Recent Developments in the Supervision and Regulation of FinancialInstitutions in Coping with Capital Flows and Volatility of Capital Markets

Speeches

02 Sep 1995

Recent Developments in the Supervision and Regulation of FinancialInstitutions in Coping with Capital Flows and Volatility of Capital Markets

David Carse, Deputy Chief Executive, Hong Kong Monetary Authority

(Speech at the Asian Development Bank Fifth Seminar on International Finance)

The increase in capital flows into the emerging markets has been one of the dominant features of the 1990s. The effect has been to bind financial markets around the world even more tightly into the global market place. While this has brought major benefits to the recipient countries, it has also left them more exposed to external shocks. Moreover, the financial intermediaries through which the flows take place are subject to greater risks. In turn, life has become tougher for the financial regulators.

During 1995 the Mexican financial crisis and the Barings incident highlighted the underlying problems that regulators in both the newly liberalised emerging markets and the developed markets face in coping with rapid capital flows and market volatility. The situation becomes more complicated when financial institutions broaden their business scope. Banks are becoming more actively involved in financial activities such as securities trading in addition to their traditional roles of being intermediaries between depositors and borrowers. Banks may find it difficult to adapt their internal control systems and culture to cope with this widening range of activities. Regulators may also find it difficult to keep up with the changes.

The volatility of markets thus poses challenges for regulators of different types of financial institution and increases the necessity for cooperation between them. However, the challenges for the banking supervisor are perhaps greatest because of the central role of the banks in the financial system. Banks tend to be the direct victims of external shocks caused by reversal of capital flows or a collapse in asset values. They also provide the channel, via the payments system and the interbank market, whereby contagion spreads more widely after an initial shock, giving rise to systemic crisis. This paper will therefore concentrate on the regulation and supervision of banks, while recognising that regulation of other types of financial institution (such as securities companies) is also vital in terms of improving market transparency, achieving a level playing field and promoting investor protection.

The paper will first review briefly the trend of capital flows and market volatility in the major developed and developing economies in the last few years. It will then discuss the problems which the banking sector can suffer from such developments. Finally, the regulatory and supervisory policy response to these problems will be addressed.


Capital flows to developing countries surged rapidly in the early 1990s

Global capital flows had been modest in the 1980s after the Latin American debt crisis and the flows were mainly amongst developed countries, with Japan and Germany being the major capital exporters and the US the principal capital importer. The situation changed markedly at the beginning of the 1990s. Global capital flows increased substantially, in particular those to the developing countries. In 1993, total net capital flows to developing economies reached US$155 billion, almost four times the amount in 1990 (see Table 1) and seventeen times the annual average in the period 1983-89. Much of this growth took the form of increased portfolio flows (into bonds and equities) which rose from only US$6 billion to US$88 billion in 1993, more than half of the overall total. During the period, Western Hemisphere countries (mainly Latin America) and Asia were the major recipients of these funds.

A major distinction between the capital flows to Latin America and to Asia is that a relatively large proportion of the net inflows to the latter took the form of direct investment, mostly in the form of acquisitions or establishment of new enterprises, though portfolio flows also rose rapidly. Between 1990-94, Asia received about 60% of the total foreign direct investment to developing economies while Western Hemisphere received 30%. As regards portfolio investment, the opposite was true. During the same period, Western Hemisphere took up 61% of such inflows while Asia received 29%1. In simple terms, it can be said that a large part of the inflows into the Asian countries was invested in real, income generating assets; while those to Latin America went mainly into potential volatile and interest sensitive financial assets.

According to World Bank estimates2, during the early 1990s, the inflow of funds into the Asian and the Latin American economies was highly concentrated in a small number of countries. In Asia, China, South Korea, Malaysia and Thailand were the leading host countries for the inflows and together received about 41% of the total inflows to developing economies between 1990 and 1993. During the same period, Mexico, Argentina, Brazil and Venezuela together absorbed 26% of such flows. Mexico was a particularly large taker of funds, receiving (according to IMF estimates3) about 20% of total net capital flows to all developing countries in 1993.

A number of developments in the world economy encouraged the rapid surge in the flow of funds to the developing economies in the early 1990s. First, the major developed economies were generally in the downturn of their economic cycles between 1991-93 (see Table 2). Slow economic growth, weak consumption and investment demand coupled with low interest rates encouraged global funds to diversify their investments in search of greater return. The growth of such funds had been encouraged by the ageing of populations in the developed world with a consequent build-up of funds in pension, insurance and mutual funds. The OECD estimates that at the end of 1991, OECD mutual, insurance and pension funds had assets amounting to US$12 trillion as against banking sector assets of US$20 trillion. Portfolio diversification by the funds in search of higher return encouraged inflows into emerging markets.

At the same time, improved macroeconomic performance made the developing economies more attractive. Between 1980 and 1994, the average annual GDP growth in East Asia was 8.4%, more than double the world average of 3.4%. Stronger growth, lower inflation, and improved fiscal and balance of payments positions in a number of the developing economies provided a powerful incentive to investors (see Table 3).

Apart from strong economic growth, financial liberalisation and economic reform in the emerging markets also helped to attract incoming capital. Moves towards a more liberalised economic environment became a major policy trend in the developing countries. Reforms included: reducing government controls over private economic activities; adoption of stable exchange rate regimes either in the form of fixed rate, currency board system or managed float backed up by appropriate monetary policies; privatisation of state-owned enterprises; lowering or removal of import tariffs; institutional reform in the capital markets; liberalisation of capital account; and the removal of direct controls on the banking system.

