The globalisation of financial markets
Professor Otmar Issing, 12 September 2000, Ottobeuren
Over recent decades, there has been a steady increase in cross-border financial flows around the world. First, various financial institutions including banks and institutional investors have expanded their activities geographically. In this process, they acted as an intermediary to channel funds from lenders to borrowers across national borders. Second, the more mature securities markets have gained a clear cross-border orientation. In many instances, newly issued securities are designed and offered to the public in such a way as to maximise their appeal to international investors.
These developments reflected the progressive dismantling of controls on cross-border financial flows as well as the liberalisation of national financial markets more generally.
This resulted in an augmentation of the range of borrowing and lending possibilities available to economic agents throughout the world. In particular, there was a marked expansion of the range of financial possibilities available for financing current account deficits and recycling current account surpluses.
In the process of developing globalisation of financial markets seen over recent decades, both technological advances and financial innovation played a key role. In the past few decades, information systems have become able to compute and store more data more rapidly. Telecommunications networks have extended their ramifications and augmented their capacity while more reliable data exchange protocols have made it possible to connect computing machines in more efficient ways. As a result, cross-border financial deals have become both easier and more secure, effectively lowering the barrier constituted by distance, be it determined by geography or other factors. Moreover, particularly over the last two decades, financial markets have become breeding ground for a wide array of rapidly evolving financial products, often described generically as "derivative" instruments. These products make it possible for borrowers and lenders to customise their risk exposures as well as adjust them over time. With derivative products, borrowers and lenders can therefore mitigate some of the problems associated with asymmetries of information in financial markets, which are particularly acute in the international context.
However, the integration of financial markets across national borders is not a new phenomenon. In the period of the gold standard, from around the middle of the nineteenth century to 1914, financial markets had been well integrated at the global level although the architecture of the international financial system was very different from what it is now. Nevertheless, several of the issues envisaged by economic policy makers in the period preceding 1914 are astonishingly similar to those posed by the globalisation of financial markets in the present day. This includes issues such as the vulnerability of the international financial system to crises or the limited room for manoeuvre which is available for economic policy pursuing specific objectives at the national level.
In this article, I would like to first review the evidence suggesting that financial markets have recently become more globalised. In this context, I will examine various indicators of globalisation and, after reviewing the evidence, try and identify an operational definition of the concept of globalisation, which will be useful in order to discuss the implications of the phenomenon for monetary policy. My second aim will be to discuss the benefits and risks associated with the globalisation of financial markets and the role that monetary policy can play in this context particularly as concerns the euro area.
2. Have financial markets become global?
Global markets are markets in which the law of one price applies, in the sense that it would be possible to buy or sell products for the same price irrespective of geographical location and local circumstances. When products are purchased and sold outside national boundaries, price differentials may remain as long as there are costs specifically associated with cross-border exchange as opposed to exchange within national boundaries. Hence, the process of internationalisation of financial markets is only a step towards global financial markets. This distinction between globalisation and internationalisation seems to apply to financial markets as well as to markets for goods and non- financial services. Over recent decades, financial markets have gained a clear cross-border orientation but, overall, it can be argued that they are still not truly global.
2.2 Review of the evidence: quantities
The international orientation of financial markets can be measured by cross-border financial flows. Bordo, Eichengreen and Kim (1998) argue that a good indicator of cross-border financial flows is the absolute value of the ratio of current account balance over GDP, averaged across a number of countries. They show that this indicator has increased somewhat since the mid-1960s but still remains below the levels seen from the mids-1870s to 1914 (see Figure 1). Furthermore, in the more developed countries the values taken by this indicator were more stable in the period preceding 1914 than over recent years. These observations suggest that net financial flows, although increasing, are currently not as large and, overall, more volatile than in the period preceding World War I. As argued by Flandreau and Rivière (1999), this analysis of developments in current account balances may be refined further when not only the time series dimension, but also the cross-sectional dimension is taken into account in econometric work. The results of Flandreau and Rivière confirm that the degree of financial integration, when measured according to net current account balances, increased over recent years but still remained somewhat below the levels seen before 1914.
Nonetheless, as noted by Bordo, Eichengreen and Kim (1998), there is evidence to suggest that, in contrast to developments in net financial flows, gross financial flows are larger today than in the period before 1914. In particular, although little data is available before 1914, it would seem that turnover in foreign exchange markets was lower then than it is now, consistent with the analysis made by Bloomfield (1963).
In summary, an analysis of the evidence on financial flows would tend to suggest that the degree of internationalisation of financial markets has increased over recent years, to reach levels broadly comparable to those seen in the period of marked financial market integration which preceded World War I. As mentioned at the outset, however, quantities alone cannot be a sufficient indicator of globalisation, because the law of one price may still not apply when cross-border financial flows become more widespread.
2.3 Review of the evidence: prices
Obviously, there are many difficulties associated with the assessment of deviations from the law of one price, as pointed out for example by Obstfeld (1994).
