INTRODUCTION
1. The Consumer Council is pleased to submit its views to the Hong Kong Monetary Authority (HKMA) regarding a discussion paper on funding and premium assessment for a deposit insurance scheme in Hong Kong.
2. This paper sets out the Council response to the HKMA in relation to the issues raised in the discussion paper.
3. In summary, the Council supports most of the suggestions made by the HKMA.
COUNCIL RESPONSE
The adoption of an ex ante funding approach
4. In the discussion paper, two funding approaches for deposit insurance are presented for consideration, i.e. building a fund on an ex ante basis or on an ex post basis. HKMA recommends that an ex ante funding approach should be adopted.
5. In its previous submission, the Council was disposed to favour an ex ante approach as this is equitable and fair to banks. Banks would also be able to better estimate future funding requirements. Further, the ex ante funding approach would provide advance funding to ensure the prompt reimbursement of depositors' claims after a bank failure. The results of the earlier public consultation also showed that an ex ante funding approach was generally favoured by the banking industry. In this connection, the Council reiterates its previous position in favour of an ex ante funding approach.
6. It is noted in the discussion paper that although it is recommended that the ex ante funding approach should be adopted, it is not government's intention to establish a large DIS fund because the DIS cost would be unnecessarily high. A back up funding is therefore in place which essentially provides a liquidity facility to enable the DIS to meet any shortfall. The Council supports the provision of back-up funding to draw from the Exchange Fund's line of credit, to provide financial support in times when the fund is inadequate to meet obligations to depositors.
7. To ensure the smooth running of the DIS fund, the Council recommends that the fund should be managed by implementing appropriate investment policies and procedures, with internal controls, disclosure and reporting systems.
The appropriate size of the DIS fund ("target fund size") and the manner in which it should be built up and maintained
Target fund size and the related assumptions
8. The HKMA makes the following assumptions: possible shortfall loss (50% asset recovery rate), funding costs of the fund (8% p.a.), the probability of default of insured banks (set by reference to banks credit rating, the level of annual premium (8bp p.a.), build-up period (3-4 years), and the upper and lower limits of the fund (+/- 30% of target fund size) to produce a target range. The Council replies on the professional judgment of the HKMA and accepts that these assumptions are reasonable.
9. In considering the question of the appropriate size of the fund and the level of annual premium, there is no single answer to this question, because it involves a trade-off. That is, too large the fund implies higher premium level which may affect the financial health of the banking industry and the cost implications to depositors, whereas too small the fund may provide inadequate funding in resolving claims from depositors in failed banks. The Council therefore considers that, although the fund size and the premium level are smaller than that estimated by the Consultant's $2-$3 billion with premium at 10bp, a smaller starting fund size of HK$1,500 million, achievable in three to four years by charging at a lower premium 8bp at build up stage is acceptable, but to keep review the fund size and premium level once the system has fully bedded down and in the light of experience.
Hard target vs. soft target funding approach
10. In the discussion paper, two possible methods are suggested for consideration, i.e. hard target vs. soft target approaches. Similar to the previous discussion, there are trade-offs between a soft target (i.e. a range) on the one hand and a hard target on the other hand. As noted in the discussion paper, a soft target has the advantage of generally being less procyclical and would enable the fund to build larger rebates and create a bigger cushion to absorb losses. But this could also impose unnecessary cost on the banks. On the other hand, a hard target results in surcharges whenever the fund is below a particular ratio and rebates whenever the fund is above that ratio. This would give rise to high volatility and low predictability of the level of premium. On balance, it appears to the Council that soft target approach is the preferred model for higher certainty of the level of premium.
11. In the discussion paper, this approach is applied to the initial target of HK$1,500 million, with a reasonable buffer of 30% above and below the soft target, the initial target range is from HK$1,050 million to HK$1,950 million. Federal Deposit Insurance Corporation (FDIC) is considering adopting the same 30% buffer. For the same reason, the Council considers a buffer of 30% above or below target is acceptable.
Apportionment of rebates and surcharges
12. In the case of surcharges, the discussion paper suggests that the amount levied on an individual bank would be assessed in the same way as premiums (i.e. on current risk credit rating). The Council considers this acceptable.
13. The discussion paper suggests that it would seem more appropriate to share rebates among banks in proportion to their past contribution to the fund. A further issue that will arise when the DIS has been up and running for a number of years is how far back it should go in aggregating past contributions to the fund. In the case of FDIC, it is proposing that there should be a limit on the length of the "look-back" period of ten years (i.e. only premium paid during that period would count towards determining share of rebates). The Council agrees with the rationale behind this and accepts that rebates should be apportioned according to their past contribution to the fund, and a limit on the length of "look-back" period should be set.
The appropriate level of the premium and the possible introduction of a differential premium system based on the supervisory ratings of the banks
Flat rate vs. differential premium systems
14. In the last consultation, the consultant recommended it would be more appropriate to introduce a flat rate assessment system at least for the early stages of the DIS, but that a move towards a risk-based assessment method could take place at a later stage once the system has fully bedded down and if considered necessary.
15. The Council, in its previous submission, expressed support for this approach and agreed that the calculation of premiums in the DIS should eventually be based on relative risk as this would reduce moral hazard and avoid implicit cross-subsidies among banks.
16. It is noted in the discussion paper that the Executive Council has agreed that the question of risk-based premium should be looked at again. In view of this decision, the HKMA proposes a relatively simple system of varying premiums which might be described as "differential" rather than "risk-based". This approach is proposed to award the better performers with lower premiums and give an incentive to the others to improve.