Capital inflows to the developing economies fell in 1994 reflecting a decline in inflows to Latin America. The main factor was the erosion of confidence in Mexico culminating in the peso devaluation in December 1994. According to IMF estimates, capital inflows to the Western Hemisphere countries fell from US$62 billion in 1993 to US$39 billion in 1994, largely due to the decline in net flows to Mexico. The overall decrease was wholly in portfolio investment while direct investment actually increased. However, total capital flows to the Asian countries held up well. In 1994, they rose slightly from US$72 billion to US$73 billion, reflecting an increase in both direct and portfolio investment (offsetting a decline in "other" flows).

Even without the special situation in Mexico, some deceleration in capital flows to the developing countries would have been expected in 1994. The economic recovery in the US and in a number of European economies induced repatriation of funds to those countries or at least slowed the growth in capital outflows. The change in the US monetary policy stance signalled by the 25 basis point rise in the US federal funds rate in early 1994 would have had a similar effect. In general, portfolio investment will tend to be more sensitive to changes in market sentiment and international financial conditions and hence more volatile4 . This volatility will in turn feed through into the markets for the financial assets in which the flows are first invested and then disinvested.


The Mexican Crisis

It is generally agreed that the fundamentals in Mexico were reasonably sound ahead of the December 1994 crisis. In particular, the current account deficit which had widened to US$28 billion in 1994 did not reflect an out-of-control budget deficit. Rather it reflected a decline in the private savings rate encouraged by domestic credit expansion. Deregulation of the banking system had given the banks the appetite to expand their lending and capital inflows during the years prior to 1994 had helped to give them the liquidity with which to do it.

During 1994, political uncertainties contributed to a growing lack of confidence that the peso/dollar exchange rate could be maintained. This led to capital outflows by both domestic and foreign residents, particularly the former, resulting in heavy loss of reserves and making it more difficult to sell peso-denominated securities. As a result, the Government was forced to sell increasing amounts of short-term US dollar-linked securities (tesobonos), the amount of which held outside the banking system rose from less than US$2 billion at the end of 1993 to US$21 billion at end-1994. Most of these were held by non-residents. At the same time, higher domestic interest rates encouraged US dollar borrowing by the banks and their customers. The sustainability of such borrowing was clearly in doubt.

The devaluation and subsequent floating of the peso in December 1994 resulted in a collapse of both the exchange rate and the stock market, illustrating the close linkages between the two. Interest rates rose sharply as attempts were made to stabilise the economy, reaching a high of 80% in mid-March 1995. This increased the financial burden on bank customers and led to a sharp increase in non-performing loans, further exacerbating the problems of the banks which were finding it difficult to refinance maturing US dollar borrowing (previously backed by tesonbonos as collateral).

In the aftermath of the peso devaluation, a number of emerging markets also experienced potentially destabilising capital flows. Many fund managers trimmed their holding of assets in the developing countries, even in the more mature economies such as Hong Kong. This happened notwithstanding the much stronger fundamentals in Hong Kong. Hong Kong has maintained a trade account surplus, large fiscal reserves and the government has no external debt. Moreover, Hong Kong has the seventh largest foreign exchange reserves in the world, amounting to US$54 billion at the end of June 1995.

While the bulk of the outflow from Hong Kong was associated with portfolio fund repatriation, there were also signs of speculative attacks on the HK dollar. However, speculators were fended off quickly when the Hong Kong Monetary Authority tightened conditions in the interbank market, thereby imposing a high funding cost on those shorting the HK dollar. The action succeeded in stabilising the HK dollar exchange rate in the space of a day or so and interbank rates soon eased back. Since the tightening only lasted for a very short period of time, the banking sector's retail deposit and lending rates were unaffected and there were no knock-on effects on the banks and their customers.

The Mexican crisis and its aftermath illustrate a number of the themes of this paper:

  • the volatility and reversibility of short-term portfolio flows in a deregulated international environment
  • the ease with which both domestic investors and international fund managers can undertake such portfolio shifts in a deregulated environment
  • the impact of reversal of portfolio flows on financial markets and on a banking system which has geared up on earlier inflows and has "invested" them in relatively illiquid assets (such as loans to the private sector)
  • the possibility of contagion spreading to other financial markets even those where the fundamentals are good


Volatility in financial markets

Equity Markets

Equity markets in East Asia have expanded rapidly in recent years (see Chart 1). Hong Kong, Malaysia, Korea, Singapore and Thailand are now amongst the world's 20 largest in terms of market capitalisation. Increased foreign portfolio investment, relatively stable exchange rates (as compared to the past) and privatisation of state enterprises are some of the main driving forces for this development.

Equity prices in almost all major developing economies experienced substantial rises in 1993 partly due to the impact of capital flows as well as strong domestic demand. However, the markets turned around completely following the US monetary tightening in early 1994. During the last quarter of 1993, share prices in all major stock markets in East Asia surged rapidly, with increases ranging from 20% in Korea and Singapore to 73% in Thailand, compared to 12.5% in the UK and 5.6% in the US. In the following quarter when markets suffered significant reversals, the fall in stock prices in East Asia was also more pronounced than those in the developed countries (see Table 4). During the quarter with the exception of South Korea, share price indexes fell by amounts ranging from 14% in Taiwan to 26% in Thailand whereas there was only a 3% fall in the Dow Jones Industrial Index. Changes in overseas sentiment were certainly a factor in some of these movements. In Hong Kong, the IMF has noted that US investors purchased a net US$674 million of Hong Kong shares in December 1993, but sold a net US$708 million the following month.5 This helped to trigger a decline in share prices in the first half of 1994, which was given added impetus by the rise in US interest rates.

A number of Asian equity markets were also affected in the first quarter of 1995 by the spillover from the Mexican crisis.

