The evolving global financial system: some implications for emerging markets

Keynote lecture by William R White, Economic Adviser, Bank for International Settlements, delivered at the Indira Gandhi Institute Mumbai, India, 13 December 2001.

1. Introduction

It is a great pleasure to be here at the Indira Ghandi Institute to give this keynote lecture. And may I add that I feel it is also a great honour to have been asked, given the illustrious names that have also shared this responsibility over the last few years. In that respect, I would like to thank Dr  Reddy of the Reserve Bank of India for having suggested my name and also Professor Radhakrishna for having organised my visit.

The title I have chosen for my paper is rather a broad one so let me state more directly the specific themes I intend to address. They will, I hope, prove relevant to the papers and discussions that will follow over the next few days. The first theme is that of capital flows into and out of emerging market economies. What can be done so that emerging market economies get the greatest benefits of such flows at the least cost? The second theme also refers to capital, albeit with the word used in a rather different way. In particular, I would like to spend some time talking about the newly proposed Basel capital standards and their applicability to emerging market economies such as India. However, by way of introduction to this second theme, it will be necessary to raise the question of international standards and codes more generally. As I am sure you are all aware, a dauntingly large number of these codes and standards have been developed in recent years, and India has been analysing their usefulness and practicality perhaps more seriously and systematically than any other emerging market economy.

Both of the themes I intend to explore today, as the title of my paper suggests, are sub-themes of a broader set of issues pertaining to the ongoing process of globalisation. Many books have already been written on this topic, some stressing its advantages and others its disadvantages. Indeed, it seems to me at least arguable that the war currently taking place in Afghanistan may well be a manifestation of this underlying conflict. However, whether in the realm of politics, trade, financial services, culture or whatever, the policy issues always come down to the same questions. Does India wish to encourage its integration with the global community, or does it wish to resist such forces. And if it does chose the path of liberalisation and integration, how fast should it go down that path?

Answering such questions always demands an assessment and weighing up of conflicting arguments. This is inherently difficult. Moreover, when we address the specific issue of liberalisation and globalisation in the financial sphere, such an evaluation is rendered still more difficult by the fact that there is inherently no right answer, only tradeoffs.

The static tradeoff is well known in the literature, and essentially involves balancing off the efficiency of resource allocation on the one hand, against safety and stability on the other. Evidently, a liberalised financial system gives more of the former and a repressed financial system more of the latter. James Tobin once famously said "It takes a heap of Harberger triangles to fill an Okun gap", but equally well-known economists have also expressed much more market friendly views. In part this may be because there is also a dynamic tradeoff as Joseph Schumpeter reminded us when he coined the phrase "creative destruction". Liberalised financial systems allow credit to be given to a much wider cast of characters, particularly if a culture of venture capital has grown up along side it. New firms with new ideas provide the basis for innovation that pushes up the rate of growth of productivity over time. The downside is that liberalised systems of this sort often over finance good ideas as well as many bad ones. This can lead to excessive competition, the bidding away of both rents and profits, and sharper cyclical downturns than would have been the case given a more repressed financial system.

Making these policy tradeoffs in the real world is further complicated by the fact that perceptions concerning these tradeoffs can, and do, change over time. Periods of financial instability lead to an enhanced focus on the costs of such instability and enhanced efforts to minimize them. This is what happened in light of the financial turbulence of the 1930s. Conversely, after long periods of financial stability, the resource costs of financial repression come to the surface. For a time, deregulation then tends to be the order of the day. In the industrial countries such a process began in the 1960s in North America before culminating with the scrapping of exchange controls in Europe in the late 1980s. In light of still more recent events - the boom and bust cycles in Japan (the 1980s), East Asia (mid to late 1990s) and the United States (still unfolding) - it remains possible that the pendulum will swing once again. As also evidenced in the choice of trade and exchange rate regimes, there seems to a human psychological predisposition in play. Greater weight tends to be given to the self evident deficiencies of the current regime than to the only hypothetical deficiencies of some alternative regime. From this perspective, regime shifts will never end.

2. Securitisation, globalisation and consolidation

As a backdrop to my discussion of the issue of international capital flows and international capital standards, it might be useful to describe briefly some of the forces driving change in the international financial system. I like to distinguish between the underlying forces for change and the manifestation of those forces, as I will shortly explain. It is also worth noting that there is often a great deal of reinforcing interaction between these elements for change, implying the development of a dynamic that can prove very hard to either contain or reverse. The practical conclusion that I would draw from this is not that change should be resisted because it could get out of hand. Rather, it is that change should be introduced in a thoughtful way with due regard for both interactions and side effects. In this reflective process, the contributions of financial practitioners, academics and public policy experts may all have a constructive role to play. It is my understanding that such an inclusive process is currently being followed in India as you weigh up the merits of various kinds of international standards to pursue financial stability.

