Review of the HKMA's Work on Banking Stability

Review of the Hong Kong Monetary Authority’s Work on Banking Stability by Mr David T R Carse

July 2008

Contents Executive summary Introduction ..............................................................................................................................................1

1. TRENDS AND ISSUES 1.1 The state of the Hong Kong banking sector ..........................................................................2 1.2 Summary of the main trends and issues ................................................................................4 Business integration with the Mainland ................................................................................4 Impact of the sub-prime crisis...............................................................................................7 Development of Hong Kong as an international financial centre .........................................9 Increasing competition ........................................................................................................10 Erosion of financial boundaries..........................................................................................13
2. GOVERNANCE ISSUES RELEVANT TO BANKING STABILITY The role of EFAC ................................................................................................................17 Other checks and balances..................................................................................................17 The role of the Advisory Committees...................................................................................17 Appeal mechanism...............................................................................................................19 The role of HKAB................................................................................................................19 Strategic planning ...............................................................................................................20
3. THE MA’S FUNCTIONS AND POWERS IN THE BANKING ORDINANCE The functions of the MA.......................................................................................................21 The HKMA’s role in consumer protection...........................................................................22 Supervision of remittance agents and money changers.......................................................24 The MA’s powers of sanction ..............................................................................................25 The MA’s powers of investigation .......................................................................................27 The MA’s powers of intervention.........................................................................................28
4. THE SUPERVISORY FRAMEWORK Summary of the current framework .....................................................................................29 Overall assessment of the current approach .......................................................................30 The need to streamline the off-site review process..............................................................30 The FSA internal review in relation to its supervision of Northern Rock............................33 The need to rationalise the on-site examination process.....................................................34 Enhancement of macro-prudential surveillance..................................................................37 High turnover of staff ..........................................................................................................39
5. THE POLICY FRAMEWORK Supervisory policies.............................................................................................................40 The organization of the HKMA’s policy framework............................................................43
6. AUTHORIZATION The three-tier structure........................................................................................................45 Licensing and promotion.....................................................................................................48
7. SAFETY NET ARRANGEMENTS.............................................................................................49 Appendix 1 Terms of reference Appendix 2 Summary of recommendations

i Executive summary

1. The objective of the review is to make recommendations on how the Hong Kong Monetary Authority (HKMA) can best discharge its functions in promoting the general stability and effective working of the banking system under the Banking Ordinance (BO), taking into account recent and likely future developments in Hong Kong’s banking system and the changing nature of the risks facing the system. The main focus is on the HKMA’s regulatory and supervisory functions. The main conclusions and recommendations of the report are summarised below. Trends and issues
2. The Hong Kong banking system is currently in robust condition and has recovered well from the Asian Financial Crisis. The global banking system is now facing a new crisis triggered by the problems in the US sub-prime market. Although the Hong Kong banking system has so far been relatively unscathed by this, the HKMA will need to absorb the various lessons to be learned from it and this will to some extent determine its supervisory agenda over the next few years. However, a more fundamental issue for the HKMA will be how to manage the growing integration with the banking system on the Mainland. The ability to expand into the Mainland will create new business opportunities for Hong Kong banks, which are facing increasing competition in their domestic market, but will also bring with it increased risks. The HKMA will need to ensure that both the banks and itself understand the nature of these risks and that there are adequate means in place to control them. Cooperation with the China Banking Regulatory Commission will be a vital factor in this.
3. The flow will not be one way. More Mainland banks will wish to enter the Hong Kong market, which will help to boost Hong Kong’s position as an international financial centre, but this trend will need to be carefully managed and supervised.
4. The continuing erosion of financial boundaries will also increase the importance of effective regulatory cooperation within Hong Kong, particularly between the HKMA and the Securities and Futures Commission (SFC). This relationship is working reasonably well but further refinements may be needed to ease the problems, as perceived by the industry, of having to deal with multiple regulators. Governance issues relevant to banking stability
5. Although the final power of decision making in the BO rests for the most part with the Monetary Authority (MA), who acts as the chief executive of the

ii HKMA, there are a number of checks and balances on the exercise of that power. The Banking Advisory Committee and Deposit-taking Companies Advisory Committee are part of that framework of checks and balances, but there are a number of ways, set out in the report, in which their role might be clarified and strengthened. Other recommendations in the report relate to the role of the Hong Kong Association of Banks, the appeal mechanism set out in the BO, and the HKMA’s strategic planning process. The MA’s functions and powers in the BO 6. Perhaps the main issue here is the extent to which the HKMA should be involved in consumer protection and whether the MA should be given a specific statutory responsibility for this. One of the lessons from the handling of the supervision of Northern Rock in the UK is the need to strike the right balance between prudential and consumer issues. An undue focus on the latter can lead to distraction from the main business of trying to reduce the risk of bank failure. Therefore the report does not recommend that the MA be given a specific statutory responsibility for consumer protection. (Nor does it consider that the HKMA should take on the additional, and potentially distracting, role of authorizing and supervising remittance agents and money changers.) 7. However, the report considers that it is legitimate and appropriate for the HKMA to remain involved in those consumer protection issues that fall within its current statutory remit to promote the general stability and effective working of the banking system. There may be advantage in the HKMA preparing a formal guideline under section 7(3) of the BO setting out what it considers that this entails. 8. The report also suggests that there is a case for strengthening the MA’s powers of sanction, investigation and intervention in the BO. The supervisory framework 9. The HKMA’s current supervisory framework is comprehensive and intensive. Coupled with the HKMA’s regular dialogue with the industry, it enables the HKMA to be aware of the main risks to the system and to individual authorized institutions (AIs). However, as the HKMA itself recognises, there is scope for improvement in the following areas: • The need to streamline the off-site review process to avoid duplication and to reduce work pressure on case officers to ensure that they are not distracted from their main task of risk assessment.

iii • The need to rationalise the on-site examination process to ensure that it is properly risk-focussed and does not impose an excessive burden on AIs and on the HKMA itself. • The need to enhance macro-prudential surveillance of the banking system, for example by improving information flows within the HKMA, in order to maximise the chances of detecting, and preparing for, future adverse trends and risks. • The need to deal with the issue of high staff turnover in the supervision area. This recognises that while the supervisory framework is important, the way it is applied by staff and the experience and expertise of those involved are equally important, if not more so. The policy framework 10. The HKMA has an extensive set of policies that largely comply with the revised Basel Core Principles for Effective Banking Supervision. In particular, it has been a world leader in early implementation of the Basel II Capital Accord. To that extent it has already fulfilled one of the principal recommendations of the Basel Committee in response to the sub-prime crisis. To a large extent, the international response represents a return to basics - focussing on strengthening capital and liquidity requirements and risk management practices. These, and other proposals of the Financial Stability Forum and the Basel Committee will help to determine the HKMA’s policy objectives over the next few years. A key priority, in addition to further enhancing the capital adequacy framework, will be to revise the liquidity regime, taking into account the guidance from the Basel Committee to be issued shortly. 11. In addition, continuing progress should be made in filling in the remaining gaps in the policy framework, including supervisory guidelines on counterparty credit risk, management of market risk and derivatives and a consolidated guideline on anti-money laundering and terrorist financing. The HKMA should also review the disclosure requirements of Hong Kong banks both to meet local needs and to take account of the forthcoming additional guidance from the Basel Committee. 12. Finally, the HKMA should aim to achieve a clearer distinction between minimum standards and best practice in its supervisory guidelines. Giving the MA a general power in the BO to make rules might facilitate this. There would also be advantage in bringing all the various elements of the policy framework together in a regulatory handbook similar to that of the SFC.

iv Authorization 13. The current three-tier structure of the banking system tries to strike a balance between flexibility of entry into the system and protection of small depositors. However, the system is now more complex than it needs to be to achieve this balance, given the decline in the number of restricted licence banks (RLBs) and deposit-taking companies (DTCs). Therefore the report recommends that the three-tier structure should be reduced to two tiers: banks and other deposit-takers (i.e. RLBs and DTCs combined). The other deposit-takers should be allowed to take call, notice and time deposits of HK$100,000 or more (or some higher figure to be determined) without restriction on maturity (other than in respect of deposits repayable on demand). 14. Otherwise, the authorization criteria set out in the BO seem appropriate and it is not recommended that these should be relaxed in order to facilitate the admission of more foreign banks. Safety net arrangements 15. The importance of effective safety net arrangements has been demonstrated by the experience with Northern Rock in the UK. The HKMA has the ability to act as Lender of Last Resort and has issued clearly articulated guidelines on how this would operate. No change is suggested in the current arrangements. This is complemented by the Deposit Protection Scheme (DPS) which was launched in September 2006. The inquest into what went wrong with Northern Rock has identified certain defects in the UK system of deposit insurance and a number of reforms have been suggested. Some of these issues were already addressed in the way in which the DPS was set up in Hong Kong – in particular, it is pre-funded (which facilitates a quick payout) and it offers 100% coverage. Moreover, since its inception the board of the DPS has mounted an extensive campaign to increase public awareness of the scheme as a means of promoting depositor confidence. 16. However, there are certain issues that require further study, such as the level of protection of deposits in the Hong Kong scheme. These should be reviewed in the light of the changes that are made to the UK scheme and the new international principles on deposit insurance that are due to be issued in response to the sub-prime crisis. Conclusion 17. It is important to emphasise that the HKMA is widely respected by the Hong Kong banking industry for its professionalism and effectiveness. It is viewed by outside commentators, such as the rating agencies, as being in the top flight of regulators internationally. The recommendations set out in this report should be viewed in this context. No fundamental deficiencies in the

v regulatory and supervisory framework or processes have been identified. But a number of enhancements can be made which will provide an even sounder foundation to cope with the challenges ahead.

1 Introduction

1. I was commissioned by the Hong Kong Monetary Authority (HKMA) to conduct a review of how it can best discharge its functions in promoting the general stability and effective working of the banking system under the Banking Ordinance (BO or the Ordinance), taking into account recent and likely future developments in Hong Kong’s banking system and the changing nature of the risks facing the system.
2. In particular, the purpose of the review is to make recommendations on the focus and priorities of the HKMA’s supervisory functions and other stability-related activities, and on the development of policies that are relevant to the stability objective. The period covered by the review is the next 5-10 years. The terms of reference for the review are set out in Appendix 1.
3. The review was not prompted by concerns about weaknesses in the current regulatory system or by concerns about the health of the Hong Kong banking system. Rather, it was considered that the completion of the programme of banking reform initiated by the last such review, by KPMG, in 1998 provided a suitable opportunity to take stock and to consider how the current arrangements for regulating the banking system can be further enhanced to cope with the challenges ahead.
4. The review has taken place against the background of the recent upheavals in the financial markets initiated by the US sub-prime crisis, the full impact of which is not yet known. The review was not commissioned in reaction to these events, but the lessons and the implications that flow from them (in relation for example to risk management) have been considered as part of the review and are discussed in more detail below.
5. The review was conducted through desktop research of external publications (including those on the websites of other regulators such as the Financial Services Authority (FSA) in the UK) and HKMA documents (reports, internal memos, policy documents and guidelines) and of case files showing how the supervisory process had been applied in practice to individual authorized institutions (AIs). In addition, discussions took place with HKMA staff at various levels and interviews were conducted with relevant external stakeholders of the HKMA: individual AIs, accounting firms, rating agencies, credit analysts, the Consumer Council, academics and members of the Exchange Fund Advisory Committee (EFAC). Written contributions were also received from the Hong Kong Association of Banks (HKAB) and the DTC Association (DTCA). My thanks to all those who provided input to the review.

2 6. This report begins by examining some of the trends and issues that will affect the Hong Kong banking sector over the period covered by the review and the implications of these for the HKMA. It then goes on to consider, and make recommendations, in respect of the various aspects of the HKMA’s regulatory and supervisory functions. These are summarised in Appendix 2. 7. It should be said at the outset that the HKMA is widely respected by the Hong Kong banking industry for its professionalism and effectiveness. It is viewed by outside commentators such as the rating agencies as being in the top flight of regulators internationally. The recommendations made in this report should be viewed in this context. No fundamental deficiencies in the current regulatory and supervisory framework and processes have been identified but there are a number of areas where enhancements can be made. The priorities over the next few years will be set to a large extent by the lessons learned internationally from the sub-prime crisis, though it should be noted that a number of the actions now being suggested (notably early implementation of the Basel II Capital Accord) have already been put in place by the HKMA. The future agenda in Hong Kong will also be set by local considerations, including particularly the need to manage the increasing business integration with the Mainland.

1. TRENDS AND ISSUES 1.1 The state of the Hong Kong banking sector
2. As noted above, the Hong Kong banking sector is in robust condition. Over the last five years, it has shaken off the impact of the Asian Financial Crisis and it is currently well-capitalised with high levels of liquidity and low levels of bad debts. Profitability has been boosted by the recovery in the domestic economy and the continuing growth momentum on the Mainland. The post-tax return on assets of retail banks in aggregate rose to 1.5% in 2007 compared with 1.2% in 2002 when the hangover effects of the Asian Financial Crisis were still being felt, and the post-tax return on equity over the same period rose from 14.3% to 22.0%. While the net interest margin remains below the level in 2002, it has risen significantly from its trough in 2004. This has been accompanied by an increase in fees and commissions as banks expand their wealth management and stock market activities: the share of non-interest income to total income rose to 44.1% in 2007.
3. Profitability has also been boosted by improved asset quality. This is illustrated by the decline in the retail banks’ classified loan ratio between 2002 and 2007 from over 5% to 0.9%. The charge-off ratio on credit cards, which peaked at around 14% in 2002, is now down to a more normal level of just under 3%, albeit there are some emerging signs of deterioration as personal bankruptcies increase. The delinquency ratio of residential mortgage loans reached a record low of 0.11% in December 2007, and the number of such

3 loans in negative equity has fallen to around 1,900 from the high point of 106,000 cases in June 2003. 10. As always, however, a number of uncertainties and potential vulnerabilities lie ahead. Domestically, lending related to stock market activities grew extremely rapidly in 2007, which contributed to the growth in the banks’ interest and fee income. Share prices fell significantly in the first quarter of 2008 reflecting nervousness at the potential global impact of the sub-prime crisis. The full effects of this on the US economy and hence on world economic growth remain to be seen. A recent econometric study by the HKMA suggests that the negative impact of a US-led economic slowdown, coupled with a consequent large and prolonged correction in domestic assets, should be moderate in terms of the resultant credit losses for retail banks in Hong Kong. But this possibility, coupled with the financial tightening on the Mainland to rein back inflation, suggests that, in the short to medium term at least, Hong Kong banks could be operating in a less benign environment than in recent years. 11. A further note of caution derives from the fact that performance among the retail banks is not evenly spread. In particular, while the profitability of the small and medium size banks (in terms of return on assets, net interest margin and return on equity) is generally respectable, it tends to lag behind the large banks (as defined for HKMA internal monitoring purposes). Moreover, while the retail banks in aggregate have ample liquidity with loan to deposit ratios of around 46% (all currencies) and 65% (Hong Kong dollars) as at end-2007, the Hong Kong dollar loan to deposit ratio of some small and medium sized banks is appreciably higher. This shows that these banks are more constrained in their funding, particularly in Hong Kong dollars, than the larger banks, which is not surprising since the latter enjoy an 80% share of the retail banks’ total deposits. 12. Small and medium size banks operate at significantly higher capital ratios than the large banks: 16-17% in aggregate compared with around 13% for the locally incorporated banks as a group. This is an indicator of balance sheet strength but can also be a constraint on balance sheet growth, particularly since some banks are at or near their limits to raise fresh Tier II capital. 13. Thus, while the performance of the small and medium size banks has improved in absolute terms since 2002 and is certainly no immediate cause for concern, they are at a competitive disadvantage compared to the large banks. This has implications for the evolving structure of the Hong Kong banking sector and for the HKMA’s supervisory stance.