17. On balance, the Council is disposed to agree that a premium system differentiated on the basis of individual-bank risk profiles can mitigate criticisms of inequitable treatment and may encourage more prudent risk-management practices at banks. However, this only applies if the information required to implement a risk-adjusted differential premium system is available.
CAMEL rating system for premium setting
18. It is proposed in the discussion paper that the differential system would be solely based on supervisory ratings, i.e. the Capital, Asset quality, Management, Earnings and Liquidity (CAMEL) rating system used by HKMA and many other regulators around the world. HKMA suggests that the CAMEL rating system might be a reasonable starting point, perhaps to be replaced with a more sophisticated approach in the light of experience.
19. Based on the HKMA's estimate under the CAMEL system, 86.7% of the banks will pay the standard rate of 8bp, 4% pays below the flat rate and 9.3% pays higher if the CAMEL system is used for premium setting.
20. The next question raised in the discussion paper is how wide the range of premiums should be. HKMA suggests that the range of premiums be set at 5, 8, 11, 14bp under 4 categories rating, a relatively small gap (3bp) between categories. The HKMA considers it is appropriate that once the system had bedded down, it might be possible to move to wider increments coupled with more objective indicators of differences in risk. The Council agrees with this approach.
21. The discussion paper notes that the main objection to the use of the CAMEL system is that it relies on the subjective judgment of the HKMA. However, the HKMA states that in any case, the objective factors that would be relevant are to a large extent already taken into account in arriving at CAMEL ratings, though in a rather judgmental manner. In addition, the system has been in use for some time and the composite CAMEL rating is disclosed to AIs on a regular basis without serious disagreements so far. If CAMEL is used for premium setting purpose, the HKMA said that it would review its internal guidelines to staff on the CAMEL system, including to incorporate the risk management rating which is derived from its risk-based examinations.
22. The Council notes that most newly established systems initially adopt a flat-rate system, given the difficulties associated with designing and implementing a risk-adjusted differential premium system. However, judging from the HKMA's estimate, 86.7% of banks will pay a standard rate under the proposed CAMEL system. Given its apparent advantage over the flat rate system in reducing moral hazard and avoiding implicit cross-subsidies among banks, the Council accepts that the subjective CAMEL supervisory ratings be adopted as a starting point for premium setting purposes, instead of the flat rate original proposed. Having said that, the Council suggests that the system would be replaced with a more sophisticated approach in the light of experience (e.g. a combination of objective and subjective factors adopted by Canada Deposit Insurance Corporation as quoted in the discussion paper).
23. The Council suggests that the bases and criteria used in a risk-adjusted differential premium system should be transparent to all participants.
Rating confidentiality
24. The HKMA notes in the discussion paper that it would be important that the CAMEL ratings assigned to individual banks remain confidential. Provisions would need to be introduced in the relevant legislation to ensure that the CAMEL ratings and premiums charges do not become public information. Similar provisions are included in the CDIC Act.
25. The Council notes that CAMEL ratings at present are kept within the HKMA and each individual bank. These ratings are not available to the public or in banks' annual reports. According to the HKMA, this non-disclosure policy is universal in other countries. The Financial Stability Forum however stated this is
"an important but delicate issue, and that policymakers should consider whether to allow the release of information related to the risk profile of each bank, or to restrict this information for confidentiality or other reasons."
26. The Council understands the rationale behind keeping sensitive information like rating and premiums confidential because the disclosure of such sensitive information, substantiated or otherwise, may put the stability of the banking sector at stake. The HKMA also mentions that not disclosing individual bank's rating would prevent lawsuits as this sensitive information might cause a bank run.
27. Further, the HKMA considers the lack of CAMEL ratings would not damage the public's right to information as there is plenty of financial information in the market to assist consumers to understand a bank's position. In fact, information on capital, asset quality, earnings and liquidity (i.e. CAMEL) are publicly available (though not on non-financial aspect such as management).
28. The Council accepts the arguments for maintaining an appropriate level of confidentiality. However, as a matter of general principle it believes that consumers should have access to information in order to make an informed choice. The manner in which a balance is to be achieved in meeting the need for confidentiality and transparency is a matter that can be left to the Government.
29. The Council also suggests that additional rating information that HKMA will use to expand the CAMEL criteria should be made transparent to the public.
30. The Council notes that the FDIC also does not release its ratings on the safety and soundness of banks and thrift institutions to the public. However, it provides a searchable database to assist consumers to find institutions and their branches in order to determine their status as insured depository institutions, their financial condition and their condition relative to other institutions. The databases also contain other financial and non-financial information about individual financial institutions as well as certain aggregate financial statistics for comparative use. As a service to consumers, FDIC has compiled a listing of several financial institution rating services. This may be one aspect to meet consumer's information needs. In this connection, the Council recommends that a database containing comparative financial and non-financial information (similar to FDIC) should be provided by the Government.
OTHER ISSUE
Public awareness
31. In relation to the dissemination of rating information, a public-awareness program should be introduced by the HKMA for promoting and facilitating public understanding of the characteristics of the DIS and its benefits and limitations. This can build public confidence in the banking sector. Additionally, such a program can help to disseminate vital information when failures occur, such as guidance regarding how to file claims and receive reimbursements.
Link to the discussion paper by Hong Kong Monetary Authority.