Turning first to the underlying forces for change in international financial markets, I would focus first on advances in technology which have led to a whole host of new financial products (through risk unbundling and rebundling) as well as new ways to deliver them (e-finance). Moreover, cheaper communication vehicles are providing more trading information to more people than ever before - real time, simultaneously and worldwide. Deregulation has been another potent force, particularly in emerging market economies in recent years. Generally speaking, the pattern has been a first stage of removing quantitative directives, a second stage of freeing up domestic prices for financial instruments, and a third stage of removing international capital controls. Demographic changes (especially in Europe) have also been a potent force for change in the financial industry. An ageing population has been tempted to save more and to invest more adventurously, recognizing that governments look increasingly unlikely to honour fully their social safety net promises.

Another underlying force for change has been the nexus of increased competition, the enhanced search for shareholder value and the spread of financial safety nets. Competition has been driving down rates of return on equity in the financial services industry even as shareholders have increasingly been demanding more. This has led to some degree of cost cutting and rationalisation in the industry, but may also have led to a greater willingness to take risks (eg banks lent heavily to EMEs in the 1970s, then moved on in turn to LBOs, property, proprietary trading and lending to EMEs again). These risk-taking propensities may have been enhanced by the growing perception that financial safety nets (eg too big to fail and IMF support packages) would cushion the blow were risks to actually materialise.

These underlying forces, and their interactions, have manifest themselves in three principal ways. The first I would refer to as securitisation. Technology allows the commoditisation of credit, and its associated risks, and their sale and purchase in market places which are now much better informed than ever before. In effect, relationship banking is increasingly being replaced by market-based transactions, with the associated danger that (in the limit) traditional financial institutions will be left holding only sub-prime credits. In this new world, ensuring the proper functioning of markets, particularly when markets are under stress, is becoming a high priority for those concerned with preventing financial crises.

The second manifestation is that of globalisation. Securities markets are increasingly linked worldwide as well as across market segments. In addition, more and more firms are now operating globally, and they increasingly have a local presence which offers direct competition to domestic financial institutions. Associated with most of these trends has been a sharp expansion in cross border capital flows, not only between industrial countries but emerging market economies as well. These global links have many advantages but they also ensure that crises anywhere have an easier mechanism for propagation than ever before. The particular problems associated with large inflows of foreign capital and subsequent large outflows are now well known and will be discussed further below.

The final manifestation has been the trend to consolidation in the financial services industry. A recent study by the Deputies of the G10 confirms that financial firms are getter bigger, more cross sector, more complex in terms of what they do, and increasingly interrelated (big dealing with big). On the one hand this could ensure a better overall handling of risks, if such firms are well managed. On the other hand, such firms could just as easily turn out to be too big to manage (inviting crisis) as well as too big to fail (or to save in some small countries). Winding down such a large complex firm, with widely diversified international activities subject to different legal jurisdictions, would not be an easy task in itself and would be made worse by the interfirm relationships just referred to. The implications of consolidation for operational risk, especially in so far as it affects the infrastructure supporting their financial system, has also been thrown into sharp relief by the events of 11 September.

3. International capital flows and emerging market economies

The Mexican crisis of 1994, the East Asian crisis of 1997/8 and the Russian/Brazilian crisis of 1998/9 focused international attention on the volatility of such capital flows. There was a strong sense that large inflows could create problems of inflation, asset price bubbles, misallocations of investment and subsequent problems in the banking sector. Moreover, sudden reversals of such inflows were felt to have the potential to precipitate both exchange rate crises and banking crises. In the light of these experiences, various reports (by the G10 Deputies, the Willard Group and the Financial Stability Forum) focused on how such movements and their repercussions might be alleviated. I would like to review briefly the main recommendations, even though more recent evidence indicates the possibility that emerging markets may now be facing a rather different problem. I will return to this possibility in a moment.

Looking at the issue first from a macroeconomic perspective, no one has ever questioned the reality of the "impossible trinity". That is, a country cannot have highly mobile capital flows, a fixed exchange rate and its own domestic monetary policy all at the same time. For a time, the consensus seemed to be that free floating was the preferred way to square this circle, but in the last few years the agreed position has become more nuanced. Pegged or semi-fixed exchange rate regimes encourage volatile capital flows and most observers would feel they should be avoided. Yet some form of managed floating is increasingly seen as having some attractions. A managed floating regime would presumably be premised on some domestic nominal anchor (some variant on inflation regime targeting?) but might also encompass a willingness to use monetary policy and other instruments (say direct FX intervention) to respond to exchange rate movements even when they did not threaten domestic price objectives. The underdevelopment and lack of liquidity in many financial markets in emerging market economies implies that their efficient functioning might well be a supplementary source of concern for the monetary authorities. The attitude to capital controls also seems to have softened in recent years, as the merits of only a measured dismantling of previous controls have become more evident. For completeness, I should also add that for small countries there is some evidence that harder pegs (currency boards or even dollarisation/eurosation) might have some attractions. However, this observation hardly applies to a country the size of India.