4 1.2 Summary of the main trends and issues 14. The main trends and issues likely to affect the Hong Kong banking sector over the period covered by this review are listed below and are discussed in more detail in the rest of this section: • Business integration with the Mainland • Impact of the sub-prime crisis • Development of Hong Kong as an international financial centre • Increasing competition • Erosion of financial boundaries Business integration with the Mainland 15. Interviewees were unanimous that the accelerating economic integration with the Mainland will be the main factor shaping the Hong Kong banking sector as it moves forward. This is also the view of the HKMA itself. 16. The growing links between the banking systems in the Mainland and Hong Kong manifest themselves in various ways: • Regulatory changes in the Mainland have made it easier for Hong Kong banks to set up and expand operations on the Mainland and to do this via locally incorporated subsidiaries rather than branches, thus enabling the full range of retail banking business to be conducted. • At the same time Mainland banks are increasing their presence in Hong Kong, including through acquisition of local Hong Kong banks. It is likely that commercial banks from the Mainland, other than the Big Four, will seek to set up branches or make acquisitions in Hong Kong as they begin to internationalise their business (out of the 31 Mainland banks that appear in the world’s Top 1000, only 6 currently have branches in Hong Kong). • Large amounts of capital are flowing from the Mainland into Hong Kong in search of investment opportunities. This is boosting liquidity in the Hong Kong banking system. • A RMB market is developing in Hong Kong, providing a testing ground for convertibility of the RMB. RMB deposits in Hong Kong grew by 43% in 2007 albeit starting from a very low base.

5 17. These trends are providing new business opportunities for Hong Kong banks, but are also leading to increased risks. In particular, the expansion of Hong Kong banks into the Mainland will enable them to break out of the narrow confines of their domestic market in which they are experiencing ever-intensifying competition. A number of banks have ambitious plans for their business on the Mainland and envisage that in 5-10 years the size of their operations there could rival or even exceed their operations in Hong Kong. If realised, this would involve a fundamental change in the structure of some Hong Kong banks whose presence outside the SAR has hitherto been modest. One interviewee commented that Hong Kong banks may effectively become Mainland banks over time. If something were to go seriously wrong in the operations on the Mainland, the survival of the bank as a whole could be jeopardised. 18. The key issue for these banks and for the HKMA is how the risks in the Mainland operations (and from increased cross-border exposure) will be monitored and controlled. A number of interviewees noted that the business environment in the Mainland is significantly different and it cannot be assumed that the risk management systems that work in Hong Kong will work there. The risk issues will be compounded if banks try to grow too fast and/or pursue inappropriate business plans in their Mainland operations, e.g. an undue focus on local small and medium enterprises. 19. A further potentially complicating factor is that the Mainland authorities naturally wish foreign-owned banking subsidiaries to be as free-standing as possible in terms of their systems of internal controls, risk management, finance and accounting and IT. At present this seems to be more of a cost issue (through the need to establish separate systems in the subsidiary) rather than a control issue, but the HKMA will need to be alert to any signs that it is hindering parental oversight. 20. Accordingly, it is recommended that: • The HKMA should ensure that Hong Kong banks have appropriate oversight and control of their business operations on the Mainland. • The HKMA should have a full understanding of the business models that Hong Kong banks are pursuing on the Mainland and of the type and volume of business being conducted there and the risk controls in place. It should be prepared to challenge business models that it considers inappropriate.

6 • The HKMA should have a full knowledge of the business and regulatory environment on the Mainland, so as to be able to understand the risks and to offer supervisory guidance to AIs. • The HKMA should seek to strengthen communication and cooperation with the China Banking Regulatory Commission (CBRC) at all levels (at municipal and provincial levels as well as national), and to cement this through help with training and two-way secondments. • While relying principally on the direct supervision of the CBRC of Hong Kong banks’ subsidiaries on the Mainland, the HKMA should be prepared to conduct its own periodic on-site examinations of these (subject to any necessary approval from the CBRC). • Part of the HKMA’s cooperation with the CBRC should be to try to harmonise regulatory policies and practices to produce a level playing field on both sides of the border and to reduce the risk of regulatory arbitrage. 21. The HKMA is already aware of these issues and the above comments should therefore be seen as an endorsement of its current approach, with the added stipulation that it should ensure that sufficient resources are made available to ensure effective implementation. 22. The cross-border expansion will not be one way. As noted, it is likely that a wider range of Mainland banks will wish to set up operations in Hong Kong. These banks may be unfamiliar with operating in a fully competitive banking market and they may inadvertently bring with them practices that do not accord with Hong Kong’s international standards. In the light of this: • The HKMA should liaise closely with the CBRC to ensure that Mainland banks which wish to come to Hong Kong are fit to do so. • The HKMA should attach a high priority to the supervision of Mainland banks once they have set up in Hong Kong. 23. The issue of whether the HKMA’s authorization criteria are appropriate to cater for the admission of Mainland and foreign banks is considered in Section 6 below. 24. Substantial amounts of funds are flowing into Hong Kong from the Mainland for investment in the property and stock markets. This is creating wealth management opportunities for Hong Kong banks as well as increasing the amount of excess funds in the system. Part of this money is entering the

7 Hong Kong system through regulated channels (such as the QDII scheme), but some is also believed to be entering through unregulated channels (such as underground banks). This has money laundering, and hence reputational, implications and it is therefore recommended that: • The HKMA should continue in its efforts to encourage the Mainland authorities to establish new, and expand existing, regulated channels to facilitate the mobility of capital from the Mainland to Hong Kong or through Hong Kong to the rest of the world. • The HKMA should ensure that Hong Kong AIs exercise adequate customer due diligence and transaction monitoring in respect of non-resident customers, including those from the Mainland. Impact of the sub-prime crisis 25. This is the most immediate issue confronting financial institutions and supervisors around the world; and the lessons to be learned from it are likely to dominate the regulatory and supervisory agenda for the next few years. This is something that the HKMA will have to take into account in setting its own priorities in the short to medium term. 26. The origins of the sub-prime crisis lie in lax underwriting standards in the US mortgage market and the use of securitisation and complex financial instruments like CDOs, which compounded the problem by spreading losses, and uncertainty about the extent of these, around the global financial system. The Hong Kong banking sector has so far been relatively unscathed although some banks made relatively large investments in assets related to sub-prime and had to make significant write-downs of these in 2007. 27. The full effects of the crisis are not yet apparent, including the extent to which credit problems will spread to other asset classes. Hong Kong banks could still experience shock waves from the effects of the US economic slowdown and the problems being experienced by major financial institutions. 28. The crisis has highlighted the risks in complex financial instruments and the fact that these risks were not fully understood not only by investors but also by those manufacturing and selling the products. It has called into question the quality of risk management even in some of the largest financial institutions and the use of credit ratings by investors as a substitute for own judgement in investment decisions. Also discredited is the originate and distribute business model which acted as a disincentive to proper credit assessment and ongoing monitoring, and made it more difficult to ascertain where the risk in the system lies. The resultant credit crunch has demonstrated the importance of liquidity management in periods of prolonged market stress and the need

8 for proper valuation and disclosure. 29. The sub-prime crisis is a wake-up call to Hong Kong banks on how they can be exposed to volatility in financial markets, the growing complexity of banking products and the expanding interconnectedness of financial institutions. It also reiterates the importance of the housing market to the financial system and the broader economy. There is a view that lessons will be learned from this episode and that things will never be the same again, i.e. there will be a return to a more conservative style of banking and to simpler products. It would however be unwise to rely on this being more than a temporary phenomenon. History shows that financial crises tend to repeat themselves, albeit in different form, and banks are likely to recover their appetite for risk once balance sheets have been rebuilt. Despite the special features of the current crisis (sub-prime mortgages, CDOs, securitisation etc), the recurrent problems of build-up of financial imbalances, excessive leverage and imprudent lending lie at the heart of the crisis, and these will not go away. 30. Although Hong Kong banks seem to have largely escaped this time, they will be at risk of involvement in the next crisis as Hong Kong grows as an international financial centre (see next section) and its banks become more intertwined with international trends and practices. 31. From the HKMA’s perspective, this will increase the need to enhance its supervision of AIs’ involvement in complex instruments and their management of interest rate, liquidity, market and operational risks. Some of the lessons to be drawn are discussed in later sections of this report. At this stage, the main priorities to be highlighted are that: • The HKMA should develop its capability to look ahead, to the extent possible, to detect global trends and pick up early warning signs of the risks that arise from these trends, even when these seem improbable. • Where major risks are identified, steps should be taken to bring these to the attention of the boards of AIs (e.g. by writing to them) and positive confirmation should be sought that these risks are being addressed. • The HKMA should continue to insist that AIs apply prudent criteria in their residential mortgage lending, including adherence to the 70% loan to value ratio. 32. The issue of macro-prudential surveillance is discussed in more detail in Section 4 below.

9 Development of Hong Kong as an international financial centre 33. Interviewees attach great importance to the development of Hong Kong as an international financial centre because of the business opportunities it will create, and they believe that the HKMA has an important role to play in achieving this goal. The HKMA has helped to develop a financial development strategy for Hong Kong which focuses on positioning Hong Kong as the international financial centre for China by developing a “mutually-assisting, complementary and inter-active relationship” between the financial systems of the Mainland and Hong Kong. In practice, this means taking initiatives in five areas: • Expanding the presence of Hong Kong financial institutions on the Mainland. • Increasing the outward mobility of the Mainland’s investors to Hong Kong. • Allowing Hong Kong financial instruments to be traded on the Mainland. • Strengthening Hong Kong’s capability to handle financial transactions denominated in RMB. • Strengthening financial infrastructural links between Hong Kong and the Mainland. 34. Hong Kong’s success as an international financial centre is shown by the fact that it has retained its number three ranking in the City of London’s Global Financial Centres Index (GFCI) for March 2008, scoring well in all five areas of competitiveness: people, business environment, market access, infrastructure and general competitiveness. Regulation is one of the key areas highlighted by the latest report and it is an area of particular strength for Hong Kong. 35. In terms of business opportunities, Hong Kong’s main strength derives from its growing relationship with the Mainland and, as noted above, this underlies its development strategy. However, there are four issues which require attention. First, a number of interviewees cited a growing skills shortage in Hong Kong (notwithstanding the positive rating that Hong Kong achieved under this category in the latest GFCI report). Second, there are some gaps in the product range in Hong Kong, notably in relation to the bond market, though continuing efforts are being made to fill this. Third, as more financial activities are carried on in the Mainland, there is a risk that Hong Kong may be marginalised. (The GFCI report notes that Shanghai is among the financial

10 centres to watch for the future.) This is not however seen as an immediate threat by interviewees while the RMB is not fully convertible and provided that Hong Kong maintains its regulatory standards, adherence to the rule of law and high quality of corporate governance. Fourth, there is a view that, while the Mainland is of vital importance, Hong Kong should not be too China-centric and should take care to foster its wider international and regional links. This reflects the fact that part of Hong Kong’s role is to intermediate between China and the rest of the world and also between different countries in the region. Reflecting this, while the growth of the RMB market will undoubtedly be an important factor in Hong Kong’s future development as a financial centre, the need to foster the growth of US$-denominated assets should also not be neglected. 36. Some of these issues are not solely a matter for the HKMA, but also fall to other regulators and the Government to address – for example, establishing whether there is a major skills shortage in Hong Kong and if so what should be done about it. However, other actions are clearly the responsibility of the HKMA, in particular: • The HKMA should ensure that it continues to provide a regulatory regime that maintains effective standards while at the same time facilitating access to new entrants. • The HKMA should continue its efforts to develop the bond market in Hong Kong (e.g. through facilitating the issue of RMB bonds). • The HKMA should pursue the development of Hong Kong as a centre for Islamic bonds and other forms of Islamic finance as a way of attracting investors from the Middle East and widening the range of products in Hong Kong. • The HKMA should continue to develop Hong Kong as a regional payment and settlement hub. • The HKMA should continue to participate in international fora as a means of raising Hong Kong’s international profile and influencing the development of international regulatory standards. Increasing competition 37. Competition in the banking sector has intensified in recent years. The deregulation in the wake of the KPMG review has no doubt been a factor in this as have initiatives such as the introduction of positive data sharing which has allowed new players to enter the market for personal loans and has encouraged a shift from revolving credit to instalment loans. The effect of

11 this has been to lower the cost of borrowing and increase choice for consumers – something that positive data sharing was designed to achieve in addition to enabling better credit assessment. 38. So far this has been accomplished without any adverse effect on banking stability which might arise for example if increased competition were to lead to a general relaxation in credit underwriting standards. But the further decline in the margin on residential mortgage loans to 300bp below prime rate is putting pressure on the profitability of the smaller banks (which as noted earlier lags behind that of the larger banks) and increasing their interest rate risk since their liabilities are linked to HIBOR to a greater extent than for the larger banks who benefit from their greater access to low cost current and savings accounts. The smaller banks are therefore vulnerable to an upward spike in interest rates. There are concerns among the smaller banks that their competitive disadvantage will be further increased because the larger banks will benefit from a lower capital weighting for mortgage loans through use of the IRB approach under Basel II. It has also been suggested that Basel II may create a disincentive on the part of the larger banks to lend to the smaller banks (including recycling funds back to them in an IPO) because of the higher capital weighting that this will attract. 39. It should be noted that not all recent developments have worked to the disadvantage of the smaller banks. The Deposit Protection Scheme (DPS) has to some extent improved their ability to attract deposits and the greater ability to expand into the Mainland has created an additional business opportunity albeit this requires extra investment and creates new types of risk. Notwithstanding this, the smaller banks are feeling under pressure, and during the interviews they mentioned a number of ways in which they thought that the HKMA could assist them – for example, by promoting better recycling of funds during IPOs; by encouraging wider sharing of default data (including by the larger banks) to enable better calculation of default probabilities and facilitate a move by the smaller banks to IRB; by establishing a team of specialists who could provide guidance to those banks seeking to enter the Mainland market; by taking a leadership role in promoting the training of local staff (e.g. in compliance skills); and by not imposing requirements on best practice which are inappropriate to the size of the bank. 40. From the HKMA’s perspective, it is desirable to have a range of banks of different sizes within the banking system since this can help to serve a range of customer needs – for example, if larger banks wish to close a branch in a particular area, a smaller bank may decide to fill the gap. This is not to say however that the current size of the smaller banks is optimal and that further consolidation of the banking sector would not be desirable. Mergers and acquisitions have occurred over the years and are expected to continue, but the pace of change has generally been slow. While the profitability of the remaining smaller banks has been relatively modest, it has not created strong

12 pressure on family owners to sell or be acquired. 41. The question for the HKMA has been, and remains, the extent to which it should try to force the pace of change. However, generally speaking, the use of the Monetary Authority’s (MA) powers under the BO is linked to prudential concerns, such as a clear threat to the interests of depositors, and in the absence of such clear concerns it would be difficult to justify any action by the MA in this area. This suggests that the HKMA’s general policy should continue to be to promote the case for more consolidation through discussions with the banks concerned and through appropriate public comment. It should of course provide the necessary technical assistance where a merger or acquisition occurs. 42. This approach is based on the view that the HKMA should not try to engineer a pre-determined structure for the Hong Kong banking system. Rather: • The HKMA should take steps to encourage AIs to adopt a more balanced funding structure, e.g. through its efforts to promote the development of the bond market. • The HKMA should monitor that increased competition does not encourage increased risk-taking (e.g. through investment in complex instruments) that is not justified in relation to capital or earnings. • The HKMA should continue its efforts to encourage improvements in the risk management of the smaller banks, ensuring in particular that they fully appreciate the interest rate risk posed by the currency board system. • The HKMA should be prepared to back its judgement and take decisive action where it believes that a business model is unsustainable or risk management is dangerously weak (to avoid a repetition of the Northern Rock situation in the UK). • Such decisive action should include pressing for merger or acquisition and taking whatever action is possible to facilitate this (again bearing in mind the Northern Rock experience). • The HKMA should consider the extent to which capital requirements are providing a disincentive to the recycling of funds to the local banks during IPOs and in effect forcing liquidity offshore.

