Comments on the 3rd Consultative paper on the New Basel Capital Accord
The Asian Bankers’ Association (ABA) supports the Basel Committee’s efforts in formulating a new capital accord that improves on various aspects of the 1988 accord. We especially support its principal objective, which we understand to be the achievement of a sounder and safer banking system through better risk management. We appreciate the Basel Committee’s efforts to take the views of the banking industry into consideration in its work, and note that the third consultative paper already reflects many of the concerns that the industry has shared with the Committee since the circulation of the first consultative paper.
Although the 1988 accord was intended primarily for internationally active banks in G-10 economies, it was eventually adopted for broader application worldwide. We anticipate that the new accord will likewise be widely implemented beyond the G-10 and in most jurisdictions within Asia. Consequently, it is our concern that the new accord also reflect fundamental realities in our region, specifically the significant role of real estate in bank lending (particularly to small and medium enterprises) and the situation of banks in rapidly evolving emerging markets, which differ from those in the long-established and mature G-10 markets.
It is therefore our view that some modifications to the current draft are indispensable for ensuring the accord’s effectiveness and the success of its implementation. We hereby express our concerns with respect to certain issues arising from the current draft and urge the Basel Committee to consider these recommendations. Our concerns focus on a few key areas in so far as they have a broad significant impact throughout our region: operational risk, the treatment of real estate under the standardized and internal ratings based approaches to credit risk, risk weights under these approaches and the supervisory review process or Pillar 2. In addition, we also offer some suggestions on other specific areas.
I. Operational risk
We recognize that the development of the framework for calculating operational risk has not yet reached a level where most banks are confident that operational risks are accurately captured in the most satisfactory way possible at this point. Some of our member banks have expressed doubts about the suitability of using gross income to quantify operational risk. Others have pointed to significant regional differences in operating scenarios that would render the application of the á and â factors, given the values assigned to them in the basic indicator and standardized approaches, irrelevant or inaccurate in certain jurisdictions.
We assume that, as drafted in CP3, the inclusion of a measure of operational risk in the denominator of banks’ capital ratio would increase capital requirements for many banks in the region’s emerging markets and burden them more heavily than their counterparts in mature markets. While the new accord contains incentives for banks to adopt the Advanced Measurement Approach (AMA), including the recognition of the risk mitigating impact of insurance, the AMA remains too complex and expensive for most banks in developing economies to adopt, in view of the quantitative and qualitative requirements for the use of sophisticated regulatory capital options. We therefore believe that further development of the framework for calculating operational risk should ensure that the costs of qualifying for the AMA do not become a disincentive to invest in more advanced operational risk management systems.
Consequently, we suggest further work by the Basel Committee on this issue. While some of our member banks have suggested that operational risk be handled under Pillar 2 instead of Pillar 1 until ways to measure it are found that benefit from extensive confidence in their accuracy and relevance in most, if not all jurisdictions, we could also agree to its continued treatment as part of Pillar 1 (subject to further development), with the following modifications:
- The á and â factors in the basic indicator and standardized approaches, which are assigned a value of 15% and a range of values from 12% to 18%, respectively, should be lowered.
- As both the range of â factors compared to the á factor and the application of a single â factor of 18% to gross income for all business lines under the alternative standardized approach do not provide incentives to move from the basic indicator to the standardized approach, even as the qualifying criteria for the standardized approach calls for significant investment, the incentive of a lower capital charge should be given to banks adopting the standardized approach compared to those using the basic indicator approach.
- The mitigating effects from business contingency planning (BCP), as a form of insurance, should also be recognized under the AMA, such as in the form of a capital offset, to provide incentives for banks to develop sound BCP. Given that the assessment of benefits from BCP is primarily qualitative in nature, the discretion as to the extent of capital reduction could be left to the national supervisory authorities under Pillar 2.
II. Credit risk and eligible collateral
Most banks in Asia have a high concentration of mortgage loans and welcome the expansion of eligible collateral to include residential and commercial real estate (RRE and CRE) in CP3. We see the new accord favoring the adoption of the IRB approach in our region, as a wider list of eligible collateral for credit risk mitigation is available under this approach compared to the standardized approach.
However, a wider recognition of real estate in both the standardized and IRB approaches to credit risk measurement remains important. Limiting the role of real estate as eligible collateral for credit risk mitigation will pose very difficult problems for banks in a number of jurisdictions within Asia, where land and buildings make up the greater part of collateral and real estate lending accounts for the majority of certain banks’ portfolio. More significantly, this issue is of special relevance to small and medium enterprises, as lending to this segment is frequently collateralized by real estate throughout most of the region. We continue therefore to urge such wider recognition of real estate, which could be achieved through the following:
Under the standardized approach,
- National discretion could be allowed in expanding eligible collateral for credit risk mitigation beyond financial collateral to include non-financial assets.