Within such a macroeconomic framework, three sets of suggestions have been made to deal with the problem of volatile capital flows. They should be treated as complementary to other suggestions to mitigate the potential damage through strengthening the financial system more generally; say, through the implementation of international standards as described below. What must also be recognized is that none of these suggestions are without potential downsides. Accordingly, some judiciously chosen combination of measures might be thought the most appropriate policy response.

The first set of proposals has to do with improvements in transparency such that the true exposure of both debtors and creditors is know to all. Behind such suggestions is the appealing notion that participants in financial markets will react more sensibly if they have better information about the current state of affairs. While likely true, the possibility of such information exacerbating herding behaviour cannot be ruled out. Moreover, it is not just the availability of information but the willingness to act on it that is ultimately important. That is a clear lesson from the Asian crisis, where BIS data on the short-term foreign exposure of the crisis hit countries was publicly available and had been commented upon (albeit discretely) by the BIS. Nevertheless, this information was systematically ignored by senior bankers, in some cases even after it had been brought to their attention by their own credit risk officers.

Regardless of these caveats, strong pressures emerged after the Mexican and Asian crises to improve the statistics pertaining to the external debts and foreign reserves of emerging market countries, and to reconcile conflicting sets of data (say creditor side versus debtor side statistics). In response, a lot of improvements have been made in this area in recent years though, as will be noted below, a number of fundamental shortcomings remain. It was also suggested at the time that more precise information should be collected about position-taking (particularly by macro hedge funds) against the exchange rates of particular countries. Here, little progress has been made although the demise of macro hedge funds in the last year or two may have rendered this less of a practical issue. Finally with respect to transparency, many commentators have called for the creation of "vulnerability indicators" for emerging market economies that would show their changing susceptibility to crises over time. In spite of great efforts by the IMF and many others, no great progress has been made, with indicators of potential banking crises being even more unreliable than indicators of exchange rate crises. In the work done to date, the greater problem has been the tendency of indicators to give false positives rather than the failure to predict crises that do in fact occur.

The second set of suggestions seek to restrain the tendency of creditors to plunge into and then back out of emerging market economies. In this regard, reliance has been put on the same three "pillars" referred to in the proposed New Basel Accord. The first of these is better internal governance, based on the recognition that loans to emerging market countries might well have special risks attached to them. This will be particularly so if, as is overwhelmingly the case, loans are denominated in a non-domestic currency. I will return to the "currency mismatch" issue below. Given the losses actually suffered in recent crises, this might seem a potent incentive to more prudent behaviour in the future. The growing unwillingness of the IMF to use public funds to shelter (at least partially) imprudent lenders will also prove helpful in this regard. However, a remaining problem for recipient countries is that even reduced exposures by a number of large creditors could still wreak havoc under appropriate circumstances. The second and third pillars are those of supervisory oversight and market discipline respectively. The problem with both of these is again related to size. Loans to any single country, or even region, are unlikely to threaten seriously the health of the creditor and thus warrant the close attention of either markets or supervisors. This is not to say that these incentive systems could not be improved. In particular, regular consultations between committees of lenders and representatives of borrowers might prove very helpful in sensitising both sides to emerging vulnerabilities. Such consultations could also help establish personal relationships and a reputation for "good faith" behaviour that might prove very helpful in the management of any crisis that did emerge.

Given the noted limitations of the first two sets of suggestions, borrowers will have to rely on protecting themselves, as best they can, against the vicissitudes of volatile capital flows. While full-blown capital controls have little appeal, some restrictive measures that fall short of this may still have some merit. Market-based penalties on short-term inflows, as practised in Chile, seem to have had some success. As well, a number of countries have recently imposed bans on lending domestic currency to foreigners wishing to sell it short. Moreover, it is common practice in many countries to prohibit domestic banks from taking net open positions in foreign currency.

In light of recent crises, most emerging market economies have taken steps to increase their foreign exchange reserves and to monitor (and manage more closely) their sovereign liabilities in foreign exchange. A recent report by the FSF (Working Group on Capital Flows) went substantially further by suggesting that governments should monitor national rather than sovereign foreign exchange exposures. Indeed, they went so far as to ask the IMF and World Bank to draw up guidelines in this regard. The principal concern expressed by the FSF was that significant "currency mismatches" by private sector borrowers could bankrupt them in the event of a devaluation. Moreover, if enough creditors found themselves in this position, they would threaten the health of their domestic creditors in turn. As valid as this concern might be, collecting the data needed to monitor such exposures adequately will not be easy. For many corporations, even on-balance-sheet exposures are opaque, to say nothing of those that are off-balance-sheet. Significantly greater attention should be paid to such corporate accounting issues, particularly by domestic lenders whose own survival might be threatened should a crisis reveal unanticipated exposures.