13 Erosion of financial boundaries 43. Wealth management and insurance business are becoming increasingly important drivers of retail banking business in Hong Kong, particularly in view of the decline in spreads on traditional lending business. Banks are becoming financial supermarkets. This is likely to increase as the Mainland’s capital account is liberalised and outward portfolio investment expands. While banks’ profits will benefit, the risk of misselling will also increase along with the risk of loss to investors through acquisition of unsuitable products or through market volatility. This raises questions about the capability of the banks to undertake such business and the appropriate role of the HKMA in supervising it. 44. Hong Kong presently operates a system of functional regulation with separate bodies responsible for supervising banking, securities and insurance business. Where securities and insurance business is undertaken in separate subsidiaries of AIs, responsibility for supervising these will rest with the relevant functional regulator, with the HKMA generally acting as lead regulator for the group as a whole. However, securities and insurance business is also undertaken within AIs themselves and this raises the question of which body should be responsible for supervising this business. This has been a particular issue in relation to securities business because of the amount of such business undertaken through banks’ branch networks or in wealth management units. The solution that was arrived at when the Securities and Futures Ordinance (SFO) was introduced was that the HKMA should act as the front-line supervisor of AIs’ securities business while the Securities and Futures Commission (SFC) retained responsibilities for standard-setting, licensing, investigation and enforcement (in liaison with the HKMA). The respective roles of the HKMA and the SFC are set out in the BO and the SFO and are amplified in a Memorandum of Understanding. This is reinforced by regular dialogue and bilateral meetings between the two bodies. 45. There were various reasons for putting the current arrangements in place: • To relieve the concerns of the banks which were being brought within the full scope of the securities regime for the first time (previously they had enjoyed exempt status). • As part of this, to provide so far as possible a single point of contact for AIs in relation to their securities business and to avoid duplication. • To enable the HKMA to have direct supervisory oversight of all AIs’ business.

14 • To achieve economies in the use of supervisory resources by enabling the SFC to rely on the HKMA’s supervision of AIs. 46. My discussions with the SFC and the HKMA indicated that they believe that the system is workable but there are areas for improvement, which are discussed below. From the industry’s point of view, it does not seem to have produced a totally satisfactory solution. The perceived problems of having to deal with multiple regulators was mentioned by a number of interviewees, and in its written submission HKAB noted that its members had observed that “sometimes the efforts of the HKMA and the SFC seem to have been duplicated, e.g. requests for records and other documentation and investigation enquiries”. It suggests that “some thought should be given as to how the investigation process can be effectively formalised and streamlined when multiple regulators are involved”. 47. There was also some criticism during interviews of the length of time that investigations can take and about a lack of transparency in the role of the HKMA’s enforcement unit. 48. The HKMA is not convinced that these criticisms are valid. It notes that duplicated requests for material have been rare: over the last three years only four cases have involved investigations by both the HKMA and the SFC. Moreover, there are time pledges for completing cases on hand and in the last 15 months the HKMA was able to complete the investigations on time in eight cases and only three other cases extended beyond the time pledged due to their complexity. Having said this, the HKMA is making changes to its procedural manual to shorten the time that investigations can take; and it is receptive to raising the transparency of its role in disciplinary cases and of the procedures that it follows, for example, by means of a guideline or article, as well as providing periodic updates on its activities and a summary of cases as a way of sharing current findings and issues with the industry. 49. A further issue is that of regulatory neutrality. There is a perception on the part of the brokers that AIs get softer treatment from the HKMA than they would from the SFC. However, it was the view of some interviewees that on any securities-related matter involving AIs the HKMA is particularly strict. From my own observations, the HKMA takes its responsibilities for supervising AIs’ securities business seriously and performs them diligently and fairly. Strengthening supervision of AIs’ securities-related business to ensure a level playing field between AIs and other intermediaries is in fact a key initiative in the HKMA’s current three-year plan. 50. Such activities do however use significant resources and also raise the question as to the extent to which the HKMA should be involved in regulation of conduct of business given that it is primarily a prudential regulator, i.e. concerned with the financial safety and soundness of AIs. As regards the

15 latter point, while it is true that the two types of regulation may require different styles, there is in practice quite a significant degree of overlap. Both types of regulator have a concern with corporate governance, the fitness and propriety of senior management and the quality of risk management, compliance and internal controls. These issues are of relevance across all the activities of AIs. In assessing the overall quality of systems and controls, it would be difficult for a banking regulator to form a judgement that did not embrace all activities carried on by an AI, particularly when investment-related activities are intertwined with conventional banking business (e.g. in products such as structured deposits, leveraged forex or share margin financing). As a banking regulator, the HKMA is in any case concerned with the conduct of banking business, as shown by its sponsorship of the Code of Banking Practice, and with the general issue of reputation risk which may arise from both banking and securities business. Reputation risk (including that from misselling) is something which the HKMA assesses in setting the minimum capital ratios of AIs, showing how conduct of business issues can feed into prudential issues. 51. There does not therefore seem to be a justification for unravelling the current arrangements (which would in any case require primary legislation) and for the HKMA to hand back the supervision of banks’ securities business to the SFC. There is no support for such a move among the firms that I interviewed and it would seem to run counter to the desire to reduce contact with multiple regulators since it would increase the direct interface between AIs and the SFC, while the HKMA would no doubt wish to retain an interest of its own in relation to wealth management issues. 52. The priority therefore is to enhance the present system. The steps that could be taken to achieve this are as follows: • The HKMA should undertake an assessment in consultation with HKAB and other interested parties of whether the industry’s perceived concerns about the operation of multiple regulators are indeed valid, and if so how the current system can be enhanced. • The HKMA should consider with the SFC how the process of consultation between the two bodies on the use of enforcement measures might be streamlined. This should recognise that, while the views of the consulted party should be taken into account, it will not always be possible to achieve consensus on the measures to be taken. • The HKMA should pursue its plan to raise its transparency in relation to its role in disciplinary cases and the procedures that it follows, and to provide periodic updates and feedback to the industry on its activities in this area.

16 • More staff secondments should be undertaken between the HKMA and the SFC to provide training and to promote harmonisation of culture and practices. 53. There is a broader issue relating to reform of the regulatory structure in Hong Kong that goes beyond the remit of this review, but should nonetheless be mentioned for sake of completeness. This relates to whether Hong Kong should introduce a unified or “mega” regulator that would be responsible for supervising the whole range of financial services provided in Hong Kong. This would reduce the need for AIs to have to deal with multiple regulators and would enable a more consistent regulatory view to be taken of the whole of the finance industry. 54. However, it is not clear that these advantages, if they could be realised in practice, would be sufficient to justify a radical change from the current system which performed well during the Asian Financial Crisis. The changes in regulatory structures in other countries in the Asian region following the Crisis were motivated at least in part by the fact that the previous arrangements had been discredited. This has not been the case in Hong Kong. While improvements can be made, strong justification would be needed before making a fundamental change which would incur significant transition costs and would run the risk of diverting attention away from the task of ongoing supervision. 55. There is also the non-trivial issue of which model of the unified regulator would be appropriate for Hong Kong, and its relationship with the monetary policy functions of the HKMA. Clearly, no model is perfect. The FSA in the UK has often been held up internationally as an example to be followed, but questions have been raised in the wake of Northern Rock as to whether it makes sense to split off responsibility for the financial stability of the system as a whole and for lender of last resort (which reside in the Bank of England) from responsibility for the stability of individual banks (which resides in the FSA). This is a particularly relevant issue for Hong Kong where monetary stability and banking stability are intimately linked through the currency board system with its implications for interest rate volatility. (That indeed was the rationale for setting up the HKMA in the first place.) 56. Changes to the overall regulatory structure do not therefore seem to be an immediate priority for Hong Kong. Nonetheless, the authorities should continue to follow the international debate on this issue, including the progress of the latest proposals for regulatory reform in the US.

17 2. GOVERNANCE ISSUES RELEVANT TO BANKING STABILITY The role of EFAC 57. The HKMA has an unconventional governance structure in that the powers in the BO and in the Exchange Fund Ordinance are vested in a single person, the MA. The MA is the chief executive of the HKMA, but the board structure to which a chief executive would normally be accountable does not exist because the HKMA is not a corporate body. However, the vacuum that would otherwise exist has been filled by extending the role of the EFAC so that it acts as the board of the HKMA for governance purposes. In this role, EFAC is assisted by a number of sub-committees dealing with governance, audit, the currency board, Exchange Fund investments and financial infrastructure. 58. The Governance Sub-committee monitors the performance of the HKMA and makes recommendations on remuneration and human resources policies, and on budgetary, administrative and governance issues. The Sub-committee also receives regular reports on the work of the HKMA, including in relation to its banking stability function. Other checks and balances 59. Within the HKMA itself, there are a number of arrangements to try to ensure the proper exercise of the MA’s powers. In particular, the Banking Supervision Review Committee has been established to consider the use of major powers under the BO and to make recommendations to the MA on the use of those powers. It comprises members of the HKMA staff who do not have direct dealings with the case concerned. Other committees within the HKMA have recently been established to deal with particular regulatory issues such as the setting of minimum capital ratios, supervisory ratings and the exercise of disciplinary powers in relation to securities business.

60. Additional restraint on the use of the MA’s powers is provided by the fact that in a number of cases (e.g. power to revoke and use of section 52 powers) these can only be exercised after consultation with the Financial Secretary (FS), albeit that the final decision continues to rest with the MA. The role of the Advisory Committees
61. The Banking Advisory Committee and the Deposit-taking Companies Advisory Committee can also be viewed as part of the system of checks and balances. The Committees are established under sections 4 and 5 of the BO for the purpose of advising the Chief Executive of the Hong Kong SAR upon any matter connected with the Ordinance and advising the Chief Executive in Council in any case where the advice of the Committees is sought under section 53(2). (This relates to the possible use of the Chief Executive in

18 Council’s powers of intervention under section 53 of the Ordinance.) 62. The normal routine work of the Committees is to meet on a quarterly basis to provide the views of members on current market developments and trends and to comment on policy initiatives by the HKMA. Since the agenda of the two Committees is normally the same, the practice has been adopted of holding joint meetings of the Committees. This has the effect of increasing the number of participants at each joint meeting. 63. Interviewees who are past or present members of either of the Advisory Committees were of the view that the current arrangements are not working well. The role of the Committees is no longer clear and meetings have become something of a formality. While some see value in joint meetings since it enables a greater range of views to be heard, the number of participants is considered too large for meaningful discussion. As a means of industry liaison, more value is derived from the HKMA’s industry working groups on issues such as anti-money laundering and consumer credit and the regular breakfast meetings that the HKMA holds with the smaller local banks. 64. Suggestions for improvement include: splitting the joint meetings back into two separate ones (though there is no unanimity on this); broadening the range of the participants (fewer bankers and more outsiders); use of sub-committees to deal with particular topics; and a greater focus on local as opposed to international issues. 65. Ultimately, the success of such committees depends on the ability and willingness of members to make their views known. There may be a better chance of achieving this in a smaller meeting. Moreover, there may be advantage in shifting the focus of the Committees so that their role is one of advising the HKMA on policy issues (though not in relation to the use of the MA’s powers in relation to individual institutions), while at the same time distancing the Committees to a greater extent from the Government. This would clarify the role of the Committees, enhance the operational autonomy of the MA and help to ensure that policy decisions are taken within an appropriate framework of advice. To this end it is recommended that:
The effectiveness of the Banking Advisory Committee and Deposit-taking Companies Advisory Committee should be strengthened. Pending legislative change, the two Committees should continue to meet jointly, but efforts should be made to reduce the size of the membership of each Committee to enable joint meetings of more manageable proportions to be held.
Unless there is a particular issue that requires his presence, the FS should normally delegate his responsibility for chairing the joint meeting to the MA.