The more volatile price movements of the emerging stock markets is illustrated by the larger standard deviations of the movements in their price indices compared to those of the developed markets (Table 5). This will reflect the disproportionately greater impact of external flows on the former markets as well as their relative lack of liquidity.


Bond markets

Bond markets were also highly volatile in 1994. On crude estimates, capital losses in the world bond markets might have reached US$1.5 trillion in 1994, roughly equivalent to 10% of OECD GDP during that year. Chart 2 shows that the US bond yield increased sharply between the last quarter of 1993 and September quarter of 1994. The rise in bond yields was at first associated with rising volatility, particularly in the March quarter of 1994. The volatility started to decline in the June quarter of 1994 but it remained notably higher than in previous years during most of 1994. The volatility has increased sharply again since mid-1995, but this time the movement was associated with a sharp fall in bond yields and markets did not experience the same kind of turbulence as in 1994.

The development of bond markets in the Asia economies only took off in the late 1980s and the size is rather small compared to the equity market capitalisation. By 1994, the total size of the Asian bond market (excluding Japan) was US$338 billion, about 2% that of the size of industrial countries. In past, the slow growth of the bond markets has reflected the lack of funding requirements on the part of Asian governments, such as those in Hong Kong, which have pursued prudent fiscal policies. However, the growing demand for long-term finance to support large-scale infrastructure projects in the Asia region is likely to give a substantial boost to bond market development. Moreover, the development of the pension fund industry, as in Hong Kong, will increase the supply of funds seeking a home in long-term investment.


Capital Flows and Volatility

One distinct feature of the volatile movements in the financial markets in 1994 was that they were more synchronous across countries and their consequences were further-reaching than in the past. This may reflect the increased influence of US interest rates on the emerging markets. One possible reason for such a development is that more Asian and Latin American currencies are linked to the US dollar, either directly or in the form of a managed float.

Recent research on the transmission of the volatility from the US to Pacific-Basin suggested that cross-country investment flows play a very important role in the process6. The effects of movements in US interest rates on those in emerging markets are more pronounced in countries with little restrictions on capital flows and less fluctuation in their nominal exchange rates. Empirical evidence shows that the importance of US interest rate shocks on the variance of domestic interest rates in the Asian economies has increased since 1989, especially in Indonesia, Korea and Thailand where measures were taken to stabilise their exchange rates and liberalise their financial markets7.

Rapid capital flows, especially sharp outflows, will tend to increase the volatility of markets, particularly when there is rapid withdrawal of non-resident funds at times of financial panic. This is likely to be more acute in the less liquid emerging markets. Moreover, liberalisation of capital outflows may also facilitate domestic investors selling their holding of domestic assets to acquire foreign assets at times of financial panic and this may also contribute to the increase in volatility.

Measures to increase the liquidity of securities markets should help to reduce this volatility and enable inflows and selling pressure to be more easily absorbed. In order to broaden the investor base, measures must be introduced to ensure adequate disclosure of information by listed companies, discourage insider trading and protect minority interests. Market infrastructure must also be enhanced to ensure transparency of pricing and to enable trades to be speedily contracted and settled. In Hong Kong, for example, a book entry computerised clearing and settlement system run by the Hong Kong Monetary Authority has been introduced on which both Exchange Fund bills and notes and private sector HK dollar paper can be settled. This has been linked with both Euroclear and Cedel, giving overseas investors easy access to the HK dollar debt market. The next stage is to develop a Real Time Gross Settlement System for clearing interbank payments, which should be ready by end-1996. This will not only strengthen the banking system, by largely eliminating payment risk, but also will eventually allow delivery against payment for securities transactions.

There is however something of a paradox in relation to measures to improve the liquidity of securities markets. Illiquid markets tend to lock investors in by increasing the transactions costs of selling. By contrast, liquid markets make it easier and less costly for investors to sell and thus may facilitate outflows and increase the selling pressure on the currency in a crisis. Overall, however, it would appear that liquid markets bring overall benefits to the economy in terms of improving the efficiency of the capital markets in resource allocation and reducing capital costs.


The impact of derivatives

The growth of derivatives is the major example of the innovation which has occurred in financial markets over the last decade. Between 1990 and 1994, the total notional principal of outstanding exchange traded derivative contracts rose from US$2.3 trillion to US$8.8 trillion. During the same period, the principal of OTC interest rate and currency swaps rose from US$3.5 trillion to US$8.0 trillion.9 Derivatives are financial contracts whose value depends on the values of one or more underlying assets or indices. They include such instruments as swaps, futures and options. Derivatives enable the risks in financial transactions to be unbundled and to be transferred to those parties who are more willing or able to bear them. This ability to transfer risk, along with the leverage and lower transaction costs in derivatives, would be expected to encourage market participants to enter into transactions which they would not otherwise have undertaken, thus increasing the volume of such transactions and improving the liquidity of the markets to which the derivatives are related.

From this perspective therefore, it is likely that the growth of derivatives has been a factor in the increase in international capital flows, eg by enabling investors in foreign capital markets to more easily hedge their currency risks, and have generally helped to moderate market volatility (through the favourable impact on liquidity).

The effects of derivatives are not however wholly benign. A recent report by a working group of the G10 Euro-currency Standing Committee10 has noted that in times of stress derivatives may amplify price movements, eg through the impact of dynamic hedging techniques, particularly when these are undertaken by a number of market participants simultaneously. Moreover, derivatives extend the range of instruments available to undertake speculative attacks against a fixed exchange rate, eg through the purchase of put options. Because such transactions do not immediately make themselves felt in the cash market, they may conceal the onset of an exchange rate crisis but intensify the selling pressure when the options are triggered.