A further factor contributing to currency mismatches in many countries is the absence of adequate financial markets to allow corporations to borrow at home in domestic currency or to allow hedging when borrowing is done in foreign currency. Efforts to develop such markets have been strongly recommended by the FSF but such efforts will obviously take time. A further problem lurking in the background for many countries is what Ricardo Hausmann calls that of "original sin". Foreign lenders (each failing to understand the fallacy of composition in their thinking) will remain hesitant to accept a foreign exchange exposure in currencies that often have a dreadful track record. Fortunately, some recent successes in floating longer-term bonds denominated in domestic currency (Chile), and obtaining longer-term foreign exchange hedges (Poland), indicates that such problems may not be insurmountable. It goes without saying that sound macro policies provide the best means to overcome foreign suspicions.

Dealing with the volatility of capital flows will clearly continue to be of longer-term concern. However, in the near term, the problem may well be a shortage of external capital rather than a dangerous surfeit. While the data are not very reliable, it seems to me the case that over the last four years there has been a net outflow of capital from emerging market countries, with the withdrawal of funds by internationally active banks being the most pronounced. To an important degree, this reflects the recent tendency of internationally active banks to establish a local presence allowing them to both raise funds and lend them in domestic currency. As implied above, this must be thought a good thing. Another good thing is that a number of countries are now running large current-account surpluses and are actively repaying debt. Yet some aspects of this capital outflow are more worrying. Consolidation among large lenders has led to reductions in combined limit exposures. Moreover, in the recent period of market turbulence there seems in industrial countries to have been a more generalised tendency to withdraw from risk taking. This trend stands in sharp contrast to the rising appetite for risk that preceded it. Countries with high domestic savings rates and healthy current-account positions (as in most of Asia) may not be much discomforted by these recent trends. However, other countries not sharing these qualities (in Latin America and eastern Europe) may feel stronger effects. I suspect that a global economic downturn would further exacerbate such withdrawal problems.

A closely related issue is the role likely to be paid by the evolving Basel Capital standards in influencing both the cost of capital to emerging market economies and the volatility of capital flows. Up until now, there is no concrete evidence of the Basel standards playing any important role in this regard. Prima facie it would not be unreasonable to expect that the 20% risk weighting for sub one year interbank loans to banks in non-OECD countries might have induced "excessive inflows" of this sort. However, a number of careful investigations (by Working Groups reporting to two of the standing committees reporting to the G10 governors) failed to substantiate this claim. Moreover, it is instructive to note that in recent years there has been a virtual collapse in this kind of international lending, even though the regulatory framework has not changed to date. This would seem to indicate that, even if the capital standards did play some role, it was by no means a definitive one. This having been said, it is also possible that the influence of the New Basel Accord could be significantly greater. I return to this issue in the next part of my presentation.

4. International capital standards and emerging market economies

(a) International standards and the prevention of financial crises

Before turning to the most recent proposals for revising the Basel capital standards, and their implications for emerging market economies, it would seem appropriate to discuss the issue of international standards more broadly. The drawing up and promulgation of international standards, intended to improve both macroeconomic and financial stability, has in some respects become " the flavour of the month". Indeed, the website of the Financial Stability Forum now lists over sixty sets of such standards and guidelines. Moreover, a growing number of countries, both industrial and emerging, are now actively participating in the Financial Stability Assessment and ROSC programs developed by the IMF. In short, the issue of international financial standards is being treated increasingly seriously by both national and international bodies. Nevertheless, a number of important questions remain.

The first and most obvious question is - why have standards at all? The fundamental reason is related to the discussion in Section 2 above. Market driven, as opposed to repressed, financial systems are inherently more unstable. That is the price to be paid for their greater allocative efficiency. This fact calls for measures to strengthen financial systems so that they will be more resilient to inevitable shocks. This will require that attention be paid to the proper functioning of each of the main elements of the financial system; infrastructure, financial institutions and financial markets. Standards should be seen as a means to this end.

A second question is a little harder to answer. Why should standards be international? One reason for common standards of good behaviour is that they provide a "more or less" level playing field that should stimulate international competition. Another reason is that liberalisation and globalisation in the area of financial services implies some knock-on effects whenever individual countries are prone to financial instability. We all have then a good reason for being concerned about the behaviour of our neighbours. Finally, given a significant degree of uncertainty about what constitutes best practice (or even minimum acceptable standards) there is a simple but compelling logic to assembling an international group of experts (drawn from relevant national jurisdictions) and letting them forge some kind of agreement.

It is also important to recall that international financial standards fall into the genus of "soft law", which in fact has a long history. What is meant by "soft law" is that these agreements have no force in international courts. Rather, they must be implemented by sovereign national authorities that become convinced of their usefulness. As for the history of such agreements, we rightly think of the early work of the Basel Committee on Banking Supervision as being path breaking in the area of financial stability. However, similar agreements designed to facilitate the conduct of cross-border financial transactions (eg technical standards like SWIFT, harmonised contract clauses, UNCITRAL) as well as international agreements to extend cross border competition in the financial area (eg the OECD Codes of Liberalisation and various free trade agreements) have also been in effect for some time.