19
In the longer term, the BO should be changed to turn the two Committees into one Advisory Committee whose role is to advise the MA on matters of policy in relation to his regulatory functions. The new Committee should be chaired by the MA with a maximum of say 12 external members (including an appropriate representation of non-bankers). Appeal mechanism 66. The arrangements for appeal against regulatory decisions constitute one of the checks and balances that apply to such decisions since they allow for quasi-judicial oversight. In the case of the BO however the standard appeal in relation to banking matters is to the Chief Executive in Council rather than to an independent tribunal. This no longer seems tenable particularly since an appeal in relation to securities-related matters is to the Securities and Futures Appeals Tribunal established under the SFO. Moreover, a new Review Tribunal, which is quasi-judicial in nature, has been established to hear appeals against decisions by the MA in relation to the calculation of the capital adequacy ratio of individual AIs. In the light of this, it is recommended that: • The current right of appeal to the Chief Executive in Council in the BO should be replaced with a right of appeal to an independent Appeals Tribunal. (This might be done by enlarging the scope of the current Review Tribunal to turn it into a Banking Appeals Tribunal.) The role of the Hong Kong Association of Banks (HKAB) 67. The HKMA relies on HKAB to assist in the development of policy, provide advice and represent the views of the banking industry. Although it is arguably incidental to this review, it is worth noting that a number of interviewees considered that HKAB is not fulfilling its full potential. One point of view ascribes this to the fact that HKAB has had to become more “democratic”. Others think that it is not democratic enough and that the smaller banks should have the opportunity to play more of a role, e.g. by being able to take the chair. It is difficult however to know how serious this feeling is, and whether any of the smaller banks would have the resources to play a more active role under the current arrangements. As Hong Kong expands as an international financial centre and consumer issues come increasingly to the fore, the need for strong industry representation will increase. It is difficult to see how this can be delivered under the current arrangements of a rotating chairman and secretariat. It is therefore recommended that:

20 • The HKMA should suggest to HKAB that it conducts a strategic review of its role, functions and organisational structure, with particular reference to whether it should appoint a professional chief executive and a permanent secretariat. Strategic planning 68. The HKMA currently produces a rolling three-year business plan for each of its main policy areas, including banking stability. The plan sets out the key trends and challenges which will influence attainment of the HKMA’s objectives for each policy area, specifies what these objectives are and lists the initiatives planned to achieve the objectives. The plan forms a basis for working out resources implications and drawing up the annual budget. It is considered by the EFAC Governance Sub-committee and is presented to EFAC itself for endorsement in the autumn of each year. 69. While the plan is a useful means of steering the work of the HKMA and directing resources, it suffers from a number of drawbacks. In particular, there is no attempt to prioritise items in the plan itself and to distinguish between those initiatives which are essential (e.g. because they have fixed deadlines or are necessary to deal with some major issue) and those which are desirable. Some of the initiatives lack specificity and are more in the nature of aspirations (although this is to some extent unavoidable in a three-year plan). This makes it difficult to monitor performance against plan. It would assist in this respect if the plan were to distinguish between those items which are “projects”, i.e. discrete initiatives with a particular aim and a beginning and an end, and those which are “business as usual”, i.e. part of the ongoing work of the area concerned (e.g. monitoring of AIs’ Mainland exposures). Items should also be allocated to specific years in the plan with details of the projected completion dates for the various milestones. A further refinement would be to try to estimate the resources implications of the various items in terms of number of man-days required. 70. There are dangers in making strategic planning too elaborate if this means that the process of preparing the plan impedes the actual execution of policy initiatives or adds undue rigidities to the HKMA’s work plan. Nonetheless, it is recommended that: • The HKMA should refine its business plan to make it more specific and make it easier to monitor implementation of the various initiatives and to assess the resource implications.

21 3. THE MA’S FUNCTIONS AND POWERS IN THE BANKING ORDINANCE 71. As noted earlier the functions and powers set out in the BO are vested personally in the MA and executed through his office (the HKMA). The Ordinance was originally enacted in 1986, though it has been subject to regular amendments since then. At some point over the next ten years it may be desirable to conduct a thorough review of the Ordinance with a view to rewriting it from scratch. But the Ordinance as it is works reasonably well, and so long as it remains relatively easy to make amendments, there is not a pressing need for a completely new Ordinance. The functions of the MA 72. The long title of the BO states, inter alia, that it is intended to make provision for the supervision of AIs so as to provide “a measure of protection for depositors”. However, section 7(1) makes it clear that the principal function of the MA shall be to “promote the general stability and effective working of the banking system”. Thus, the MA’s role is mainly systemic in nature from two perspectives – one concerned with the stability of the banking system and the other with how effectively it is performing its functions (e.g. taking deposits, money transmission and financial intermediation). 73. Section 7(2) of the Ordinance goes on to list the matters that the MA should involve himself in. Some items relate to the business activities of AIs. For example, the MA is to: • Take all reasonable steps to ensure that the principal places of business [etc] …..are operated in a responsible, honest and business-like manner. • Promote and encourage proper standards of conduct and sound and prudent business practice amongst AIs. • Suppress or aid in suppressing illegal, dishonourable or improper practices in relation to the business practices of AIs. • Take all reasonable steps to ensure that any banking business, any business of taking deposits, or any other business carried on by an AI, is carried on with integrity, prudence and the appropriate degree of professional competence.

22 74. These specific functions elaborate on how the MA’s functions under section 7(1) may be discharged and are therefore integral, rather than incidental, to the HKMA’s overall responsibilities set out in that section. This suggests that the remit of the HKMA is quite broad. While the HKMA is concerned mainly with prudential issues (safety and soundness of AIs and the quality of their risk management and internal controls), it also has a role in overseeing the more general way in which AIs conduct their business as a whole and the ethical standards that they follow. This reinforces the point that the dividing line between prudential and conduct of business issues is not always clear-cut and that both are relevant to the stability and effective working of the banking system. If the banking industry consistently mistreats its customers, it runs the risk of loss of confidence and can hardly be said to be working effectively. The HKMA’s role in consumer protection 75. The above is relevant in particular to the HKMA’s role in consumer protection. The MA has no specific statutory responsibility for consumer protection set out in the BO; and the HKMA (correctly) regards its main responsibility under the BO as being to ensure that AIs are financially sound and prudently managed. An undue focus on consumer issues can result in a loss of focus on the ultimate objective of consumer protection, which is to reduce the risk that banks fail. This seems to have been a factor in the supervisory failings in relation to Northern Rock and the FSA has announced that it will shift its priorities more in the direction of prudential issues. However, there is a balance to be struck and, as noted above, the scope of the MA’s responsibilities under section 7 of the BO for the overall stability and effective working of the banking system and for the manner in which AIs conduct their business would appear to extend to certain consumer and counterparty issues. Deception, sharp practices, deliberately misleading advertisements etc can be said to damage the industry as well as the individual institution. 76. In practice, the HKMA is already engaged in a number of consumer protection activities: • Receiving complaints about banking services and ensuring that AIs have effective systems for dealing with such complaints. • Promotion of good banking practices through, inter alia, its endorsement and enforcement of the Code of Banking Practice. • Addressing ad hoc public concerns about consumer issues in the banking sector (e.g. ATM frauds, closure of bank branches, basic banking services). • Supervision of AIs’ wealth management business.

23 • Providing security tips to consumers on internet banking. 77. Such activities appear to be legitimate for the HKMA to undertake and consistent with the MA’s functions set out in section 7 of the Ordinance. The question is whether the HKMA should be doing more. 78. Views among interviewees on this issue were mixed. The Consumer Council would like the MA to be given a specific statutory duty for consumer protection (as is the case with the UK FSA), which would be overseen by a Consumer Panel, and for more of the HKMA’s resources to be devoted to the issue. The majority view among other interviewees was however that the HKMA is doing about as much as it should in this area, and that there were risks that taking on more responsibility could result in an overload of issues that are not central to the MA’s functions, leading to the loss of focus mentioned above. 79. In drawing the appropriate line, the HKMA’s main attention should be systemic in nature: focussing on behaviour by AIs that could have a bearing on confidence in the banking system or its effective working. It is legitimate for example for the HKMA to be concerned about branch closures if the result is that parts of the community lose convenient access to banking services. But the remedy in this instance would not be to prevent the bank concerned from closing its branches, but to discuss with the industry how mitigating measures might be implemented, such as achieving interoperability of the two ATM networks in Hong Kong and making it easier to switch bank accounts. Where possible, initiatives in this area should emanate from the industry itself rather than from the HKMA though the latter may be able to play a facilitating role. 80. In considering what more might be done, it is clear that the HKMA should not be involved in arbitrating in disputes between AIs and their customers and in awarding compensation. That is the role of an ombudsman and as consumers access a broader range of financial services in Hong Kong it may be sensible at some point to establish such a body as Singapore has recently done. The role of the HKMA in relation to complaints should be to pursue those complaints which have supervisory implications and to deal with those which reveal improper conduct, such as deliberate or reckless misselling, and which if left unchecked could bring the system as a whole into disrepute. 81. This raises the question of whether the MA has a sufficient range of powers to deal with cases of misconduct that are uncovered. The HKMA does have a wide range of supervisory measures that it can take, but some of these (such as removal of management or revocation of a licence) are not necessarily suitable to deal with consumer issues. A broader range of sanctions may therefore be called for. In the case of securities-related business, recourse can be had to

24 the sanctions available to the SFC, but there are other wealth management products that do not fall within the SFC regime (e.g. interest-linked or forex-linked products), and for these it may be appropriate for the MA to be given similar sanction powers to those of the SFC. This is dealt with in more detail below. 82. The final point is whether it is necessary for the MA to be given an explicit statutory role in relation to consumer protection. This does not seem to be necessary given the broad provisions of section 7 of the Ordinance. There could also be drawbacks in giving the MA a specific statutory duty in this respect, since in the absence of a financial services ombudsman, it could set up expectations, in relation to dispute resolution, that the MA would not be able to fulfil. It could moreover lead to an undue diversion of resources away from prudential issues to consumer issues such as seems to have occurred in the UK FSA. 83. In summary, it is recommended that: • The HKMA should maintain its primary focus on prudential issues but should also remain involved in consumer protection issues that fall within its current statutory remit to promote the general stability and effective working of the banking system. It does not appear necessary for the MA to be given a specific statutory responsibility for consumer protection. • There may be advantage in the HKMA preparing a formal guideline under section 7(3) of the BO setting out how it considers it should exercise this aspect of the MA’s functions and what the limits on it are. • The HKMA should continue not to be involved in resolution of individual disputes between banks and their customers. That is the role of a banking (or more generally a financial services) ombudsman for which there may at some point be a need in Hong Kong, particularly as wealth management activities expand. Supervision of remittance agents and money changers (RAMCs) 84. RAMCs provide remittance and money exchange services. Under the Organised and Serious Crimes Ordinance they are required to register with the police and to identify customers and keep records of transactions of HK$8,000 or above and to report suspicious transactions. However, while they are liable to investigation and enforcement action when non-compliance is suspected, they are not subject to authorization on the basis of fit and proper criteria or to ongoing supervision and monitoring in relation to anti-money laundering and counter-terrorist financing (AML/CFT) systems and controls.

25 This is contrary to FATF Recommendation 23 and was commented upon by a number of interviewees who saw it as posing a threat to Hong Kong’s reputation as a financial centre as well as creating difficulties for the AIs that provide banking facilities to these firms. 85. Different jurisdictions have used different approaches to the problem of how to accommodate authorization and supervision of RAMCs. In the UK, the responsibility rests with Customs and Excise because of its historical role in this area. In Singapore and in Jersey, the role has been taken on by the unified regulators in the two jurisdictions, reflecting the wide scope of their responsibilities for supervision of the financial sector. In Hong Kong, with its functional approach to supervision, the issue is more difficult. The HKMA considers it inappropriate to take on the role feeling that it has no experience in this area and that it would distract it from its main functions. This seems to be a legitimate concern since, from my own experience, taking on responsibility for part of the financial system that has been previously unregulated can be a time-consuming and difficult task. Accordingly, it is recommended that: • The HKMA should not assume responsibility for the AML/CFT supervision of RAMCs, but some other solution should be found. The MA’s powers of sanction 86. The MA has a range of powers under the BO to deal with prudential problems in AIs, including removal of management and directors, appointment of a Manager under section 52 and suspension or revocation of authorization. These are supplemented by the criminal sanctions that apply to breaches of specific provisions of the Ordinance. However, such powers are somewhat draconian in nature and may have undesirable, albeit unavoidable, systemic implications if used. They are also less suited to dealing with issues that may arise as AIs become involved in the provision of a broader range of financial services. A further point is that it is a recommendation of this report that there should be consideration of whether the MA should be given additional powers to make rules under the BO. This would raise the question of what sanctions would apply if these rules were to be broken. This is dealt with in Section 5 below, which discusses the policy framework. 87. The most obvious gaps in the MA’s present armoury of sanctions are that he has no power to levy financial penalties or to issue public reprimands. A number of interviewees queried whether such “intermediate” penalties are necessary. They considered that the present approach had produced a good culture of compliance in the industry. They also pointed to the difficulties of making sanctions public given the adverse impact that this could have on the AI concerned. It appears that AIs in general regard a public reprimand as a more severe penalty than a fine, which is a reversal of the normal hierarchy.

26 Both the UK FSA and the SFC regard a fine as more severe than a reprimand. This is consistent with the fact that fines would in any case be publicised and therefore would inevitably contain an element of reprimand. 88. Publicity does raise difficult issues for a banking regulator since public reaction to news that an AI has been disciplined may be unpredictable. Against this however, it has to be recognised that publicity may be precisely what gives a sanction its deterrent effect. Moreover, the issue which gives rise to the possible use of sanctions may already be in the public domain and the HKMA can be placed in a difficult situation if it is constrained by the confidentiality provisions in the BO from revealing what action it has taken to deal with the issue. Finally, it should be noted that AIs are subject to the SFC’s sanctions regime in relation to their securities-related business, which would include liability to public reprimands and fines. It would seem anomalous if they were not to be subject to similar penalties in relation to that part of their wealth management business which is not covered by the SFC regime. 89. Only the HKMA is in a position to know, from its practical experience of dealing with AIs, whether the advantages of having additional powers of sanction outweigh the possible disadvantages. The latter would include not only the possible destabilising impact of adverse publicity, but also the risk that it would lead to a more legalistic and arm’s length relationship with the banking industry. The additional resources that would need to be devoted to enforcement and to possible appeal cases would also have to be considered. However, on the basis that the HKMA is increasingly likely to have to confront issues within AIs which are not purely prudential in nature, an extension of the range of sanctions available to it would seem to be desirable. The type of situations where these might be used might include: • Late payment of fees or late submission of returns. • Other technical breaches of the Ordinance (e.g. the rules on connected lending) where the offence is not serious enough to warrant criminal prosecution. • Intentional or reckless misselling of financial products not covered by the SFC regime. • Intentional or reckless breaches of regulatory requirements (assuming that these have the required degree of specificity). • Failure to implement an agreed action plan following an HKMA examination.