There are also concerns at the micro level that individual market participants may over-extend themselves in trading in derivatives and that the resultant collapse may trigger off a chain reaction among counterparties, producing a systemic crisis. While the Barings collapse lends some support to this view, there is a consensus that the problems there were due to a failure of internal controls rather than to derivatives per se (see below). Moreover, the Barings collapse was quickly resolved with little contagion to other market participants. The speed with which Barings' trading positions were unwound and resolved contrasts with the much more intractable problems that are caused by bad loans, the more traditional cause of banking difficulties (as seen currently in Japan).

There are therefore different schools of thought on the regulation of derivatives. On the one hand, there are those who argue that the risks involved - credit, market, liquidity, operational etc - are not really different from those which apply to other types of financial instruments. This would argue for little additional need for regulation. Against this, the risks in derivatives can be packaged together in more complex ways and even the more simple types of product may be only imperfectly understood by senior management of financial institutions which trade them (as the Barings case seems to suggest) and by the end-users (Orange County, Metallgesellschaft, Procter & Gamble etc). This leads towards the view that trading in derivatives should be more strictly controlled and even prohibited for certain types of financial institution and end-user.

Knowledge of derivatives cannot be "unlearned" however and an overly heavy-handed regulatory approach toward derivatives in today's globalized environment is likely to drive trading activities into unregulated entities and/or markets. The main focus of supervisory effort therefore in respect of derivatives is to foster prudent risk management by the market participants themselves, as described below in respect of banks.


Problems arising from capital flows

Capital flows can give rise to a number of macroeconomic difficulties. If the objective is to maintain a stable exchange rate, which has generally been the case with Asian economies, the authorities may be obliged to intervene in the foreign exchange market and to sterilise the monetary impact in an attempt to insulate the domestic market from the increase in liquidity that would otherwise result. However, this may involve substantial cost arising from the difference between the cost of the debt issued by the monetary authorities and the return earned on the increased foreign exchange reserves. Moreover, attempts to mop up surplus liquidity can force up interest rates, thus encouraging more inflow. However, to the extent that the inflows are only partly sterilised, domestic credit will be increased as was the case with Mexico.

Conversely, allowing the exchange rate to float upwards will lead to an appreciation in both the nominal and real exchange rates which will create problems for the export sector and may conflict with the Government's objectives for export-led growth.

Of course, there are also be substantial benefits from capital inflows in terms of encouraging increased investment and the development of the financial markets. The resultant integration into the global capital market will however be at the expense of some loss of discretion to pursue an independent monetary policy, particularly under a stable exchange rate regime. This may be a useful market discipline, but markets have a tendency to overshoot. Thus, policy mistakes will tend to be punished more quickly and disproportionately by capital flight. This reinforces the point, although it is not the main focus of this paper, that the best way to maintain the stability of the financial and banking systems is to pursue sound macroeconomic policies. From this perspective, a sound supervisory system is a vital but secondary issue.


Adverse impact on banking sector

Volatile capital flows will, in particular, tend to cause problems for the banking sector because of its central role in the financial system. Even if bank deposits are not the ultimate home for capital inflows, the banks provide the channel by which funds will flow into and perhaps more importantly out of the country. Banks will also be affected by any more general deterioration in macroeconomic conditions arising from volatile capital flows both through the effects on their customers and on their own holdings of other financial assets. Once individual banks become affected by such problems there is a danger of the contagion spreading thorough the payments system and interbank markets, raising the possibility of a systemic crisis. This may in turn set up a vicious circle further undermining depositor confidence and giving new impetus to capital flight. A robust banking system is thus an essential element in enabling the authorities to cope with capital flows and attendant instability in the capital markets. Before going on to consider how this objective can be achieved, it is necessary to review briefly the specific risks to which the banking sector is exposed:

- as already noted in the case of Mexico, failure to fully sterilise the monetary impact of intervention in the foreign exchange market will add to the liquidity of the banking system and encourage expansion of lending. Rapid expansion of bank loans is usually a warning signal because it tends to be associated with a lowering of credit standards as banks lend to customers whose creditworthiness is more marginal. If the banking system is in the process of being deregulated while this domestic credit expansion is taking place, the credit risk of the banks will be increased because lending skills may not have developed sufficiently to cope with the more competitive environment.

- credit risk will be accompanied by liquidity risk if the foreign funding on which the banks have relied shrinks in a crisis and the loans which the banks have made cannot easily be unwound. In a world of volatile capital flows the ways in which banks match sources of funds and uses of funds become vitally important.

- even if foreign funding can be obtained it may become more expensive as lenders insist on a risk premium. The banks will be exposed to additional interest rate risk if the authorities are forced to raise interest rates to protect the currency in the face of inflows or to stabilise the economy after a devaluation. This problem will be compounded if the liabilities side of the balance sheet has been made more interest rate sensitive through the removal of interest rate controls. Moreover, to the extent that the banks pass on higher cost of funds to their customers, this will weaken the latter's financial position and lead to a further deterioration of asset quality. The present situation of the Mexican banks is a graphic illustration of these risks.

- if the banks or their customers have been encouraged by relatively high domestic interest rates to borrow in foreign currencies and to switch into the domestic currency they will also face a foreign exchange risk if the domestic currency is subsequently devalued. Even if the banks themselves are running a fully or largely matched foreign currency position, they will be exposed indirectly if they have lent in foreign currencies to their customers whose revenues are largely in the domestic currency and whose repayment liability in that currency will have been increased as a result of the devaluation. Again, this was a feature of the Mexican situation.

- the operational and legal risks faced by banks may also be increased in a crisis, e.g. if the settlement or payments systems breakdown.