Assuming that one accepts the legitimacy of this process, there remains a third question. Who should participate in the international groups of experts forging these international standards? While some standards have been set by universal bodies like the IMF, the vast majority have been produced by much smaller and more exclusive groups, many of them associated with the BIS. There is an important tension here that needs to be recognized. On the one hand, any negotiating process must have a limited number of participants if it is to be successful. Moreover, the more expert are the participants in the subject matter, the greater the likelihood of a good final product. Recognising the relative development of financial markets in the major industrial countries, this kind of logic explains the current participation in many standard setting bodies. On the other hand, the special circumstances of financial systems in emerging market countries also need to be recognised, and opportunities given for "graduation" as levels of sophistication increase. Various models are currently being experimented with in order to resolve these tensions which, I repeat, are real.

A fourth set of questions has to do with which standards to implement and in which order. At the national level, no country has the resources to do everything at once. And given that there are 150 sovereign countries, all potentially needing international assistance (of which more below), that too is simply not going to happen. While the Financial Stability Forum has identified 12 Key Standards as having priority, individual countries may well feel that their particular circumstances demand some other subset. Particular circumstances, in association with the recognition that many standards are intimately interconnected, may also mean that individual countries will order their priorities quite differently. Having said this, may I just express my personal view that accounting and auditing standards, corporate governance and the core principles of banking supervision would seem worthy of particular attention in many emerging market economies.

A fifth question is how best to use incentive systems to ensure that chosen standards are actually implemented in emerging market countries where (as also in some industrial countries) there are many impediments to doing the right thing; inertia, vested interests, inadequate laws and often the lack of an independent judiciary. Again, something like the "Three Pillars" approach might be recommended. The most important incentive is self-interest. Those national authorities (public or private) who take responsibility for implementing a given set of the standards must be convinced that it is in the best interests of the country to do so. In India it seems that no more convincing of the authorities is needed, but that is not true in many other countries. A second set of incentives can be given by the official sector in the international community. FSAPs and ROSCs would be a more formal aspect of such a process, while informal peer pressure might also prove useful. Just last month, for example, there was an extended discussion of such issues at the G-20 meeting in Ottawa. The provision of technical assistance could also play a role. So too could "host" supervisors in major financial markets when assessing their willingness to provide access rights to financial institutions from emerging market countries. The third set of incentives should be provided by the markets. They should reward borrowers from healthy financial jurisdictions by agreeing to lower borrowing costs and increase access. Of course, for the markets to be able to do this, they need: to be familiar with the financial standards being recommended; to, believe that they are relevant to risk assessment; and to have ready access to information about observance. While the FSF is working hard on all of the above, it is clear to date that these preconditions for market discipline with respect to standards have not yet been fully put into place.

The last and perhaps the most important question with respect to the implementation of financial standards is where the resources are to come from. In many countries there is a clear shortage of people with the appropriate skills. Moreover, as individuals become more familiar with a technical area, say banking supervision, they are very likely to be bid away by an expanding private sector financial institution. As for international support for such efforts, this too is likely to be in relatively short supply. This argues in favour of a careful attention to priorities in emerging market countries and perhaps rather simpler (not necessarily lower) standards than might be applicable in other circumstances. At the international level, resource shortages may argue in favour of focusing attention on the financial systems of larger countries or those having deeper linkages with outside markets. In both cases, the possibility of financial instability having external implications would be greater. Of course, the difficulty with this suggestion is that it is not easy to reconcile with the status and mandate of the Fund and World Bank as universal and non-discriminatory institutions.

(b) The implications of the New Basel Accord

Implicit in the above discussion is that great efforts are being made to improve risk management procedures at many financial institutions. This is certainly the case in the industrial countries and is increasingly the case in emerging market economies. Perhaps the most widely publicized initiative of this sort has been the New Basel Accord, the most recent version of which was circulated for comment in January 2001. In this section, after briefly summarizing the reasoning behind the new proposals, I would like to address two issues pertinent to emerging market economies. The first question is whether banks in EMEs should strive to adopt the New Basel Accord, and if so in what form? The second question is the likely implication of the New Accord for the behaviour of internationally active banks in industrial countries. This may, in turn, affect capital flows to EMEs.

The old Accord largely achieved its basic objectives. These were to raise the level of capital in the major banking systems and to do it in a way that respected the need for a "level playing field". Moreover, because the old Accord used a small number of fixed risk weights to discriminate between more and less risky credits, it had the major advantage of simplicity. As a result, it was rather quickly adopted as a standard for all banks in over 100 countries, even though it had originally been designed to apply only to internationally active banks in the G10 countries.

With time, however, the deficiencies in the old Accord also became increasingly evident. Given the small number of risk weights, credits of markedly different quality were often lumped together encouraging regulatory arbitrage. In particular, there was an incentive for banks to remove good quality loans from their portfolios, often through securitisation. This process threatened to increase the overall risk exposure of the banking system, leading to concerns about systemic effects given the important role played by banks in the payments and settlement systems in most countries. Moreover, as banks began to engage in new and more complex forms of risk taking, it became clear that the regulatory requirements of the old Accord were deviating more and more from the actual risks being undertaken.