27 • Failure to maintain adequate AML/CFT controls. • System or other operational failures that lead to loss or inconvenience to the public. 90. The HKMA would in all such cases need to consider whether the breach or other misconduct was serious enough to warrant it being placed in the public domain and whether the use of other sanctions or remedial measures that did not require publicity were more appropriate. Rather than a public reprimand, it may be more appropriate for the HKMA to issue a more neutral “public statement” which summarises the facts of the case, the HKMA’s findings and the remedial action taken. It could be stipulated in the BO that before issuing any such statement the HKMA would need to be of the opinion that it was in the public interest to issue the statement (e.g. for its deterrent effect and to demonstrate that regulatory standards are being upheld) and not contrary to the interests of depositors or the general stability of the system. The ability to levy financial penalties would add an additional element of being able to fine-tune the size of the penalty in relation to the seriousness of the breach. 91. Taking the above considerations into account, it is recommended that: • The HKMA should consider whether it feels the need for an extension of the MA’s powers of sanction to include the issuing of public statements and the ability to fine. If so, it should present the case to the Government and the banking industry. • Whether or not such powers are granted, the HKMA should consider issuing a guideline under section 7(3) of the BO describing the various enforcement powers available to it and the broad principles that would govern the use of such powers. The MA’s powers of investigation 92. The MA has the general power under section 55 of the BO to enter the premises of AIs and their branches, subsidiaries etc in order to examine and investigate books, accounts and transactions etc. In relation to the securities business of AIs, the MA may also authorize a person (generally a member of his staff) to conduct an investigation under section 180 of the SFO. The SFO is different from section 55 of the BO in two major respects. Section 180 has a wider range of coverage in terms of the persons to whom enquiries can be directed (e.g. a related corporation of the institution concerned or any other person who may have relevant information) and is more prescriptive in terms of the powers that can be exercised by the authorized person (e.g. the ability to make inquiries of the institution, related corporation or other relevant person regarding the records, documents or transactions that are the subject of the

28 inspection). Given that the MA has such a power under the SFO it would seem sensible to incorporate it also in the BO in relation to banking matters. Similar considerations would apply to replicating in the BO the SFC’s powers of investigation under section 182 of the SFO. Accordingly it is recommended that: • The powers of investigation of the MA under the BO should be extended so that they are no less than those available to the SFC. The MA’s powers of intervention 93. The MA already has extensive powers of intervention under Part X of the BO in relation to AIs that have failed to meet their obligations or may be about to do so. These include the power, after consultation with the FS, to issue directions, to appoint an Advisor or to appoint a Manager to take charge of the affairs, business and property of the AI. The latter appointment may be the prelude to the appointment of a provisional liquidator or may involve managing the business over the longer term to give it a chance to recover from its difficulties. The appointment of a Manager was recently tested in relation to a local deposit-taking company and while there are no doubt lessons to be learned from this experience, the process seems to have worked well. 94. The Manager is given wide powers under the Ninth Schedule of the Ordinance, including the power to sell or otherwise dispose of the business or property of the institution concerned. This would appear to allow the use of “purchase and assumption” transactions where a healthy AI (or a “bridge bank” established by the Government or the HKMA for the purpose) would purchase all or part of the assets of a failing AI and assume all or part of its liabilities. At a minimum the liabilities would include the deposits insured under the DPS, but might also include uninsured deposits if that was considered to be appropriate. The advantage of such a scheme is that if it can be implemented quickly it provides continuity of service to depositors (because they can use the AI to which their deposits have been transferred) and it avoids triggering the DPS. It would therefore provide another option for resolving a failing AI. 95. Such a scheme is currently under consideration in the UK in the wake of Northern Rock. To some extent this will involve the UK catching up with the powers that are already available to the MA under Part X of the BO. But the specific proposals that emerge in the UK and the legislation that is used to implement them may provide useful pointers to how the arrangements in Hong Kong could be further enhanced (e.g. in ensuring that the transfer of assets and liabilities of a failing bank cannot be blocked by shareholders). It is therefore recommended that: • The HKMA should follow closely the debate in the UK post-Northern Rock about how to deal with failing banks and apply any relevant

29 lessons learned to the situation in Hong Kong, including any necessary enhancements to Part X of the BO. 4. THE SUPERVISORY FRAMEWORK Summary of the current framework 96. The HKMA’s supervisory framework is built around five major components: • CAMEL rating system • Risk-based supervisory approach (RSA) • Supervisory review process (SRP) • Macro- and micro-prudential review • Stress testing programme 97. These are intended to provide an assessment of the various types of risk facing AIs, the quality of their systems for managing these risks and their resilience in the face of adverse shocks. The core of the framework is the RSA which assesses the risk profile of AIs by reference to eight types of risk: credit, market, liquidity, interest rate, operational, legal, reputation and strategic. Under each type of risk, the level of inherent risk and the quality of risk management systems for controlling that risk are assessed in order to arrive at the overall risk profile. This should then feed into the SRP, which was introduced in January 2007. The SRP specifically addresses the risks not covered by Pillar 1 of Basel II and the end-product is the minimum capital adequacy ratio (CAR) for locally incorporated AIs. The SRP in turn is used (for locally incorporated AIs) as input for the overall assessment of their safety and soundness under the CAMEL system. This is summarised in a supervisory rating (based on a 13 point scale for local AIs – a different system is used for overseas AIs) which encapsulates the HKMA’s view of the institution and is also used to set the risk-based annual premium which each bank pays to the DPS. 98. The macro-prudential review is a means of off-site surveillance which is produced half yearly and focuses on identifying risks emerging in the banking system as a whole. The quarterly micro-prudential review on the other hand addresses the risk of failure in individual AIs by focussing on those which have recorded unfavourable changes or have maintained a weaker performance relative to the industry average. This is supplemented by the stress-testing programme which monitors the financial performance (CAR and profitability) of selected AIs under various hypothetical adverse scenarios (mild, medium and severe). This can help to identify those outlier AIs that

30 are more vulnerable to stress as well as the main high risk areas (e.g. property exposure or interest rate mismatches) that give rise to that stress. Industry-level stress tests are also carried out to test the exposure of the system as a whole to adverse events such as the US sub-prime crisis. 99. The HKMA uses a variety of tools to gather information for the above assessments: on-site examinations of various types, off-site review based on data received from AIs (prudential returns, internal MIS reports etc) and from external sources, prudential meetings with the senior management of banks, tripartite meetings with AIs and their auditors, meetings with the board of directors or with board committees and use of external auditors (e.g. to prepare reports on particular issues). Overall assessment of the current approach 100. The HKMA’s current supervisory framework is comprehensive and intensive. Coupled with the HKMA’s regular dialogue with the industry, it enables it to be aware of the main risks to the system and to individual AIs. The HKMA’s supervision is generally well-regarded by the industry for its combination of effectiveness, professionalism, approachability and flexibility and by the rating agencies which rank the HKMA’s supervision highly in global let alone regional terms. However, this is not to say that there is not room for improvement and among the general praise from interviewees there was also mention of various aspects of the current system that should be addressed. This was also reflected in some of the comments made by HKMA management and staff during my discussions with them. 101. The main issues that require attention are summarised below and discussed further in the rest of this section: • The need to streamline the off-site review process • The need to rationalise the on-site examination process • Enhancement of macro-prudential surveillance • The issue of high staff turnover The need to streamline the off-site review process 102. As noted above, the off-site review process is built around the RSA, the SRP and the CAMEL rating. The main components of the RSA, as set out in module SA-1 of the Supervisory Policy Manual (SPM) are: the Institutional Overview which sets out the HKMA’s current state of knowledge of the structure and financial condition of the AI; the Risk Assessment Narrative which analyses the level of the risks facing the AI, the quality of the systems

31 used to manage these risks and the direction of the risks (i.e. are they decreasing, increasing or stable); and the risk profile matrix which for each type of risk summarises the level of inherent risk, the quality of risk management systems and the direction of risk and produces an overall rating of the AI’s risk profile (i.e. inherent risk after allowing for controls). This may be low, moderate or high. 103. As noted above, this risk assessment should feed into the SRP and CAMEL rating. The various aspects of the assessment of the AI’s financial condition, corporate governance and risk profile should then be used to direct resources to the areas of highest risk, which will be reflected in the issues that are selected for on-site examination (as set out in the scoping memorandum) and the topics that are raised for discussion in prudential interviews and tripartite meetings. The issues uncovered in the examinations, interviews and meetings then provide material for updating the risk assessment in what is an ongoing circular process. 104. In my discussions with HKMA staff and from my own observations derived from case files, the following issues emerged: • The current off-site review process is still evolving following the introduction of the SRP and there is a need to ensure that the overall process is efficient and that each of the elements within it continues to be applicable and useful. • In particular, there may be areas of duplication in the RSA, SRP and CAMEL, which would increase the workload of staff and the risk that process could get in the way of substantive risk assessment. • There is a need to ensure that the ability of case officers to fully understand the risk profile of AIs is not undermined by the wide range of routine and non-routine tasks, not directly related to the risk assessment, that they have to undertake, and that systems such as the EPSS database are designed to best support the core objective. Any gaps in the ability of staff to understand more complex instruments and risks must also be addressed. • Care must be taken to avoid any tendency for risk assessments to gravitate towards “neutral” ratings (i.e. “moderate” and “acceptable”). • The link between the assessment of areas of higher risk and the areas that are selected for on-site examination must be made clear. Steps should be taken to ensure that case officers do not adopt an insufficiently selective approach towards the preparation of scoping

32 memoranda to avoid the risk of failing to spot an issue and leaving things out. • The policy on the intensity of supervision that should be applied to the smaller institutions - notably the deposit-taking companies (DTCs) - needs to be clarified. • Related to this is the general issue of the HKMA’s risk tolerance, i.e. the extent to which it is prepared to tolerate the occasional failure. 105. Senior management in the HKMA’s supervisory function seem to be aware of these issues and they are best placed to arrive at solutions. However, in terms of eliminating duplication and streamlining the risk assessment process, one approach would be to increase the emphasis on the SRP (at least for locally incorporated AIs). This reflects the fact that it provides a more granular and precise approach towards the assessment of the level of the various types of inherent risk and the quality of risk controls. Moreover, it generates not only a minimum capital ratio but also an overall risk profile, which is then used to establish the aggregate CAMEL rating. That being so, it is not clear why CAMEL is still needed in addition to the SRP. If it is felt that SRP is not comprehensive enough (e.g. because it does not deal explicitly with items such as the quality of risk assets) and therefore needs to be supplemented with CAMEL, the answer may be to develop the SRP to address any perceived gaps in coverage (such as the types of risk exposure revealed by the sub-prime crisis), rather than running parallel systems. 106. A further reason for greater focus on the SRP is that the international regulatory efforts to improve the quality of banks’ risk management will centre on use of Pillar 2 guidance for this purpose (see the April report of the Financial Stability Forum (FSF) on Enhancing Market and Institutional Resilience), and the SRP provides a vehicle for implementing this. 107. The SRP should however be supplemented by retaining the risk profile matrix because this summarises in a neat way the relationship between each inherent risk and the related risk controls. Most of the input for this would come direct from the SRP, though it would be necessary to arrive at a composite rating for credit and operational risk and a rating for market risk. The SRP does not deal comprehensively with these risks because they are covered by the 8% minimum capital ratio under Pillar 1 of Basel II. It would also be necessary to retain a mechanism for assessing the level of risk and the quality of risk controls by main business activities as a guide to the areas that should be subject to on-site examination (i.e. the main “auditable” areas). However, this should be done without the need to write a lengthy and time-consuming risk narrative, which is the case at present. In general, written comments

33 should be kept throughout the process to the minimum required to justify the risk assessments. The FSA internal review in relation to its supervision of Northern Rock 108. It is important to bear in mind, as shown by the FSA’s internal review of Northern Rock that the way in which the risk assessment framework is actually applied by staff is a vital factor in ensuring effective supervision. In this respect the FSA’s review reveals significant shortcomings in supervisory practices including: • Lack of comprehensive financial analysis. • Failure to identify and pursue risks arising from Northern Rock’s increasing business risk profile and control framework. • Infrequent contact with the firm despite the fact that it was supposed to be under “close and continuous” supervision. • Lack of formal risk re-assessment, recording of issues in the FSA’s internal database and escalation of issues to more senior management. • Other failures to comply with established processes (for example recording of key meetings). This was made worse by the work pressures on case officers. • An inadequate level of engagement and oversight by supervisory line management (exacerbated by the fact that there was high turnover of the divisional heads responsible for Northern Rock and the fact that they had to cover gaps due to manager turnover). 109. At the more strategic level, as noted above, the FSA has also acknowledged the need to reassess the balance of resources committed to conduct of business regulation relative to prudential supervision. 110. Bearing these points in mind, it is recommended that: • The current off-site review process should be streamlined to reduce the need to run different systems side by side and to produce a logical and efficient sequence of actions leading up to on-site examination. The SRP should be revised to take account of the further Pillar 2 guidance to be issued by the Basel Committee. • Risk assessment should be an on-going process throughout the year with re-assessment of the risk profile as new information emerges.

34 There should be a clear mechanism for escalating emerging risk issues to senior management and appropriate arrangements for quality control of the assessment process. • The work pressures on case officers should be alleviated by reviewing the purpose of their various tasks, by making sure, on an ongoing basis, that the limited supervisory resources are allocated to tasks that are of a higher priority, and by giving them better tools to access the EPSS database and conduct financial analysis (including peer group review). • The ability of case officers to assess complex instruments and risks (including strategic risks) should be addressed though enhanced training in this area, through better knowledge transfer within the HKMA (e.g. from the specialised examination teams) and through increased use of short-term secondments to industry. • The current risk ratings (e.g. low, moderate and high for risks and strong, acceptable and weak for controls) should be made more granular to avoid fence-sitting by case officers and to allow more precise ratings. For example, using an approach similar to that of the UK FSA would give four ratings for risk: low, moderate-low, moderate-high and high, with the two latter ratings being the main area of focus for on-site examination. • An explicit impact rating should be assigned to each AI including DTCs. (Impact being measured in relation to the effect on depositors and systemic stability if the firm were to fail or become involved in financial scandal of some kind, e.g. related to money laundering or fraud.) In general, reduced-scope risk assessment and supervisory monitoring may be appropriate for low impact firms (although the risk profile of the firms concerned will also be a relevant factor in deciding the intensity of supervision). The HKMA should publish a guideline setting out its supervisory policy in this respect. • Related to this is the need for the HKMA to establish its tolerance in relation to the risk of failure or major scandal. This should help to establish the intensity of the HKMA’s supervision by type of firm and the resources it needs to ensure that its risk tolerance is not exceeded. The need to rationalise the on-site examination process 111. The issue of risk tolerance also applies to the HKMA’s approach to on-site examination. This has evolved in recent years with an increased use of examination teams with in-depth expertise to conduct specialised

35 examinations in areas like treasury and derivative products and also to conduct thematic examinations. The latter enable particular issues (such as controls on RMB business or wealth management activities) to be examined across a range of different AIs with a view to assessing the risk for the industry as a whole and deriving guidance on sound practice that can be fed back to industry. Other than specialised and thematic examinations, the HKMA also conducts risk-based examinations as part of the RSA process described above – these are regarded as the HKMA’s regular type of examination, although the numbers have declined in recent years as thematic examinations have expanded. 112. While the overall number of examinations has not increased (it has remained at around 245 since 2003), the shift in emphasis towards specialised and thematic examinations has put pressure on the HKMA’s resources because it is more onerous to carry out different types of examinations and coordinate these, and requires the use of staff with more specialised expertise. AIs are also feeling the strain and during interviews a number of complaints were voiced about the increase in the number and type of examinations that certain AIs are facing. A review of the examinations carried out on selected local banks in 2006 and 2007 showed that it is not unusual for them to be subject to six or seven examinations in a year, including several thematic and specialised examinations. Each of these takes several weeks to complete and requires the production of an often lengthy examination report which may take several months to emerge and whose findings then need to be digested by the AI concerned. It is not uncommon for different aspects of the same issue (e.g. AML/CFT) to be addressed in a risk-based examination and shortly before or afterwards in a specialised or thematic examination. AIs which are considered exemplars of best practice feel particularly aggrieved at being repeatedly selected for examination for benchmarking or fact-finding purposes. 113. As regards the quality of examinations and of examination reports, the attitude of industry seems to be mixed as would be expected since individual AIs will be interfacing with different examination teams, some of whom may be more knowledgeable and flexible than others. Some interviewees rate the quality of examination teams as good and find their recommendations to be of value. The more negative comments revolve around the view that some examinations are more detailed than they need to be (“top-to-bottom audits”) and that some of the points that emerge appear to be relatively minor. This is reflected in some of the examination reports which, from my observation, lack prioritisation of action points in terms of seriousness and urgency of implementation (though this will usually be addressed in the covering letter) and can be extremely lengthy. Having said this, the ability of examination teams to delve into detail and uncover issues is also a strength where the situation demands it.

