives, which reduce the amount of loss in cases of default. Credit-default swaps (CDS) are the most extensively used credit derivative. They provide banks with the opportunity to buy protection against default events on one or more of its assets, which would reduce its credit risk. For example, a bank could purchase credit-risk protection on a specified set of loans or corporate securities held by the bank by issuing a CDS and paying a premium (this would be paid six monthly at the agreed rate) to the party that decides to accept the credit risk (such as a bond investment manager or another bank that wants to diversify its credit risks). The protection seller faces the obligation to compensate the protection buyer should pre-defined default events occur on the specified parcel of loans or securities. Should a default event occur the bank would receive a compensation payment and this lowers its loss given default, whereas should no default event occur the seller would receive the premium payments without having to compensate the bank.
Basel II explicitly recognises the role of banks’ increasing use of instruments such as credit derivatives. Since these instruments reduce the risk of loss they reduce a bank’s capital requirement. The reduction depends on the credit standing of the provider of the credit-risk mitigation instrument, such as the protection seller in the case of CDS. Thus the risk-weighted assets are adjusted using a risk weight appropriate to the risk class of the protection seller. For protection sold under the CDS, the same process is followed but the risk weight applied is that appropriate to the reference credit being protected.
No comments:
Post a Comment