- finally, capital inflows may lead to asset price inflation, particularly in the property and stock markets, with a risk of a bursting of the bubble if the inflows are reversed. Banks are typically heavy lenders in the property market and may hold securities for investment and trading purposes. They may also hold such assets as collateral for loans. Severe volatility in capital markets thus exposes the banks to both market and credit risks. The failure of Barings Bank shows how a bank can be brought down almost overnight through improperly controlled trading activities.

This is quite a catalogue of potential dangers for banks: credit, liquidity, interest rate, exchange rate, market, legal and operational risks. Banks are being increasingly exposed to them through the forces of globalization, liberalisation and innovation. In other words, the same trends which have encouraged the development of global capital flows have also increasingly freed banks from direct controls on interest rates and on lending and trading activities. This freedom has brought about increased opportunities for the banks and has been a powerful stimulus to increased efficiency and innovation in the banking sector. However, it has also brought new risks as the banks have not always used their new freedoms well. In particular, risk management techniques have not always kept up with the more competitive and complex market conditions which result from deregulation.


Supervision in a deregulated environment

There is thus another paradox that deregulation must be accompanied increased prudential supervision if the costs are not to outweigh the benefits. This is however easier said than done. Much attention has been focused in recent years on the risks of deregulation in emerging markets which may leave their banks exposed to both internal and external shocks such as a tightening of monetary policy or a reversal of capital flows. However, it is a fact that problems from deregulation have tended to arise also in developed countries with well established banking and supervisory systems. Of course, it is always possible to look back after the event and see that those systems did in fact contain flaws which contributed to the problems. The more difficult task is however to detect such flaws beforehand and undertake pre-emptive reforms which will reduce the risks from deregulation in advance.

The main characteristics of a supervisory system which might achieve this objective are described below. But first it is necessary to make the distinction between regulation and supervision or to put it another way, between direct and indirect controls. Put briefly, regulation involves telling the banks what they can and cannot do in terms of running their business. Supervision involves the establishment of a framework of rules and guidelines which are intended to establish minimum standards of prudent conduct within which banks should be free, or at least more free, to take commercial decisions. The latter approach allows the banks greater freedom to allocate credit in line with their own commercial judgement and is thus consistent with the general trend towards a market economy which is a feature of the Asian region.

The change in approach is indeed unavoidable. Direct controls on the banking system become more difficult to maintain when the rest of the economy is being deregulated. Lending and interest rate ceilings will prompt disintermediation from the banking system, typically into unregulated or loosely regulated "fringe" institutions. Such institutions may turn out to be the cause of financial sector instability. Freedom of capital flows also enables borrowers and lenders to seek or place funds abroad with potentially destabilizing effects at time of stress, as in the case of Mexico.

This is not to say that there may not be a case for some direct regulatory controls, for example where there are market imperfections or to cope with sudden emergencies such as a sharp increase in capital inflows. Moreover, there is always a danger in removing existing controls in that competitive pressures may cause markets to overshoot. Even in Hong Kong, which is largely liberalised, there are still direct controls on the interest rates on certain types of retail deposit, though these are currently being carefully dismantled. Moreover, while entry into the Hong Kong banking market is available to those institutions which satisfy the prudential criteria, they are restricted to one branch in Hong Kong in order not to create excessive competition for retail banking deposits.

Nonetheless, the general trend, and indeed the necessity, is to move towards what the Chairman of the Basle Committee, Dr Padoa-Schioppa, has called "market-friendly supervision".11 Of course, all supervision is intended to be market-friendly in the sense that it is directed towards maintaining the soundness of the financial market system. But the essence of the new approach is to look for solutions to supervisory problems within the market itself rather than by imposing a "top-down" approach by the supervisors. Looked at from this point of view, the role of supervision is, in Dr Padoa Schioppa's words, to strengthen "the disciplinary factors already present in the banking firm and the market, and to complement them where they fail or where their failure would entail undue social costs". This market-based approach is particularly relevant to dealing with today's more complex world of free capital flows and fast moving financial market conditions and trading portfolios. It is thus particularly appropriate for dealing with the supervision of derivatives and other trading activities by banks.


The elements of an effective supervisory system

As already indicated, the prudent management of capital flows and of market volatility requires that the whole supervisory approach is effective.

There are three main elements to this:

  • the legal framework
  • the policy framework
  • supervisory enforcement

These will be summarised in turn before discussion of the specific aspects which are relevant to coping with the problem of capital flows and volatility in capital markets.


The legal framework

It is necessary that there should be a solid legal framework of banking legislation which clearly defines the objectives, powers and responsibilities of the banking supervisor. This should also restrict the business of carrying on of banking business and the taking of deposits to a clearly defined group of authorised institutions. Care should be taken to include "fringe" deposit-taking institutions within the scope of the supervisory net since the experience of many countries, including the UK, Hong Kong, Malaysia and more recently Japan shows that financial sector problems may frequently arise in such fringe institutions if they are subject to little or no regulation.

The legal framework must also impose controls on the ownership and management of banks since this is likely to be a major determinant of whether the banks are prudently and honestly managed and the interests of depositors are safeguarded. The supervisors must be given the powers to set minimum standards of prudential conduct, relating to such matters as capital adequacy and liquidity, and to monitor adherence to these standards by gathering information and conducting examinations. Finally, the supervisors must have the ability to deal quickly with problem institutions by taking control of them or in the last resort closing them down.


The policy framework

As already indicated, the banking legislation should give the supervisors the backing to be able to construct a policy framework of supervisory rules and guidelines. The objective is to control excessive risk-taking by banks and to reduce the chances of bank failure. These supervisory policies will be discussed in more detail later on, but in general the main elements can be summed up in the familiar CAMEL rating which stands for Capital, Asset quality, Management, Earnings and Liquidity. Many countries rate their banks against these headings or a variant of them and this can provide a useful means of assessing the relative quality of banks and identifying those which require special supervisory attention.