36 114. HKAB has pointed out that some of its members believe that the HKMA should consider focussing on “high-level supervision” and minimizing examination of detailed day-to-day operations. Clearly the latter type of examination will sometimes be appropriate but there is a balance to be struck. Part of the problem may relate to the lack of focus in risk-based examinations already referred to and the range of issues covered by these.

115. The problem of how to deal with follow-up on action points in examination reports is also an issue cited by some HKMA staff. Often a standard item in risk-based examinations is to check to see that recommendations made in previous examinations have in fact been implemented. Such on-site examination may sometimes be necessary but the primary reliance should be on the assurance of the chief executive of the AI that the matter has been implemented with use of sanctions if it turns out not to have been. Making this a standard item in risk-based examinations uses up resources and creates the risk that unimplemented action points will continuously be rolled forward.

116. At present examination reports present a list of findings with comments of the management of the business area concerned and the action they intend to take on the basis of the findings. However, the intended action is not always clear-cut (e.g. if it is an undertaking to “review” the issue) and it is not clear whether it represents the official response of the institution, as ratified by the board and/or chief executive of the AI.

117. In the light of the above comments, the following recommendations are made: • The HKMA should make efforts to reduce the number of separate examinations to which AIs are subject in any given year. • In particular, the HKMA should re-consider the role of risk-based examinations in the light of the move to specialised and thematic examinations. Unless the risk profile of a local bank includes high or moderate-high risks or control weaknesses, consideration should be given to whether it requires a risk-based examination each year (particularly if it is being subject to other types of examination that year). • The HKMA should exercise restraint on the thematic examination of particular AIs to obtain a benchmark of best practice. Other methods should be used including discussions with, or presentations from, the AI concerned on the practices they use in the area under review. • The HKMA should review the amount of detail that examiners should normally cover, perhaps in the case of well-run AIs allowing

37 examinations to be confined to a relatively high level unless that reveals particular issues that require more detailed investigation. • The HKMA should present its examination findings in a more concise way by focussing on the high priority items and itself specifying the action points to be taken on the basis of the findings and the deadline for these to be implemented. • Follow-up should be conducted primarily by the chief executive of the AI confirming that the action points have been implemented by the deadline specified by the HKMA. Confirmation of this could be provided by the external auditors and be subject to spot checks by the HKMA. Enforcement action should be taken if it turns out that the HKMA has been misled. • More use should be made of section 59 of the BO to commission reports on AIs’ systems and controls from external auditors (only 3 such reports were commissioned in 2007 and 1 in 2006). • In general, the HKMA should consider ways in which it can leverage its use of external auditors, for example by allowing them to attend meetings with the boards of directors as observers. Enhancement of macro-prudential surveillance 118. The HKMA’s macro-prudential surveillance complements its supervision of individual AIs by focussing on system-wide risks and attempting to detect early warning signs of impending shocks or stress. This would enable both the HKMA and the AIs to put protective measures in place. The HKMA has made commendable efforts to improve this aspect of its work in recent years, including production of the macro- and micro-prudential reviews by the Banking Policy Department (BPD), the half-yearly Monetary and Financial Stability Report published by the HKMA, macro stress testing of the banking sector’s vulnerability to shocks such as the sub-prime crisis and the use of the macroeconomic credit risk models developed by the Market Research Division (MRD) of the Research Department (RD) to improve the stress testing of banks’ residential mortgage loans and other loan portfolios. 119. As noted earlier in the context of the sub-prime crisis, it will be important to further enhance the HKMA’s ability to foresee issues, both domestic and international, which might have an adverse impact on the Hong Kong banking system. The need for this will increase as Hong Kong develops as an international financial centre and its interconnectedness with the global financial markets increases still further.

38 120. The key issue for the HKMA is to improve information flows within the HKMA and the cooperation and collaboration between Banking Supervision Department (BSD), BPD and RD. The bottom-up analysis of trends and emerging risks contained in the macro-prudential and micro-prudential reviews which is distilled from the banking statistics needs to be supplemented by a more top-down review based on macro developments (such as recession in the US). This requires the ability to identify emerging trends and the transmission mechanism of these to the Hong Kong banking sector. It is important therefore that BSD communicates clearly to RD what it requires in this area so that RD can appropriately focus its research efforts and suggest themes for on-site examination. This dialogue does seem to be working to some extent (as shown by the use of MRD’s macroeconomic credit risk models), but from my discussions with HKMA staff it appeared that the interaction between RD and the banking areas is somewhat ad hoc and could be formalised. It should be noted that RD is also a source of advice on prudential policy issues, but where suggestions have been made on these by RD, it appears that it has not received feedback from the banking side. 121. A further point is that vertical communication within BSD should be addressed so that information on emerging risks is fed down to case officers and examination teams so that they know what to look out for. This was a gap identified by the FSA’s internal review of its handling of Northern Rock. 122. BSD and BPD should also ensure that they exploit to the full other information sources within the HKMA, including the External Department and the Monetary Management and Infrastructure Department. The latter is well-placed to pick up market developments and trends, including new financial products or favourite trades which may create vulnerabilities. This should be supplemented by BSD’s own regular discussions with AIs, which could include asking them which financial products are currently being marketed to them by the investment banks. 123. Accordingly, it is recommended that: • The HKMA should further enhance its efforts on macro-prudential surveillance, in particular to ensure that there is full cooperation and exchange of information among its relevant departments and that the expertise of the Research Department is fully utilised. It is important that risks and priorities identified during the course of surveillance feed down to, and influence the activities of, line supervisors. • To ensure that there is an opportunity for dialogue involving interested parties, macro-prudential trends and issues should be a regular item (say monthly) on the agenda of the HKMA Chief Executive’s Committee.

39 High turnover of staff

124. High turnover of staff at the junior/middle levels in BSD is making the HKMA’s supervisory task more difficult and was commented on by both senior management and some interviewees. It puts additional pressure on resources, creates a deficit in terms of experience and expertise and makes it more difficult to maintain relationships between HKMA staff and the firms they are supervising. I detected no evidence however of a serious morale problem in the HKMA. Other regulators (such as the UK FSA) are also experiencing high staff turnover (which was a factor in its failings in relation to Northern Rock) and it is natural that this will occur when industry conditions are buoyant as they have been in recent years. Where there is a general shortage of staff and the industry is being encouraged by the regulator to boost compliance and risk management capability, industry will regard the regulator itself as one of the sources of new personnel. Interchange of staff between the regulator and industry is a healthy thing, but it is problematic if too many staff are leaving, particularly if this happens around the same time.
125. There is no simple solution to this problem. Pay is obviously an issue but, while the HKMA tries to keep its salary levels generally competitive with industry, it is difficult in individual cases to match what firms are prepared to pay. Rather, a multi-faceted approach is required to ensure that the HKMA’s overall working environment is as conducive as possible to retaining the staff that it wants to keep. This would include: • Providing more in-built flexibility to reward good performance and give a better sense of career progression within BSD itself. • Introducing a new long-term incentive programme. • Improving and expanding training (e.g. in complex instruments), including overseas, to ensure staff feel confident in their ability to do the job (note that it is important that sufficient time for this is allowed so that training does not become simply another source of work pressure). • Ensuring that IT and other tools are appropriate and enable staff to do their job. • Prioritising work clearly so that staff can focus on the main tasks in hand. • Improving vertical as well as horizontal communication within the HKMA to promote a sense of inclusion as well as knowledge-sharing.

40 126. In summary, it is recommended that: • The HKMA should adopt a multi-faceted approach to try to manage turnover among its supervisory staff, including consideration of a long-term incentive programme for staff. 5. THE POLICY FRAMEWORK Supervisory policies 127. The HKMA has an extensive set of policies that largely comply with the revised Basel Core Principles for Effective Banking Supervision. In particular, it has been a world leader in the implementation of Basel II. The advantage of being an early mover in this respect is illustrated by the fact that timely implementation of Basel II is at the heart of the proposed response of the FSF and the Basel Committee to the sub-prime crisis. The fact that this has already been done in Hong Kong means that the HKMA has a solid platform to build on for the future. This will include the need to give effect to the various proposals of the Basel Committee to further strengthen Pillars 1 and 2 in order to improve the resilience of the banking system. The broad thrust of the international supervisory initiatives is, in a sense, a return to basics – focussing on strengthening capital and liquidity requirements and risk management practices. These, and the other proposals of the FSF and the Basel Committee, will help to set the policy agenda of the HKMA over the next few years in addition to those policy issues dictated by domestic considerations. 128. The HKMA’s policies are set out in the various modules of the SPM and in other guidelines and circulars. In addition, the Guide to Authorization describes the HKMA’s approach to licensing issues. 129. A key aim of the HKMA is to provide substantive guidance in relation to the eight inherent risks which are the focus of its risk-based approach. Towards this end, a SPM module on strategic risk has recently been issued and a module on reputation risk is out for consultation. A guideline on counterparty credit risk is now under preparation, which is timely since concerns about counterparty risk have been at the heart of the global credit crunch. Following this, the main remaining gap will be that there is no SPM module covering market risk as a whole and the existing guidelines on risk management of derivatives date from 1994 and 1996. This is a significant gap, particularly in the light of recent market developments, and one which should be filled as soon as possible. (I understand that a guideline on market risk management is under preparation though this is still at a very preliminary stage.) Some other policies also need to be updated in the light of Basel II and changes in accounting standards, including loan classification and

41 provisioning, and work on this is currently underway. 130. As noted above, the Basel Committee is also reviewing liquidity requirements in response to the recent financial market stress. The HKMA is represented on a Basel Working Group on Liquidity which is currently conducting a fundamental review of the Basel Committee’s published guidance on Sound Practices for Managing Liquidity in Banking Organisations. The Committee plans to issue the enhanced guidance for consultation in July 2008. The HKMA will obviously wish to take this into account in revising its own liquidity regime. This should include addressing the following points: • The stress tests on liquidity conducted by both the HKMA and individual AIs should be enhanced to reflect recent experience, including particularly the need to cater for prolonged periods of market stress. • The HKMA should consider whether it needs to offer more prescriptive guidance on the maturity mismatch limits set by AIs and behavioural assumptions used by them in projecting future cash flow. • While the main focus may be on cash flow management, recent events have demonstrated the importance of having a stock of genuinely high quality liquid assets on which institutions can rely to generate liquidity when funding dries up completely, either by sale or by repo. • The current statutory minimum liquidity ratio does not fulfil this role (the investments that are allowable within liquefiable assets are not necessarily of high quality and loan repayments due in the short-term are also included as a source of liquidity). • A statutory liquidity ratio provides a useful lever to force remedial action or trigger use of the HKMA’s powers in times of stress, but the purpose and composition of the current ratio (including its use of credit agency ratings for quality purposes) needs to be reviewed along with its relationship with the short end of the cash flow regime. 131. One of the reasons for the current financial crisis is the uncertainty about the actual exposure of individual institutions to credit and market risks. This emphasises the importance of valuation and adequate disclosure. The credit analysts and rating agencies whom I interviewed considered that, despite the advances in recent years, there are still gaps in the external reporting by Hong Kong banks in relation to such issues as the breakdown of their investments, China exposure and exposure to the various industry segments. This is an area that should be addressed by the HKMA in the light of the first round of

42 annual Pillar 3 disclosures in Hong Kong and the further recommendations on disclosure and valuation to be issued by the Basel Committee. 132. In summary, it is recommended that: • The HKMA should develop its supervisory policies in line with the detailed recommendations to be issued by the Basel Committee in response to the sub-prime crisis. This will include review of the current liquidity regime. A particular need in Hong Kong is to revamp the current statutory minimum liquidity ratio, including re-examination of its use of credit agency ratings, to ensure that it is a proper measure of high quality liquidity. • The HKMA should fill in the remaining gaps in its supervisory guidelines, including counterparty credit risk and management of market risk and derivatives. • The HKMA should review the disclosure requirements of Hong Kong banks both to meet local market needs and to take account of the forthcoming additional guidance from the Basel Committee. 133. The HKMA’s existing Guideline on AML/CFT dates from 1997 and has been updated to take account of later developments such as the revised FATF Forty Recommendations by means of a Supplement which was first issued in 2004 and most recently revised in November 2007 to take effect from May 2008. Taken together the Guideline and the Supplement provide comprehensive guidance to AIs. However, the use of a Supplement was supposed to be a short-term expedient and it should be consolidated as soon as possible into a SPM module on money laundering. The HKMA’s own AML/CFT guidance is being augmented by guidance papers on particular topics (such as the handling of politically exposed persons) issued by the Industry Working Group on Prevention of Money Laundering and Terrorist Financing, which is chaired by the HKMA. The work of this group is a very positive development. 134. Less positive are the gaps in the overall AML/CFT regime in Hong Kong. The lack of authorization and supervision of RAMCs has been already commented on. To comply with the FATF Recommendations, it will also be necessary for the Government to set out the basic customer due diligence and record keeping obligations for financial institutions and other relevant persons in legislation. This will provide the overall statutory framework within which the HKMA’s guidelines can set out the specific requirements that apply to AIs. 135. In summary, it is recommended that:

43 • The HKMA should consolidate the existing Guideline and Supplement on AML/CFT into one document, taking account of the recommendations that arise from the FATF review. This should be made consistent with the legislation on the customer due diligence and record keeping requirements for financial institutions when this is eventually produced. The organization of the HKMA’s policy framework 136. At present, SPM modules are issued in the following variants: • Statutory guidelines issued by the HKMA under the statutory powers in the BO (usually section 7(3)). These are intended to set out the minimum standards with which AIs are expected to comply, but may also embody best practices or advisory standards. • Non-statutory guidelines issued to AIs as guidance notes. These are best practice guides setting out the HKMA’s recommendations to AIs in respect of the standards they should aim to achieve. • Other non-statutory guidelines issued to AIs. These are usually technical in nature, for the purpose of clarifying the HKMA’s interpretation of regulatory and reporting matters. 137. In practice, the distinction between statutory and non-statutory guidelines and the policy for assigning statutory status to particular guidelines but not to others are not clear. For example, the guidelines on credit risk and liquidity risk are classified as statutory while those on operational risk, strategic risk and interest rate risk are classified as non-statutory guidance notes. In practice, both types of guideline contain a mixture of minimum requirements and best practice, e.g. paragraph 2.4.1 of the non-statutory guideline on supervision of e-banking (TM-E-1) reads: “The senior management of an AI are required to appoint trusted independent experts …to carry out an independent assessment before the launch of new e-banking services….” 138. Moreover, the consequences of a breach of a statutory or non-statutory guideline seem to be similar in that “any failure to adhere to any of these guidelines, whether statutory or non-statutory, may call into question whether the AI concerned meets the minimum criteria for authorization under the Banking Ordinance” (Introduction to the SPM).