The New Basel Accord tries to address these deficiencies head on. Accordingly, it has to be more complex. It relies for its potential success on the "Three Pillars" of minimum capital requirements, supervisory overview and market discipline. Under the first of these, an attempt has been made to differentiate more clearly between loans of different quality when measuring capital requirements, and to provide incentives for banks themselves to continuously improve their risk management practices. An important corollary is that the evaluated quality of loans can now change over time, which it did not do before. As for supervisory overview, the emphasis is now more on the adequacy of the bank's internal systems rather than detailed regulatory requirements. Finally, market discipline is to be encouraged by greater transparency about bank's operations and their overall attitudes to risk taking.

Another novel aspect of the New Basel Accord is that banks can choose among alternative and increasingly sophisticated regimes in calculating their regulatory capital requirements. As an incentive for banks to put in place increasingly effective risk management systems, these alternative regimes will be "calibrated" such that more sophistication translates into a lower capital requirement. The difference among these regimes essentially rests on who sets the risk weights. Under the old Basel Accord, the regulator set them, but in the future this will be much less the case. However, a continuing and important role for the regulators is that they will have to assess whether particular banks meet certain pre-established criteria allowing them to opt for the adoption of more sophisticated regimes.

Under the New Accord, the first possibility is referred to as the Standardised Approach (SA) The basic idea is that the risk weights will be based on some external assessment of the relative credit worthiness of borrowers; provided by ratings agencies and credit agencies in particular. The second and third possibilities are part of the Internal Ratings Based Approach (IRB). Both assume a more sophisticated capacity on the part of the bank to make a reliable internal assessment of the probability of default on the part of different borrowers. Under the so called Foundation version of the IRB, the regulatory authorities provide the banks with estimation of the loss given default, while under the advanced approach, banks are permitted to estimate these losses themselves.

The publication of these proposals has generated a lively debate as to whether they should apply to emerging market economies, and if so in what form. On the one hand, no one questions that the basic objective of the New Basel Accord - better risk management in a more complex world - is an appropriate one. I would also argue that the ongoing process of market liberalisation and globalisation implies that all banks in all countries will have to allocate capital in ways that are more closely related to actually risks. This will probably mean that banks will eventually have to work within broadly the same regulatory framework. This line of thought leads to the conclusion that emerging market economies should embrace the New Accord in principle and take steps to apply it in practice. On the other hand, a number of commentators have emphasized the limitations of the New Accord and the particular difficulties faced in applying it in emerging market economies. I think it would be useful to look at some of these arguments, not only to see where they are flawed, but also to get some insights as to how to move forward in practice.

As to the limitations of the Accord itself, there are at least four issues that still need to be addressed. The first one is that of the maturity of loans and the associated capital requirement. The current draft of the New Accord proposes that capital requirements increase continuously as the maturity of the loan increases. Many Europeans have questioned this, noting that long-term loans are more likely to be provided only to firms with which the bank has a close relationship. A second issue also has a transatlantic flavour; namely the treatment of loans to small and medium size enterprises. Some people have questioned whether the capital requirements on loans to small businesses are too high particularly given the idiosyncratic nature of the risk. A third issue has to do with the "calibration" referred to above. Simulations indicate that the Accord as currently specified implies a sharp increase in capital requirements for those using the Foundation IRB, a result which could encourage banks to opt for the less sophisticated Standard Approach. And finally, there is some concern that all versions of the New Accord may encourage what seems to be a natural tendency for financial systems to amplify cycles in the real economy. The reason for this is that risk weights (previously fixed) can now change over time, and may themselves be subject to the changing moods of bankers and rating agencies.

All these concerns about the current draft of the New Accord are valid, but do not provide grounds for rejecting use of the Accord in emerging market economies. The first point to note is that these issues are being addressed in a cooperative way which should eventually lead to an improved version finally emerging. It is also important to note that the issues above all relate to the calculation of the minimum capital requirements, which is only one of the three pillars in the New Accord. Should rigorous application of the letter of the law not provide the intended results, the supervisory authorities might be expected to step in. The second point is that there are no compelling grounds for belief that current flaws in the New Accord cause any more significant problems to emerging market economies than they do to some of the larger industrial ones. This having been said, where such differences can be identified, emerging market countries should use the Liaison Committee set up by the Basel Committee as well as all other available channels to identify their primary concerns and suggest improvements. Since India is a member of the Liaison Committee this option is clearly available to it.

Turning now to the applicability of the New Accord to emerging market economies, perhaps the criticism that at first sight seems most damaging is that the capital requirements are simply irrelevant. One strand of thought relies on the empirical observation that capital levels in emerging market countries have no predictive power with respect to subsequent banking crises. This is because accounting standards are often very loose and that, in the event of a capital shortfall, it can easily be made up through some form of connected lending. In such circumstances, measured capital might go up but there would have been no effective dispersion of risk bearing to independent parties outside the banking sector. A second strand of thought points to the numerous problems endemic in the banking systems of many emerging market economies. The conclusion drawn is that scarce resources would be better spent addressing these issues rather than trying to apply capital standards formulated for countries without such endemic problems.