One important aspect of supervisory policies is that they are becoming increasingly internationalised, largely thanks to the work of the Basle Committee on Banking Supervision. This is most apparent in the widespread adoption of the risk-based Basle capital adequacy ratio, but it also applies in many other supervisory areas. The quality of individual countries' supervisory systems is thus increasingly judged by reference to international standards. It is therefore important to be seen to be adhering to these standards as a means of attracting and, more importantly, retaining international capital flows through the domestic banking system.


Supervisory enforcement

It is actually quite easy to construct what looks like an effective supervisory system on paper. What is more difficult is to ensure that the system is effective in practice and is properly administered. This partly depends on the method of enforcement. Most countries have tended to opt for inspection-based systems on the basis that there is no substitute for being able to go into banks on a regular basis to examine directly their books and records and internal control systems. However, this also needs to be backed up by effective off-site analysis of statistical returns and other information so that broader picture can be obtained of the strengths and weaknesses of both individual institutions and of the banking sector as a whole. Whichever blend of methods is used it is obviously essential that the supervisory function has the right number and type of staff to do the job and they are given adequate resources and training. It is also necessary as noted earlier to cast the supervisory net wide enough to ensure that it covers also "fringe" banks where problems might otherwise arise. (In Hong Kong, for example, deposit-taking companies have been brought within the scope of the Banking Ordinance and are supervised largely on the same basis as banks.)

An effective supervisory system along the general lines described above is a prerequisite for dealing with the problems of volatility of capital flows and of capital markets. But there are a number of specific measures and policies which can be adopted to deal with these problems.


Capital adequacy

Capital adequacy has been the main supervisory preoccupation in recent years. Capital provides banks with a cushion against losses and helps to instil confidence among depositors and other counterparties. As already noted, the Basle Committee's risk adjusted capital adequacy regime, incorporating an 8% minimum ratio, has effectively become the world standard. This has helped to establish a more level playing field in the increasingly global banking market and has prevented banks with inadequate capital ratios from establishing a competitive advantage over their peers. Banks need to ensure that the higher level of capital they are required to maintain is adequately remunerated so that there is some disincentive to engage in aggressive cutting of margins. At the same time the need to preserve an adequate margin of comfort above the 8% minimum ratio places some constraint on over-expansion of the balance sheet which might otherwise arise, for example from an increase in domestic liquidity as a result of inflows.

Capital adequacy is not of course a sufficient assurance by itself of the health of a bank. The other elements of the CAMEL rating also need to be taken into account. In particular, the capital ratio will be meaningless if the assets of the bank, and particularly its loans, are not properly valued. Despite the growth of more complex, securitised instruments which will be discussed later, credit risk is still the main risk to which banks are exposed. The risk is likely to be most acute when competitive conditions are in a state of flux, for example because of deregulation or because of rapid growth in domestic credit expansion.

In a market economy, credit assessment is, or should be, a matter for the banks. However, the supervisors can try to ensure that banks do this within a prudent framework of internal controls for measuring, monitoring and controlling credit risk. The principal means of doing this is through regular examination by supervisors of loan books and lending systems. However, an industry-wide loan classification system based on standard definitions, such as has recently been introduced in Hong Kong, is also an important preventative measure. Such systems require the banks to classify their loans into such categories as performing, special mention, substandard, doubtful and loss. This makes it easier for the supervisors to conduct peer group comparisons, in particular of the adequacy of provisions against non-performing loans, and also to gain an industry wide picture of asset quality and changes in the overall trend.

A further step is for the supervisor to mandate the minimum level of provisions which have to be held against the unsecured portion of the various loan categories, for example 50% against the unsecured portion of "doubtful" loans. This may be a useful discipline when bank management cannot be relied upon to make such judgements themselves. Equally, however, there is a risk that the stipulated provisioning level will come to be seen as a norm rather than a minimum.

Whether or not such norms for provisions are mandated, the question of loss recognition in respect of bad loans is a crucial issue. The question, as discussed recently by the IMF,12 is whether supervisors should pursue a policy of forbearance which allows institutions to write-down their losses over time in the hope that they can grow their way out of trouble (the Japanese solution), or whether losses should be recognized more immediately. The latter carries the risk of a loss of confidence among depositors and may require financial assistance from Government to the banks whose capital has been impaired. However, once the initial pain has been suffered, the banks should be on a sounder footing. The policy of forbearance prolongs the agony and increases the chances that the ultimate losses may be higher.


Risk concentrations

The supervisor can also try to ensure that banks do not incur undue concentrations of risk at the geographical, sectoral and individual borrower level. Exposure to the property and stock markets can be a particular risk if speculative bubbles develop as result of inflows. Property lending in particular has been a common factor in the banking problems which have affected both developed and developing countries in the late 1980s and 1990s. This reflects the fact that the property market is cyclical and the banks' lending to it is generally long-term. During the life of a typical property loan, its quality may be adversely affected by factors over which the lending bank has little control, such as the general level of interest rates and the volume and direction of capital flows. Moreover, the funding of property lending will generally involve running a liquidity and interest rate mismatch. These risks may become acute if capital flows reverse and interest rates have to be raised sharply. Even if the rates on property loans such as residential mortgages are variable at the discretion of the banks, there may be practical constraints in raising sharply the interest cost of such loans. By doing so, banks may simply substitute credit losses for interest rate losses as borrowers become unable to service their loans.