44 139. There are problems with this approach. First, it does not meet the requirement of certainty – AIs may not be clear about those parts of the policy framework with which they must comply in order to continue to satisfy the criteria for authorization under the BO. Second, it gives rise to the risk of inconsistency of application and failure of understanding by AIs and HKMA staff alike. Many interviewees expressed the need for a clearer distinction between minimum standards and best practice. This could be done within the current format by highlighting more clearly those parts of a guideline (whether statutory or non-statutory) which set minimum requirements (for example by using bold italics and using the word “must” to highlight these). 140. The distinction between best practice and minimum standards could be further strengthened by embodying at least some of the latter in formal rules made by the MA using a new general rule-making power within the BO. This would put the HKMA on a par with other regulators such the SFC and the UK FSA. It would be desirable to set such rules within the context of an overarching set of fundamental principles (which themselves would have the status of rules). Arguably significant parts of the BO (such as the authorization criteria) already provide a principle-based code for AIs, the breach of which can lead to supervisory action. But there may be advantage in making such a code of principles more explicit. This would have the benefit of imposing order and structure on a multiplicity of rules on different subjects as well as enabling action to be taken in respect of bad behaviour which may not be covered by a specific rule but clearly breaches a fundamental principle. 141. A further step that should be taken is to bring all the relevant rules, guidelines, guidance and circulars on each supervisory topic into a consolidated “regulatory handbook” similar to that of the SFC (see its website). At present it is awkward to have to search separately through the miscellaneous circulars and guidelines as well as the SPM in order to obtain all the information on a particular topic. This consolidation should include the material in the Guide to Authorization which would bring the guidelines on licensing within the overall policy framework. 142. In summary, it is recommended that: • The HKMA should review its policy on the distinction between statutory and non-statutory guidelines. • The HKMA should also make a clearer distinction in its guidelines between minimum requirements and best practice. The aim should be to set out clearly and simply the “rules of the game”, supplemented by guidance on how the minimum requirements may, but not must, be complied with.

45 • To strengthen this, there should be consideration of whether the MA should be given a general power to make formal rules under the BO. • The HKMA’s regulatory requirements should be set within an explicit framework of high level fundamental principles. • All the various elements of the policy framework should be brought together in a regulatory handbook on similar lines to that of the SFC. 6. AUTHORIZATION The three-tier structure 143. Institutions that wish to carry on banking business or the business of taking deposits in Hong Kong are required to be authorized by the MA under section 16(1) of the BO. The criteria for authorization are set out in the Seventh Schedule of the Ordinance. 144. Hong Kong maintains a three-tier structure of AIs: banks, restricted licence banks (RLBs) and DTCs. The main distinguishing features of banks are that: • Only banks may carry on banking business, i.e. provide current and savings accounts and demand deposits, without restriction on size or maturity, and pay or collect cheques. • Banks must have a paid up capital of at least HK$300 million. • Banks must join the Real Time Gross Settlement (RTGS) payments system. • Banks must join the DPS. • Only banks are allowed to use a banking name without restriction (e.g. XYZ Bank). 145. RLBs may not carry on banking business and are restricted to taking call, notice or time deposits from the public in amounts of HK$500,000 and above. DTCs are restricted to taking deposits of HK$100,000 and above and with a maturity of at least three months. For both types of institution, the taking of deposits will usually be incidental to their main business and this, together with the restrictions on their deposit-taking capability, means that their combined share of total industry deposits is minimal.

46 146. The main original rationale of the three-tier system was to preserve the integrity of the interest rate rules. Now that these have been abolished, that purpose is no longer served. The main remaining reason for segmentation is to allow participation in the banking system, within a regulated environment, to a broader range of financial institutions, while confining access to the market for retail deposits to fully fledged banks with the necessary financial strength and systems capability. In other words, the current system seeks to strike a balance between flexibility of entry and protection for small depositors. 147. However, it appears that the system is now more complex than it needs to be to achieve that balance, particularly given the decline in the number of RLBs and DTCs (to 29 each at end-2007) and their minimal share of deposits. This was the conclusion of KPMG when the matter was last reviewed in 1998. The recommendation at that time was that the three tiers should be reduced to two. However, while this change was accepted by the HKMA, it was placed in abeyance while other reforms (such as the abolition of the interest rate rules) worked their way through the system. 148. Now that the other reforms have bedded down, it seems an opportune time to return to the issue of the three-tier structure. Some change in the current arrangements was generally supported by interviewees, and naturally the DTCA is strongly in favour of reform of a system which it sees as “antiquated and anti-competitive”. 149. In considering what changes might be appropriate, while maintaining the balance between flexibility and depositor protection cited above, the following considerations seem relevant: • It is appropriate to maintain a distinction between full banks and “other deposit-takers” whose interest in taking deposits may be peripheral. • The retail banking public should be made clearly aware of the status of the AI they are dealing with. • AIs should be given a realistic chance to upgrade to full banking status if they have the financial and systems capability to do so. • Any changes made should not have the effect of driving significant numbers of existing RLBs/DTCs into giving up authorized status and joining the unregulated sector. 150. Based on these principles, the current system might be changed as follows: • The three-tier structure should be reduced to two tiers: banks and other deposit-takers (i.e. RLBs and DTCs combined).

47 • As at present, only banks should be able to conduct the full range of banking business. • The other deposit-takers should be allowed to take call, notice and time deposits of, say, $100,000 or more without restriction on maturity (other than in respect of deposits repayable on demand). • The other deposit-takers should have the opportunity to access the RTGS system. • Membership of the DPS should continue to be reserved for full banks. Other deposit-takers should be required to state on advertisements, notepaper etc that they are not members of the DPS. • The other deposit-takers would not be allowed to use banking names unless they are a branch of a foreign bank or a subsidiary of a bank. • The requirement to have at least HK$3 billion of unconnected deposits in order to be authorized as a bank should be removed to make it more feasible to upgrade. 151. The key question in the above is what the minimum size of deposit for the other deposit-takers should be. This essentially sets the boundary line between retail and non-retail deposits. The figure of HK$100,000 has a logic since it corresponds with the limit on the coverage provided by the DPS. However, it has arguably become too low with the passage of time, and there would be a case for setting a higher minimum size of deposits, say HK$200,000. In this case, consideration should also be given to increasing the coverage limit of the DPS, a point which is dealt with below. 152. Another issue is what the new class of AI should be called. It would be inappropriate for DTCs to be called RLBs and RLBs may not want to be called DTCs, particularly since the DTCA regards the Chinese legal name for “deposit-taking companies” as significantly pejorative in the local market (a point which should in any case be addressed by the HKMA). The best solution might be to devise a new name, which might simply be that of “deposit-taker” as used above, with an appropriate Chinese translation. 153. The proposal to make it easier for RLBs or DTCs to upgrade to become full banks could add further to competition in the local banking market. However, it is unlikely to open the floodgates since it would require a major change in business model and significant investment which only a small number of institutions are likely to want to make.

48 154. In summary, it is recommended that: • The current three-tier structure of authorized institutions should be reduced to two tiers: banks and other “deposit-takers”. • The principal dividing line between the two tiers should be the ability to conduct banking business and to take “retail deposits”, both of which (along with membership of the DPS) should be reserved for banks. Licensing and promotion 155. As already noted, more overseas banks (including those from the Mainland) are likely to want to establish a presence in Hong Kong. The HKMA is keen to encourage this as a means of developing Hong Kong as an international financial centre. This raises the question of whether there is scope for relaxing the current authorization criteria in order to make it easier to attract more overseas banks to Hong Kong. 156. However, the authorization criteria set out in the Seventh Schedule of the BO comply with international standards and do not appear to be excessive. They do not set a threshold which a financially sound and properly managed bank would not be able to meet. There is therefore no case for relaxation of the standard criteria and to do so would be to store up problems for the future. 157. The question has been raised however whether Hong Kong should attempt to make itself more attractive to niche-market financial institutions, such as private banks, by providing a special licensing regime for them. However, it is not clear why a private bank should not be able to meet the basic standards that apply to other banks, and in any case the licensing regime is unlikely to be only one factor in attracting such banks to come to Hong Kong. Singapore has for example implemented a variety of measures to attract private bank business, including strengthening its bank secrecy and trust law. However, there are dangers in an approach that relies heavily on use of bank secrecy as a marketing device. In Europe, the net is closing on those jurisdictions that are regarded as being excessively secretive such as Lichtenstein, and it cannot be ruled out that at some point increased international pressure will be brought to bear on centres in other regions such as Asia. 158. There is no reason why the HKMA should not engage in marketing efforts to try to attract more banks to come to Hong Kong, provided that it does not drop its entry standards. In this context, a better focus for the HKMA’s efforts may be to ensure that the licensing process is as efficient and as speedy as possible. Clear performance targets should be set for the team involved in licensing. In addition, the process will be eased if the HKMA establishes ongoing bilateral relationships with those regulators whose banks may wish to

49 come to Hong Kong (e.g. countries such as the Middle East, India and Russia). 159. Accordingly it is recommended that: • The HKMA should not relax its present licensing criteria for the admission of foreign banks, but should focus instead on ensuring that the licensing process is as efficient and as speedy as possible. • Towards this end, the HKMA should establish ongoing bilateral links with regulators from those jurisdictions whose banks may wish to enter the Hong Kong market. 7. SAFETY NET ARRANGEMENTS 160. Regulation and supervision play the most important part in fostering banking stability by reducing the risk that problems arise in the first place. But when problems do arise, recent UK experience shows that it is necessary to have in place effective safety net arrangements to mitigate their impact. In Hong Kong, these are provided by the Lender of Last Resort (LOLR) facility and the DPS. A further vital role is fulfilled by the Hong Kong Mortgage Corporation (HKMC) which stands ready to acquire residential mortgage loans which meet its criteria. This adds to the liquidity of AIs’ balance sheets. The advantage of this is illustrated by the ad hoc arrangements that have had to be put in place in the UK and US in order to ease liquidity shortages. 161. Interviewees had little to say about the LOLR which has yet to be tested. The readiness to accept a wide range of collateral is seen as a plus point, the value of which has been demonstrated in other jurisdictions during the credit crunch. But one bank commented that it was not appropriate to set limits on support (at present the HKMA will lend up to between 100-200% of an institution’s capital base). However, this is the amount that can be lent at the discretion of the MA himself – it is possible to go beyond this with the prior approval of the FS. This seems appropriate, since the Northern Rock experience confirms that the decision to commit large amounts of public funds to a troubled bank is inevitably a political one and should be taken at that level. The crucial point is that the decision should be taken quickly. The mechanics of providing support should also work smoothly and the HKMA needs to ensure that there will be no barriers to the quick provision of support (e.g. up to date security documentation is in place and AIs have clearly identified those assets, such as HKMC-compatible mortgages, which can be used to obtain LOLR support). Such support should also be provided as discreetly as possible so that news of the LOLR assistance does not make the problem worse.

50 162. Speed is also of the essence if a bank in difficulties actually goes on to fail and the DPS is triggered. The DPS was launched in September 2006 and stands ready to deal with a bank collapse which hopefully will never occur. Indeed, part of the rationale of the DPS is to reduce the risk of failure through the provision of protection that will make it less likely that small depositors will rush to take their money out of a bank rumoured to be in trouble. But this clearly did not work in the UK in relation to Northern Rock. The inquest into what went wrong identified certain defects in the UK scheme and the UK authorities are currently consulting on possible changes to it. One immediate change already made is to provide 100% coverage up to the protection limit of £35,000 (previously only 90% of the amount in excess of the first £2,000 had been protected). Other key priorities are to increase public awareness of the scheme, to achieve a much faster payout to minimise the length of time that depositors’ funds are tied up, and to make gross rather than net payments: at present a depositor’s liability to the bank is netted off against the amount due on his deposit before compensation is paid. It is also possible that the coverage limit of £35,000 may be increased though the need for this is less clear-cut (the present limit covers over 90% of UK depositors by number). 163. There are some implications here for the Hong Kong scheme (which I understand are currently being studied). A number of the defects revealed by Northern Rock were already addressed in the way in which the DPS was set up – in particular, it is pre-funded (which facilitates quick payouts) and it offers 100% coverage. It aims to make at least an interim payment to depositors within two weeks. Moreover, since its inception the board of the DPS has mounted an extensive campaign to increase public awareness of the scheme as a means of promoting depositor confidence. 164. However, there are two issues in particular which deserve further study. The first is the level of coverage in the Hong Kong scheme which is low in absolute terms (at HK$100,000), albeit that it offers full protection to over 80% of depositors by number. Views among interviewees were mixed as to whether the level of coverage should be increased. As might be expected, the larger banks are against, while the smaller banks (and the Consumer Council) are in favour. However, a number of the banks in favour are only of this mind if the premium does not increase. If this can be achieved (which appears possible) and if an increase in the coverage limit (say to HK$200,000) does result in a material increase in the percentage of depositors who are covered, this would seem to be a desirable step. 165. The other key issue is whether payouts should be made gross or net of the depositor’s liabilities to the bank (e.g. in respect of residential mortgage loans). Net payments are consistent with the treatment that is applied in the liquidation of a bank and reduce the risks and costs to the DPS of making a payout. But netting does potentially increase the time to make a full payout and may result in a loss of liquidity on the part of the depositor. The DPS

51 has already undertaken some research on this issue by reference to two large medium size banks and it appears that the majority of depositors (around 70%) do not have liabilities to the banks, and of those who do, the average gross deposit balance is quite small in comparison with their liabilities. This suggests that such “small depositors” may keep their deposit balances with the bank for servicing their liabilities rather than for savings purposes, and spread their savings around other banks in the system. The netting issue may therefore be less critical in Hong Kong. Nonetheless the issue will be worth further study in the light of developments in the UK, not least what proposals to change the bank insolvency regime eventually emerge. In addition, international principles for deposit insurance arrangements are due to be agreed as part of the FSF’s package of measures in response to the sub-prime crisis, and Hong Kong will obviously want to take account of these when they emerge. 166. In the light of this, it is recommended that: • The DPS should review the level of deposit protection provided in Hong Kong and the potential for raising the protection limit without increasing the annual premium. • The DPS should study the changes that are made to the UK scheme in the wake of Northern Rock and the new international principles for deposit insurance when they emerge to consider what application they may have to Hong Kong.