These arguments are not new. The definition of capital has been a subject of on-going debate for years, with the impetus being provided by the introduction of novel hybrid instruments whose true status as equity could easily be debated. The appropriate measurement of non-performing loans (particularly in Japan) and the treatment of provisions against losses continue to generate discussion. As for the resource constraints in emerging market countries, it must be agreed that these are formidable. Nevertheless, simply ignoring the question of capital would also seem imprudent. Good internal governance of financial institutions is the primary requirement for ensuring financial stability, and the best assurance of such good governance is that someone's own money is at stake. Some would argue that capital requirements should be even higher when accounting is more suspect as is the case in some EME's.

A second set of arguments also has some force. The general assertion is that, in some countries, the implementation of the New Capital Accord would weaken risk management, not strengthen it. In many Latin American countries, to quote one example often cited, there are already rigorous and forward looking rules in place having to do with provisions against prospective losses. The New Accord does not signify detailed rules in this area. In other countries, other examples could also be given. As well, some argue that various opt out clauses in the New Accord might prove harmful. For example, loans in domestic currency to domestic governments can be assigned a zero risk weight, even though history teaches us that the default risk is not nonexistent.

By way of counterargument, it could be noted that the thrust of this second set of arguments actually conflicts with the first set of arguments noted above. In general it should be presumed that the standards in emerging markets are generally lower than in the industrial countries, although this might not always be the case. In any event, an emerging market economy with higher standards is in no way being encouraged to roll them back to meet minimum standards elsewhere. The same kind of logic applies to "opt-out" clauses. The fact that a local authority chooses to act in a particular way should not be interpreted as a design flaw in the Accord itself.

Still other criticisms have been made of the specific risk measurement regimes that are proposed under the New Accord. Relative to the current Accord, even the Standard Approach is felt by some to be too complicated and to put too many demands on both bankers and supervisors. Moreover, it fundamentally relies on external assessments of credit risk and these may simply be unavailable in EMEs. Corporate ratings, for example, are almost unknown in many countries, especially for small and medium size enterprises. Moreover, the fact that unrated corporations have a lower risk weight than that assigned to the most risky of the rated corporations has been seen as an impediment to companies seeking such a rating.

Concerns have also been expressed that a rush to seek ratings would in any event overwhelm the ratings agencies. However, and somewhat in contradiction, others argue that there will be such competition for business among the rating agencies that the end result will be an unwelcome "competition in laxity". The counterargument to this last proposition is that the ratings agencies will resist any such tendency, knowing that they have to provide value to investors over time. In any event, regulators could always step in if they felt that such a process was underway. Absent sources for these external ratings of the credit quality of bank loans, a choice of the SA approach could in the limit imply effective maintenance of the status quo. Whether this would be thought desirable or undesirable might itself be a topic for vigorous debate.

As for the Internal Ratings Based Approach, critics would first argue that the sophistication it demands is even further beyond the reach of banks and supervisors in EMEs than in the case of Standardised Approach. Implementation challenges for banks include the development of a new risk culture among managers as well as the compilation of a databank indicating prospective loan losses over a full cycle. Meeting both will clearly take time. Cross-border issues may also complicate implementation if home and host country regulators attempt to impose different capital regimes. As for challenges faced by regulators, this would include shortages of qualified staff as well as a lack of experience in validating bank's internal rating models. Supervisors will also have to be aware of the potential dangers of capital procyclicality under the IRB approach. Even more difficult, solutions must be proposed that try to do something about it under Pillar 2 of the New Accord, without at the same time compromising the incentives for good internal governance of the bank itself.

The list of criteria which must be met before the IRB approach can be sanctioned is in fact very long. In principle, this might imply that very few banks will meet the standards. In turn, this would imply a more common use of the SA approach which may not (discussed above) lead to any major shift from the status quo, for better or for worse. In practice, there may be a risk that some regulators will allow banks to use the internal ratings approach even if they are not fit to do so. This could be because of competition from foreign banks (discussed below) or because national pride precludes admitting inadequacies in risk management capabilities. In such cases, the risks to individual banks and the system as a whole could be significantly increased, particularly if banks relied mechanically on packaged software to do their credit analysis for them. Another concern is that the "calibration" techniques used to establish the overall capital requirements under these regimes are overly based on the experience of large banks in the G10 area. In contrast, loans made in EMEs are inherently more risky and these risks may be less diversifiable. Thus, using G-10 experience may result in capital requirements which are actually too low given the inherent risks involved.