In Hong Kong banks have a large exposure to the property market, principally in the form of residential mortgages of which they are the only significant provider. Such lending is of high quality as evidenced by the loan default rate of less than 0.5% of the outstanding amount. Nonetheless, an element of concentration risk remains. The principal means of dealing with this from a supervisory point of view has been to encourage the banks to apply prudent lending criteria. In particular, the Hong Kong Monetary Authority has supported the application by banks of a loan to value ratio of not more than 70% in respect of their residential mortgage lending. With property prices having fallen by up to 30% from their April 1994 peak, this measure has clearly provided the banks who lent at the top of the market with a prudential cushion during the downturn.

The Hong Kong Monetary Authority is also considering means by which the liquidity of the banks' mortgage portfolios could be improved with the aim of reducing liquidity and interest rate risk. The market solution to this problem is of course securitization and a number of issues of mortgage-backed securities were made in Hong Kong during 1994. However, in general the market for mortgage-backed securities has tended to lack liquidity and to be somewhat fragmented. The Monetary Authority is therefore looking at means by which market liquidity could be increased. One possibility which is being considered is to establish a Mortgage Corporation, along the broad lines of the US Fannie Mae, which could acquire mortgages from individual banks and package them into homogeneous securities which would be more acceptable to investors. If successful, this would reduce the liquidity risks and concentration risks of residential mortgages, thereby contributing to the stability of the banking system.


Liquidity

Prudent management of liquidity risk also requires more direct measures. The maintenance of adequate liquidity is the first line of defence for a bank in protecting itself against an outflow of funds. There are however problems in devising a supervisory policy for liquidity. First, there is not accepted international standard for setting liquidity requirements as there is for capital adequacy. Secondly, the appropriate liquidity requirements vary widely between different types of bank, notably between those carrying on retail business and wholesale business. Further complications arise in the case of branches of foreign banks where it is difficult to distinguish the liquidity needs of the branch in the host country from those of the bank as a whole.

Having said this it is possible to lay down certain basic principles which should govern prudent management of liquidity. In particular, in Hong Kong, the liquidity regime aims to ensure that:

Banks can meet their obligations as they fall due under normal circumstances

- banks have the ability to cope with an abrupt withdrawal of funds under crisis conditions

The first of these objectives requires the banks to manage their funding and cashflow in a prudent manner on a day-to-day basis. This means having regard to maturity mismatches which are being run, not putting too much strain on borrowing capacity in the interbank market, trying to ensure a stable and diverse deposit base and maintaining a sensible loan to deposit ratio.

The second objective involves the holding of a stock of high quality liquid assets which can be run down in a funding crisis in order to provide a breathing space during which liquidity support can be obtained from other sources. In Hong Kong 's experience, an institution's ability to survive a funding crisis depends to a great extent on its capacity to meet depositors' withdrawals in the first few days. If the bank is basically sound and can get through this period, a run will tend to die out naturally.

In Hong Kong the banks are required to maintain a liquidity ratio of at least 25% of their 1-month liabilities.. There is no requirement to hold liquidity in Government debt, though this is encouraged (partly as a means of obtaining access to the discount window). The mandating of holdings of a particular amount of Government debt is however one means which a number of countries have used for coping with capital inflows and in particular with the sterilization problem mentioned above. Increased reserve requirements on banks may also serve the same purpose, particularly if these can be targeted specifically at deposits taken from non-residents. This may achieve the dual purpose of forcing the banks to hold more liquidity (albeit at some cost to their profitability if the required reserves pay low or no interest) and at the same time discouraging inflows. Limitations on foreign borrowing by banks also fall into this category and may also help to protect the banks from exchange rate risk (e.g. the foreign currency liabilities of the Mexican banks were limited to 10% of total peso and foreign currency liabilities which seems to have helped to insulate them to some extent from the direct impact of the peso devaluation).

However, as pointed out earlier, such controls on inflows may be difficult to maintain in a deregulated environment since the scope for circumvention is much greater.


Market risk

As already noted, credit risk remains the main risk to which banks are exposed. However, the trend towards securitization, the increasing involvement of banks in proprietary trading in both on and off-balance sheet instruments and the increased volatility of interest rates, exchange rates and capital markets has meant that banks have become more exposed to market risk.

"Market risk" is defined as the risk of losses in on and off-balance sheet positions arising from movements in market prices, interest rates or exchange rates. This type of risk is presently only partially catered for in the Basle capital adequacy framework which is mainly concerned with credit risk. The Basle Committee has however been working for a number of years on a system for measuring the capital requirement in respect of market risk. This will not be described in any detail. Suffice it to say that the current proposals would apply a capital charge to the risks pertaining to debt securities and equities in the trading book and to the foreign exchange risk and commodities risk throughout the bank.13 Derivatives of these items would also be included.

The new regime will clearly be an improvement on crude supervisory systems which attempt to control market risk by placing limits on the size of exposures which may be undertaken in the securities and foreign exchange markets. (In Hong Kong, for example, banks are generally required to limit their foreign exchange open positions to not more than 5% of capital; although those with a high degree of market proficiency might be allowed to go as high as 15%.) However, the new arrangements are not without their own difficulties as is evidenced by the length of time which it has taken for them to be produced. The problem faced by the Basle Committee has been how to keep pace with the rapid development in financial markets and instruments, such as derivatives, and in market participants' own techniques for managing risk.

The Basle Committee's original proposals, as embodied in its so-called "standardised approach", were complicated in their own right. They were however criticised by major players in the market as providing an insufficiently accurate measurement of market risk and insufficient incentive for institutions to improve risk management techniques. As a result, the Basle Committee has agreed that the more sophisticated banks may use their own proprietary risk management models to calculate the capital requirements for market risk. This would be subject to a number of quantitative and qualitative criteria laid down by the Basle Committee, including that there should be an independent risk control unit with active involvement of senior management in the process. This use of in house-models is seen as a major innovation by the Basle Committee, consistent with the move towards "market-friendly" solutions which has already been described. Even this approach is not wholly uncontroversial. The results produced by different "value-at-risk" models* may vary for the same portfolio, thus giving rise to competitive inequalities. Moreover, the Basle Committee's proposal that the "value-at-risk" produced by the models should be multiplied by a safety factor of at least 3, has led to criticism from international banks that the capital requirements will be too onerous and could vary widely if different supervisors apply different safety factors (i.e. higher than 3).