- Banks could be allowed to allocate the portion of claims secured by RRE to the corresponding 35% risk weight bucket, and to treat the remaining unsecured portions similar to either corporate or retail claims.
Under the IRB approach,
- The criteria for including CRE as eligible collateral, which we consider too strict, could be relaxed or a set of risk weight adjustment factors introduced.
- The 10% floor set for Loss Given Default (LGD) for retail exposures secured by residential properties could be removed or reduced. We believe that this floor considerably limits the recognition of property collateral, and is unnecessary, considering that banks are already required to validate their LGD estimates before adopting the IRB approach for retail exposures, and that validation will have to be an ongoing process to ensure that LGD benchmarks remain relevant.
- The LGD range for real estate secured lending, including the 35% LGD floor, could be extended. We note that LGD ranges from 35% to 45% regardless of the loan-tovalue ratio (LVR), and that the LGD remains floored at 35% even if LVR is greater than 70%; extending the LGD range would thus serve to recognize the effect of a lower LVR. We also consider the 35% floor as too conservative, especially when compared to the 45% LGD for unsecured lending. Additional consideration needs to be given to the reality that LGD for real estate lending can vary substantially between different markets and between different types of properties within the same market.
III. Risk weights used in measuring credit risk
We propose the following modifications in line with practices and conditions in our region:
Under the standardized approach,
- We are concerned that the lower risk weight applied to unrated borrowers (100%) than to those rated under BB- (150%) would discourage many borrowers from being rated, as well as lower-rated borrowers from disclosing their ratings. We therefore suggest that both unrated and lower-rated borrowers be placed on an equal footing, though keeping in mind that too high a risk weight for unrated claims may damage this market segment, especially SMEs, most of which are unable to afford the cost of ratings or obtain favorable credit ratings.
- We suggest a reduction of risk weights for some claims to be more in line with local practice and conditions. These include the 35% risk weight applied to lending fully secured by mortgages on residential property and the 75% risk weight for exposures included in a regulatory retail portfolio. In addition, applying the 100% risk weight to qualifying residential mortgage loans that are past due for more than 180 days instead of 90 days could be considered. Auto loans could also be risk weighted somewhere between the regular retail portfolio (75%) and fully secured mortgage loans (35%).
- As credit funds set up by governments play an important role in some economies in the region, we suggest that such funds, though unrated, be given risk weights that are the same or one notch lower than sovereigns.
Under the IRB approach,
- We suggest a reduction of risk weights applied to specialized lending, as the high risk weights would eventually result in much higher capital cost compared to other types of assets, such as general corporate exposures and retail lending, and unduly disadvantage banks that are involved in specialized lending.
- Under the IRB approach to securitization, we suggest a reduction of the 7% risk weight applied to an AAA rated super senior position, which we deem excessively high in view of the already significant protection arising from the structural subordination as well as the minimal default correlation, especially from a highly rated protection provider.
IV. Other specific issues concerning minimum capital requirement (Pillar 1)
In addition to the above, we would like to share our member banks’s views on other specific issues concerning Pillar 1. We invite the Basel Committee to consider the following:
- Eligible collateral: Other forms of collateral aside from those currently proposed in CP3 may be considered to be recognized for credit risk mitigation, subject to national discretion, such as assignment of contract income and sales proceeds from third parties and assignment of insurance policy.
- Eligible credit derivatives: Eligible credit derivatives for recognition under the standardized approach to credit risk measurement need not be limited to credit default swaps and total return swaps, but could include more products.
- Definition of SMEs: The definition of SMEs (where banks are permitted to separately distinguish exposures to SME borrowers under the IRB approach for corporate credits) needs to be adapted to diverse conditions in various jurisdictions. The definition given to SMEs as firms with total annual sales of less than €50 million is not suitable for most developing economies, where a lower threshold would be more applicable.
- Definition of default: The exercise of national discretion by supervisory authorities in providing guidance related to the definition of default (under the IRB approach to measuring credit risk) carries the risk of significant differences across jurisdictions where banking groups operate. The importance of harmonizing rules among supervisory authorities may need to be emphasized.
- Calculation of capital for retail exposures under IRB approach: The option of using the Expected Loss (EL) approach has been dropped in CP3, which now only provides for the use of the PD/LGD approach. As this is not the way most banks measure and manage risk in their retail portfolios, the Basel Committee may wish to reconsider the reinstatement of EL as an option.