1.1 Appendix 1 TERMS OF REFERENCE

1. The Consultant shall assess and prepare a report on how the MA should best discharge its functions under the Banking Ordinance, taking into account recent and likely future developments in Hong Kong’s banking system and the changing nature of the risks facing Hong Kong’s banking system, including: (a) the globalisation of banking business and of finance generally; (b) the increasing integration of the financial systems of Hong Kong and the Mainland of China; (c) the growing competition from other financial centres; (d) the participation of banks in securities, insurance and other forms of business, and the growing complexity of banking products; (e) the increasing reliance of banks on information technology; (f) the development of new ways of delivering business, including the internet and other electronic forms; (g) the increasing need to combat financial crime (such as money laundering, terrorist financing, and fraud); (h) intensified competition for business; (i) the changing nature of supervision, including the implementation of Basel II, the emphasis on thematic, risk-based examinations, and the greater focus on managing special areas of risk; and (j) the expectations of consumers and of the community generally.
2. In his report, the Consultant shall make recommendations on: (a) the focus and priorities of the MA’s banking supervisory functions in the coming 5 to 10 years; (b) specific policies to be developed by the MA for promoting the general stability and effective working of the banking system; (c) whether any functions and activities carried out by the MA in relation to the banking system should be expanded, reduced or discontinued;

1.2 (d) the resources, including staff, skills, expertise and training, required by the MA to carry out its responsibilities in relation to the banking system; and (e) any legislative, governance or organisational changes required for the MA to undertake new or modified functions resulting from the Review.

2.1 Appendix 2 SUMMARY OF RECOMMENDATIONS

1. TRENDS AND ISSUES Business integration with the Mainland • The HKMA should ensure that Hong Kong banks have appropriate oversight and control of their business operations on the Mainland. • The HKMA should have a full understanding of the business models that Hong Kong banks are pursuing on the Mainland and of the type and volume of business being conducted there and the risk controls in place. It should be prepared to challenge business models that it considers inappropriate. • The HKMA should have a full knowledge of the business and regulatory environment on the Mainland, so as to be able to understand the risks and to offer supervisory guidance to authorized institutions (AIs). • The HKMA should seek to strengthen communication and cooperation with the China Banking Regulatory Commission (CBRC) at all levels (at municipal and provincial levels as well as national), and to cement this through help with training and two-way secondments. • While relying principally on the direct supervision of the CBRC of Hong Kong banks’ subsidiaries on the Mainland, the HKMA should be prepared to conduct its own periodic on-site examinations of these (subject to any necessary approval from the CBRC). • Part of the HKMA’s cooperation with the CBRC should be to try to harmonise regulatory policies and practices to produce a level playing field on both sides of the border and to reduce the risk of regulatory arbitrage. • The HKMA should liaise closely with the CBRC to ensure that Mainland banks which wish to come to Hong Kong are fit to do so. • The HKMA should attach a high priority to the supervision of Mainland banks once they have set up in Hong Kong. • The HKMA should continue in its efforts to encourage the Mainland authorities to establish new, and expand existing, regulated channels to

2.2 facilitate the mobility of capital from the Mainland to Hong Kong or through Hong Kong to the rest of the world. • The HKMA should ensure that Hong Kong AIs exercise adequate customer due diligence and transaction monitoring in respect of non-resident customers, including those from the Mainland. Impact of the sub-prime crisis • The HKMA should develop its capability to look ahead, to the extent possible, to detect global trends and pick up early warning signs of the risks that arise from these trends, even when these seem improbable. • Where major risks are identified, steps should be taken to bring these to the attention of the boards of AIs (e.g. by writing to them) and positive confirmation should be sought that these risks are being addressed. • The HKMA should continue to insist that AIs apply prudent criteria in their residential mortgage lending, including adherence to the 70% loan to value ratio. Development of Hong Kong as an international financial centre • The HKMA should ensure that it continues to provide a regulatory regime that maintains effective standards while at the same time facilitating access to new entrants. • The HKMA should continue its efforts to develop the bond market in Hong Kong (e.g. through facilitating the issue of RMB bonds). • The HKMA should pursue the development of Hong Kong as a centre for Islamic bonds and other forms of Islamic finance as a way of attracting investors from the Middle East and widening the range of products in Hong Kong. • The HKMA should continue to develop Hong Kong as a regional payment and settlement hub. • The HKMA should continue to participate in international fora as a means of raising Hong Kong’s international profile and influencing the development of international regulatory standards.

2.3 Increasing competition • The HKMA should take steps to encourage AIs to adopt a more balanced funding structure, e.g. through its efforts to promote the development of the bond market. • The HKMA should monitor that increased competition does not encourage increased risk-taking (e.g. through investment in complex instruments) that is not justified in relation to capital or earnings. • The HKMA should continue its efforts to encourage improvements in the risk management of the smaller banks, ensuring in particular that they fully appreciate the interest rate risk posed by the currency board system. • The HKMA should be prepared to back its judgement and take decisive action where it believes that a business model is unsustainable or risk management is dangerously weak (to avoid a repetition of the Northern Rock situation in the UK). • Such decisive action should include pressing for merger or acquisition and taking whatever action is possible to facilitate this (again bearing in mind the Northern Rock experience). • The HKMA should consider the extent to which capital requirements are providing a disincentive to the recycling of funds to the local banks during IPOs and in effect forcing liquidity offshore. Erosion of financial boundaries • The HKMA should undertake an assessment in consultation with HKAB and other interested parties of whether the industry’s perceived concerns about the operation of multiple regulators are indeed valid, and if so how the current system can be enhanced. • The HKMA should consider with the SFC how the process of consultation between the two bodies on the use of enforcement measures might be streamlined. This should recognise that, while the views of the consulted party should be taken into account, it will not always be possible to achieve consensus on the measures to be taken. • The HKMA should pursue its plan to raise its transparency in relation to its role in disciplinary cases and the procedures that it follows, and to provide periodic updates and feedback to the industry on its activities in this area.

2.4 • More staff secondments should be undertaken between the HKMA and the SFC to provide training and to promote harmonisation of culture and practices. 2. GOVERNANCE ISSUES RELEVANT TO BANKING STABILITY The role of the Advisory Committees • The effectiveness of the Banking Advisory Committee and Deposit-taking Companies Advisory Committee should be strengthened. Pending legislative change, the two Committees should continue to meet jointly, but efforts should be made to reduce the size of the membership of each Committee to enable joint meetings of more manageable proportions to be held. • Unless there is a particular issue that requires his presence, the FS should normally delegate his responsibility for chairing the joint meeting to the MA. • In the longer term, the Banking Ordinance (BO) should be changed to turn the two Committees into one Advisory Committee whose role is to advise the MA on matters of policy in relation to his regulatory functions. The new Committee should be chaired by the MA with a maximum of say 12 external members (including an appropriate representation of non-bankers). Appeal mechanism • The current right of appeal to the Chief Executive in Council in the BO should be replaced with a right of appeal to an independent Appeals Tribunal. (This might be done by enlarging the scope of the current Review Tribunal to turn it into a Banking Appeals Tribunal.) The role of the Hong Kong Association of Banks • The HKMA should suggest to HKAB that it conducts a strategic review of its role, functions and organisational structure, with particular reference to whether it should appoint a professional chief executive and a permanent secretariat. Strategic planning • The HKMA should refine its business plan to make it more specific and make it easier to monitor implementation of the various initiatives and to assess the resource implications.

2.5 3. THE MONETARY AUTHORITY’S FUNCTIONS AND POWERS IN THE BANKING ORDINANCE The HKMA’s role in consumer protection • The HKMA should maintain its primary focus on prudential issues but should also remain involved in consumer protection issues that fall within its current statutory remit to promote the general stability and effective working of the banking system. It does not appear necessary for the MA to be given a specific statutory responsibility for consumer protection. • There may be advantage in the HKMA preparing a formal guideline under section 7(3) of the BO setting out how it considers it should exercise this aspect of the MA’s functions and what the limits on it are. • The HKMA should continue not to be involved in resolution of individual disputes between banks and their customers. That is the role of a banking (or more generally a financial services) ombudsman for which there may at some point be a need in Hong Kong, particularly as wealth management activities expand. Supervision of remittance agents and money changers • The HKMA should not assume responsibility for the AML/CFT supervision of RAMCs, but some other solution should be found. The MA’s powers of sanction • The HKMA should consider whether it feels the need for an extension of the MA’s powers of sanction to include the issuing of public statements and the ability to fine. If so, it should present the case to the Government and the banking industry. • Whether or not such powers are granted, the HKMA should consider issuing a guideline under section 7(3) of the BO describing the various enforcement powers available to it and the broad principles that would govern the use of such powers.

2.6 The MA’s powers of investigation • The powers of investigation of the MA under the BO should be extended so that they are no less than those available to the SFC. The MA’s powers of intervention • The HKMA should follow closely the debate in the UK post-Northern Rock about how to deal with failing banks and apply any relevant lessons learned to the situation in Hong Kong, including any necessary enhancements to Part X of the BO. 4. THE SUPERVISORY FRAMEWORK The need to streamline the off-site review process • The current off-site review process should be streamlined to reduce the need to run different systems side by side and to produce a logical and efficient sequence of actions leading up to on-site examination. The Supervisory Review Process (SRP) should be revised to take account of the further Pillar 2 guidance to be issued by the Basel Committee. • Risk assessment should be an on-going process throughout the year with re-assessment of the risk profile as new information emerges. There should be a clear mechanism for escalating emerging risk issues to senior management and appropriate arrangements for quality control of the assessment process. • The work pressures on case officers should be alleviated by reviewing the purpose of their various tasks, by making sure, on an ongoing basis, that the limited supervisory resources are allocated to tasks that are of a higher priority, and by giving them better tools to access the EPSS database and conduct financial analysis (including peer group review). • The ability of case officers to assess complex instruments and risks (including strategic risks) should be addressed though enhanced training in this area, through better knowledge transfer within the HKMA (e.g. from the specialised examination teams) and through increased use of short-term secondments to industry. • The current risk ratings (e.g. low, moderate and high for risks and strong, acceptable and weak for controls) should be made more granular to avoid fence-sitting by case officers and to allow more precise ratings. For example, using an approach similar to that of the UK FSA would

2.7 give four ratings for risk: low, moderate-low, moderate-high and high, with the two latter ratings being the main area of focus for on-site examination. • An explicit impact rating should be assigned to each AI including DTCs. (Impact being measured in relation to the effect on depositors and systemic stability if the firm were to fail or become involved in financial scandal of some kind, e.g. related to money laundering or fraud.) In general, reduced-scope risk assessment and supervisory monitoring may be appropriate for low impact firms (although the risk profile of the firms concerned will also be a relevant factor in deciding the intensity of supervision). The HKMA should publish a guideline setting out its supervisory policy in this respect. • Related to this is the need for the HKMA to establish its tolerance in relation to the risk of failure or major scandal. This should help to establish the intensity of the HKMA’s supervision by type of firm and the resources it needs to ensure that its risk tolerance is not exceeded. The need to rationalise the on-site examination process • The HKMA should make efforts to reduce the number of separate examinations to which AIs are subject in any given year. • In particular, the HKMA should re-consider the role of risk-based examinations in the light of the move to specialised and thematic examinations. Unless the risk profile of a local bank includes high or moderate-high risks or control weaknesses, consideration should be given to whether it requires a risk-based examination each year (particularly if it is being subject to other types of examination that year). • The HKMA should exercise restraint on the thematic examination of particular AIs to obtain a benchmark of best practice. Other methods should be used including discussions with, or presentations from, the AI concerned on the practices they use in the area under review. • The HKMA should review the amount of detail that examiners should normally cover, perhaps in the case of well-run AIs allowing examinations to be confined to a relatively high level unless that reveals particular issues that require more detailed investigation. • The HKMA should present its examination findings in a more concise way by focussing on the high priority items and itself specifying the

2.8 action points to be taken on the basis of the findings and the deadline for these to be implemented. • Follow-up should be conducted primarily by the chief executive of the AI confirming that the action points have been implemented by the deadline specified by the HKMA. Confirmation of this could be provided by the external auditors and be subject to spot checks by the HKMA. Enforcement action should be taken if it turns out that the HKMA has been misled. • More use should be made of section 59 of the BO to commission reports on AIs’ systems and controls from external auditors (only 3 such reports were commissioned in 2007 and 1 in 2006). • In general, the HKMA should consider ways in which it can leverage its use of external auditors, for example by allowing them to attend meetings with the boards of directors as observers. Enhancement of macro-prudential surveillance • The HKMA should further enhance its efforts on macro-prudential surveillance, in particular to ensure that there is full cooperation and exchange of information among its relevant departments and that the expertise of the Research Department is fully utilised. It is important that risks and priorities identified during the course of surveillance feed down to, and influence the activities of, line supervisors. • To ensure that there is an opportunity for dialogue involving interested parties, macro-prudential trends and issues should be a regular item (say monthly) on the agenda of the HKMA Chief Executive’s Committee. High turnover of staff • The HKMA should adopt a multi-faceted approach to try to manage turnover among its supervisory staff, including consideration of a long-term incentive programme for staff. 5. THE POLICY FRAMEWORK Supervisory policies • The HKMA should develop its supervisory policies in line with the detailed recommendations to be issued by the Basel Committee in response to the sub-prime crisis. This will include review of the current liquidity regime. A particular need in Hong Kong is to revamp the

2.9 current statutory minimum liquidity ratio, including re-examination of its use of credit agency ratings, to ensure that it is a proper measure of high quality liquidity. • The HKMA should fill in the remaining gaps in its supervisory guidelines, including counterparty credit risk and management of market risk and derivatives. • The HKMA should review the disclosure requirements of Hong Kong banks both to meet local market needs and to take account of the forthcoming additional guidance from the Basel Committee. Prevention of money laundering and terrorist financing (AML/CFT) • The HKMA should consolidate the existing Guideline and Supplement on AML/CFT into one document, taking account of the recommendations that arise from the FATF review. This should be made consistent with the legislation on the customer due diligence and record keeping requirements for financial institutions when this is eventually produced. The organization of the HKMA’s policy framework • The HKMA should review its policy on the distinction between statutory and non-statutory guidelines. • The HKMA should also make a clearer distinction in its guidelines between minimum requirements and best practice. The aim should be to set out clearly and simply the “rules of the game”, supplemented by guidance on how the minimum requirements may, but not must, be complied with. • To strengthen this, there should be consideration of whether the MA should be given a general power to make formal rules under the BO. • The HKMA’s regulatory requirements should be set within an explicit framework of high level fundamental principles. • All the various elements of the policy framework should be brought together in a regulatory handbook on similar lines to that of the SFC.

2.10 6. AUTHORIZATION The three tier structure • The current three tier structure of authorized institutions should be reduced to two tiers: banks and other “deposit-takers”. • The principal dividing line between the two tiers should be the ability to conduct banking business and to take “retail deposits”, both of which (along with membership of the Deposit Protection Scheme) should be reserved for banks. Licensing and promotion • The HKMA should not relax its present licensing criteria for the admission of foreign banks, but should focus instead on ensuring that the licensing process is as efficient and as speedy as possible. • Towards this end, the HKMA should establish ongoing bilateral links with regulators from those jurisdictions whose banks may wish to enter the Hong Kong market. 7. SAFETY NET ARRANGEMENTS Deposit protection scheme (DPS) • The DPS should review the level of deposit protection provided in Hong Kong and the potential for raising the protection limit without increasing the annual premium. • The DPS should study the changes that are made to the UK scheme in the wake of Northern Rock and the new international principles for deposit insurance when they emerge to consider what application they may have to Hong Kong.