The Basle Committee's proposals are due to be finalised at the end of this year and to be implemented by the end of 1997. The proposals are initially directed towards internationally active banks, particularly those from the countries represented on the Committee. However, they are likely to become more generally applicable, including to countries in the Asian region. We have already seen how the original 1988 Capital Accord has become a world standard. The same is likely to happen with the market risk proposals. This poses a number of challenges for supervisors. They will have to extend their own skills and techniques to be able to understand the types of instruments which are being traded and to validate the "value-at-risk" models which the larger institutions are using to measure the risks. They will also need to cater for the fact that less sophisticated institutions will begin to operate on a greater scale in the capital and foreign exchange markets. The model-based approach towards capital adequacy will not be suitable for such institutions.

These issues are being addressed by the Hong Kong Monetary Authority through setting up a specialist group of supervisors who understand the trading and mathematics of traded instruments, including derivatives. They will be used to conduct on-site examinations of trading operations and to devise policy guidelines for such activities. As regards the capital regime for market risk, a three-tier approach is envisaged. This would allow the more active and sophisticated institutions to use their own internal models, while the less active would be subject to the Basle standardised regime. Those institutions with negligible trading activities would probably be granted a complete exemption from the market risk requirements, except possibly in respect of foreign exchange exposure since this is an area in which banks have traditionally been involved.


The emphasis on risk management

The two trends which have been identified, namely the move away from direct regulatory controls and the increasing involvement by some banks in more complex trading activities, have led to an increased emphasis on the importance of prudent risk management by banks themselves.

This acknowledges that where the risk profile of banks can change more quickly and drastically than before, the traditional approach of on-site examination by the supervisors (which can only be periodic in nature) becomes less relevant. Key aspects of risk management are effective oversight of all business activities by the board of directors and senior management; a comprehensive set of risk management procedures for measuring, monitoring and controlling the various types of exposure; and an internal audit function to ensure that the procedures are working properly. The Basle Committee last year issued a guideline on risk management for derivatives which embodies these principles and on which the Hong Kong Monetary Authority has based its own guidelines on the subject in Hong Kong.14 Similar guidelines were also issued at the same by the Technical Committee of the International Organisation of Securities Commissions. This is a good example of supervisory co-operation aimed at working towards a level playing field for banks and securities companies doing the same type of business.

Prudent management by the banks themselves can be enhanced by the forces of market discipline, eg through improved disclosure by banks of their financial performance and risk exposures (including in respect of off-balance sheet activities). If a bank's performance and appetite for risk can readily be compared with those of its peers, this should provide an incentive for management to achieve better results without taking undue risks. Hong Kong has in the past been near the bottom of the league table in terms of the quantity and quality of financial disclosure by banks in their annual accounts. However the joint efforts of the financial regulators in Hong Kong have required banks to greatly increase the amount of information disclosed in their 1994 accounts, including disclosure of actual profits before transfer to inner reserves and the amounts of such transfer. Additional items of disclosure will be introduced in the 1995 accounts.

The importance of sound risk management and co-operation among supervisors in a more volatile and globalized trading environment was clearly demonstrated by the collapse of Barings Bank in February of this year. The key lessons of that episode are not so much that derivatives are intrinsically dangerous but that the risks need to be properly controlled. In the Barings case, it is evident that top management did not fully understand the business that was being undertaken in their Singapore subsidiary.15 Moreover, the internal controls in that subsidiary were fundamentally lacking, in particular as regards the degree of separation between the front and back offices and the arrangements for making margin payments.

There are also lessons for the supervisors. In particular, they need to have a full understanding of the risks in non-banking as well as banking businesses within a group. There also needs to be close collaboration between the different regulators both within a country and overseas.


Cooperation between regulators

Similar conclusions have been drawn by the Tripartite Group of bank, securities and insurance regulators which was brought together by the Basle Committee to address a range of issues relating to the supervision of financial conglomerates. The Group's report points out that the deregulation of domestic markets and growing internationalisation of markets over the past decade has led to the emergence of corporate groups providing a wide range of financial services including banking, securities business and insurance.16 Many of these operate across a wide range of countries. The fundamental recommendation of the Group is that the supervision of the individual entities within a conglomerate must be complemented by an assessment from a group-wide perspective. Clearly, this requires cooperation between different supervisions and the ability (and willingness) to exchange information. A "lead supervisor" may need to be appointed to gather group-wide information and to coordinate supervisory action affecting different entities in a conglomerate. The responsibilities of individual supervisors may be embodied in a Memorandum of Understanding ("MOU"). The Hong Kong Monetary Authority and the Securities and Futures Commission have recently concluded such a MOU within Hong Kong in order to define their respective supervision responsibilities more clearly, provide for exchange of information and minimise supervisory overlap and underlap of institutions carrying on both securities and banking business.

Such MOUs can be useful in both the international and the domestic context. However, their limitations need to be clearly understood. In particular, individual supervisors cannot simply divest themselves of, or delegate, their statutory responsibilities through such MOUs. Moreover, exchange of information must be subject to certain safeguards, including that there are adequate means for the recipient supervisor to maintain confidentiality requirements.

Latest Speeches
Last revision date : 02 September 1995