- Risk weights for High Volatility Commercial Real Estate (HVCRE): For specialized lending, the possibility of allowing banks to apply the advanced IRB approach to derive risk weights for HVCRE without the need for approval by the supervisory authority, though with a larger weighting, may be considered.
- Specific risk capital treatments under the standardized methodology within the trading book: The 8% specific risk capital charge for government paper with an external credit assessment below BBB- appears too high, and could be reduced.
- Recognition of provisions: As specific and general provisions are allowed to offset capital requirements (subject to cap) for exposures in the banking book, by the same token, valuation adjustments or reserves arising from marking to market of counterparty exposures in the trading book to account for potential deterioration in the credit quality of the trading counterparty may also be accorded similar treatment as provisions in the banking book for reducing regulatory capital requirements.
- Recognition of general provisions as offsets for expected loss under the IRB approach: CP3 allows excess general provisions not in Tier 2 to be used to offset expected loss only to the extent that EL exceeds the general provisions in Tier 2. As there may be tax incentives for maintaining general provisions in excess of EL in certain jurisdictions, there could be greater flexibility in allowing excess general provisions to offset risk weighted assets.
- Supervisory slotting criteria for specialized lending (supervisory rating grades for income-producing real estate exposures under the IRB approach): While the wording of the criteria is suited to developments for lease, it is not very appropriate for developments for sale. For greater clarity, slotting criteria for special purpose vehicles used for property development for sale may be included.
- Transitional arrangements for the IRB approach: The requirement to have a minimum two (2) years of data by the end of 2006 does not give banks intending to use the IRB approach sufficient time to undertake preparations, as these require substantial resources and time to enhance management systems and collection of data. The Basel Committee may wish to consider the reduction of the data requirement period to one (1) year upon the implementation of the new accord at the end of 2006.
V. Supervisory review process (Pillar 2)
We are conscious of the important role that supervisory authorities will have to play in ensuring the successful implementation of the new accord. Specifically, we believe that greater consideration should be given to how the supervisory review process could be harnessed to address the following issues:
- Procyclicality of regulatory capital requirements: Various ways to address this issue have been proposed by both academic experts and practitioners, such as reducing the slope of the IRB capital function and varying percentages of capital required for each dollar of risk weighted asset over the cycle, among others. Supervisory authorities need to seriously address this issue and work with industry to develop appropriate solutions.
- Concentration risk: While this issue may not have been effectively addressed by the granularity adjustment factor, which was earlier proposed but now dropped, it still needs to be taken into consideration as a key risk driver to bank portfolio that has not been captured in the new accord. Until full borne portfolio credit risk modeling, which may be able to adequately capture this risk, is allowed for use in calculating capital requirements, the issue of concentration risk could be addressed in Pillar 2, such as through stress testing of correlation in the portfolio.
In addition, we remain concerned about the possible impact of national discretion on areas such as recognition of additional eligible collateral on competitive equality among jurisdictions. Developing detailed guidelines for supervisory authorities in the exercise of discretion may be considered, in order to ensure a level playing field.
VI. Issues needing clarification
There are also a number of issues where greater clarity is needed, and more detailed explanations in the text of new accord or relevant annexes would be useful. Among these are the following:
- Treatment of residential mortgage loans that are extended to individuals renting the property to a third party;
- Rules for purchased receivables, whether to use invoice amount or loan amount to calculate EAD and the definition of and limitations applying to the top-down and bottom-up approaches related to purchased receivables;
- Criteria for distinguishing between HVCRE and Income Producing Real Estate (IPRE), and possibility of moving a borrower from one specialized lending category to another as the business moves on to a new stage;
- Various issues related to the alternative standardized approach to measuring operational risk, including scope of application, which activities are included in each business line, procedures for validation and review of bank’s operational risk management processes and assessment system, and calculation of granularity to capture major risk drivers, among others;
- The qualifying criteria for recognizing insurance as an operational risk mitigant for banks using the AMA, in particular haircuts to reflect the residual term of the policy and the minimum notice period for cancellation;
- The mapping of insurance coverage to actual operational risk loss exposure;
- Definition of “significant market maker” related to the valuation methodology of treasury activities; and
- Rules governing regulatory reporting of an international banking group using different approaches in different jurisdictions.
In forwarding the above views and recommendations to the Basel Committee, the Asian Bankers’ Association wishes to contribute to the formulation of an accord that can be effectively implemented across jurisdictions, taking into account current conditions in both developed and emerging markets. We hope that these views and recommendations would be taken into consideration in finalizing the new capital accord, and look forward to continued interaction with the Basel Committee in the future.