All of the above criticisms imply that the New Accord will not work well enough in EMEs. However, another whole set of arguments seem premised on the notion that they might work all too well. Many seem concerned that relatively risky borrowers will find that credit becomes more expensive or that it even becomes unobtainable. However, given that the purpose of the exercise is precisely that credit risk should be better assessed and priced, it is hard to believe that this outcome should be unwanted. In any event, this outcome is not likely to be different from that foreseen in the industrial countries, particularly in Japan and continental Europe. In all these jurisdictions, such altered behaviour is not only likely to contribute to healthier banks but it should also help reallocate real capital to areas where it can be used more productively.

The discussion of the impact of the New Capital Accord has thus far focussed on the effects on domestic banks in EMEs. A last topic to consider is the effect on internationally active banks and how this might feed back on emerging market economies. The first point to make is that most such banks are likely to wind up being allowed to use the IRB regime. Basing IRB calculations solely on the basis of current ratings, some economists have expressed concern that the capital requirements associated with the current level of loans to EMEs would rise appreciably under the present draft of the New Accord. Moreover, since higher ranked credits would actually have lower capital requirements, the effects on lower rated credits would be even more onerous. Assuming that this extra cost could eventually be reflected in interest rate spreads, the implication drawn by some commentators is that some countries could be materially affected. For countries rated B or C, spreads could rise enough that external financing might simply not be available under normal conditions. Furthermore, were the New Accord to prove more procyclical than the previous one, it is likely that lower rated credits would be even more affected during periods of economic downturn and rising risk aversion. Concerns such as this, arising from the current version of the draft Accord, will of course be taken into account in the on-going efforts to redraft and improve the current proposal.

But even if the costs of external finance were to rise, is it clear that this would be a bad thing? If EMEs in fact have relied excessively on foreign debt, particularly the less creditworthy, shortening the rope with which they can hang themselves might well be desirable. In contrast, a generally more procyclical capital regime would likely be undesirable, even taking into account some of the Schumpeterian arguments invoked above. Finally, it could be suggested that one side effect of the new regime might be to encourage internationally active banks to reduce cross border loans in foreign currency and to rely more on locally sourced funds to support local loans. This would be desirable in that it would help to reduce the dangerous problem of currency mismatches referred to in Section 3 above. Moreover, the domestic presence of well capitalised foreign banks would help to introduce new technology, better working practices and products, and would increase competition. It must, of course, be recognised that large inroads by foreign banks may also have downsides and could well prove politically controversial. As a result, the transition to a new competitive equilibrium in the banking sector will have to be carefully managed.

5. Conclusions

The international financial system has been in a process of rapid change, with technology and liberalisation being among the primary driving forces. It seems doubtful to me that this general trend will be reversed. Nevertheless, the recent financial crises in many emerging markets and the current global economic slowdown may, in the near term, temper the speed at which further such changes are likely to take place. Such considerations should not be used as an excuse to try to roll back the clock in the area of financial sector reform. The efficiency costs of highly repressed financial systems are simply too great.

The problem of large capital flows, both in and out, of emerging market economies is a serious one. Nevertheless, the practical solutions seem to lie largely in the hands of the recipient countries. They must be aware of ongoing developments and be prepared to take both strategic and tactical measures to deal with them. At the strategic level, the problem of currency mismatches can be alleviated by reduced dependence on fixed peg exchange rate regimes as well as by the development of more complete domestic financial markets. This will make reliance on unhedged borrowing in foreign currency less common. At the tactical level, better management of sovereign assets and liabilities in foreign exchange could prove useful. Greater attention to the overall exposure of the national balance sheet would also seem desirable. In this regard, banks and other domestic lenders should make it their business to know the currency risk exposure of those to whom they lend.

Turning to the general issue of international standards in the pursuit of financial stability, a number of issues have been raised. It seems agreed that such standards are useful, and that they must have an international dimension. What is more open to discussion is the process through which such standards are set and the issue of who should participate in the process. Also open to discussion is the question of implementation. This requires the setting of priorities as well as the establishment of incentive systems to ensure that participants in the financial system want to do "the right thing". Finally and most important, implementation requires the provision of adequate resources to ensure that what people want to happen actually materializes on the ground.

The last issue has to do with the more specific question of the New Basel Accord, and how it might have an impact on emerging market countries. The Accord is, of course, still under discussion and will finally emerge in a somewhat different form than it currently takes. This having been said, there can be little question that the philosophy underlining the New Accord applies to emerging market countries as well as industrial ones. That is, the governance mechanism for an increasingly liberalized financial system must rely on self-regulation and market discipline as well as the application of more traditional regulatory rules. All of the above will be required to ensure risk management systems that can adequately price and deal with risk in a more rapidly changing world. However, individual emerging market countries do differ in their stage of financial development likely implying that they may also choose to adopt different variants of the Accord. Moreover, since most are still at a relatively early stage of financial development, compared to the major industrial countries, we might also observe a general tendency to begin with less demanding approaches with a general evolution towards more complicated variants over time. In this policy area, as in many others, it is better to begin with standards which are less perfect but more practical, rather than standards that are more perfect but simply cannot be applied in a practical setting.