First, as financial and "real" prices are closely intertwined, it is difficult to test the hypothesis of the existence of a globalised financial market in isolation. For example, if there are different patterns of adjustment of relative consumer prices to shocks in different economies, then co-movements in financial prices should preferably be examined in real terms instead of nominal terms, so as to control for the effect of different nominal rigidities. This will be rendered complicated by the fact that inflation expectations, which determine the differential between real variables and nominal variables, are generally not observed ex ante. Hence, tests of the law of one price are obviously based on approximations. In many cases, the test results are such that one cannot make a clear separation between the hypothesis of global financial markets on the one hand and, on the other hand, the hypothesis of global markets for goods and non-financial services.
Second, it is difficult, if at all possible, to identify financial products which are fully comparable in the various national financial markets. In particular, debt securities issued by governments typically provide the basis for measuring "risk-free" long-term interest rates in a given currency. However, a wide array of specific factors beside general economic and financial conditions may affect government bond yields. Such factors include in particular sovereign default risk, the maturity structure of outstanding government debt and its evolution over time, the depth and liquidity of the secondary market for government bonds or the tax regime applicable to capital gains and interest receipts. These difficulties become even more challenging when considering financial instruments issued by the private sector, such as equities.
Conversely, when the evidence points towards more co-integration between financial prices at short-term horizons, this may not reflect more financial market integration but the spreading of relevant information for market participants more widely and more rapidly than before. Take for example the case of the use of the telegraph and later the telephone for the transmission of financial market information in the nineteenth century, which resulted in a more rapid circulation of information across financial markets. In such a case, the degree of co-integration between financial markets may seem to increase following the introduction of the telegraph, because pieces of information which previously affected financial prices at different points in time are subsequently able to exert a nearly simultaneous impact on financial prices.
With these difficulties in mind, it should nevertheless be acknowledged that the degree of co- integration in financial returns around the world seems to currently be rather large, and consistent with relatively high financial market integration. For example, using a model allowing for short-run deviations from long-run determinants, Bordo, Eichengreen and Kim (1998) assess the degree of long- run co-integration between real interest rates in the United States, the United Kingdom, Germany and France. They show that there was a relatively high degree of co-integration both over the period from 1880 to 1914 and over the period from 1983 to 1998, but less co-integration over the intermediary periods (see Figure 2).
Interestingly, there have been occasions when the existence of relatively close co-integration between financial variables has been mistaken for phenomena of excessive and unexplained contagion in financial markets. Forbes and Rigobon (1999) address the issue of the measurement of co-movements between financial prices They argue that, when volatility changes over time, standard methods of estimation for coefficients of correlation can be misleading. In particular, when volatility increases, standard methods of estimation would suggest increasing coefficients of correlation, which is not always the case in reality. Forbes and Rigobon (1999) show that, in several episodes of stock market turbulence, including the 1987 US stock market crash, the 1994 Mexican peso collapse and the 1997 East Asian crises, there was no evidence of changes in coefficients of correlation. Hence, the co-movements in stock prices observed during these episodes of turbulence reflected the high degree of integration, or interdependence, between stock markets, and not, as has sometimes been argued, the spreading of contagion.
2.4 Review of the evidence: qualitative indicators
There are other indicators of globalisation in financial markets, which may be classified into two broad categories.
First, there are indicators based on an assessment of whether the provision of financial services is open to competition, within the local economy and from the outside. When the provision of financial services becomes open to competition, as a result of the heightened competitive pressures faced by financial intermediaries interest rates paid by borrowers may decline while interest rates received by lenders may increase. More generally, for a wide range of financial products bid-ask spreads are likely to decline to lower levels as financial intermediaries respond to competitive pressures. Hence, the convergence of average bid-ask spreads on comparable financial products will be an indirect indicator of the opening of local financial markets to competition, which in many cases will reflect the lowering of the barriers heretofore limiting the integration of the local market within the global market place.
Second, going further into this line of reasoning it is possible to argue that another indicator of financial market integration is the absence of opportunities of arbitrage across borders. By taking advantage of anomalous price differences, risk-free arbitrage would make it possible to realise profits with certainty while incurring very low costs. A typical example of the effect of arbitrage trading is the high degree of smoothness observed in the term structure of government bond yield curves. When government bond markets are very deep and liquid, arbitrageurs can take advantage of even very small pricing differences between bonds with similar maturity dates. Such arbitrage activity, when carried out frequently and on a large scale, contributes to quickly smoothing out any pricing anomaly such as those which occur, for example, when certain less professional market participants place orders at prices slightly below or above market prices. Hence, the degree of smoothness of yield curves will provide indications on the degree of depth and liquidity of the government bond market as well as the degree of arbitrage activity carried out by market participants.
Chen and Knez (1995) develop an indicator which exploits the idea of absence of arbitrage opportunity to derive a necessary and sufficient condition for the law of one price to hold across two markets. Ayuso and Blanco (1999) apply a refined version of the method suggested by Chen and Knez to stock price data for the United States , Germany and Spain. They find results suggesting that, during the 1990s, there was an increase in financial market integration across these countries. Obviously, this approach is subject to various estimation uncertainties reflecting its complexity, as noted by Kan and Zhou (1999).  However, it is noteworthy that, in principle, the results permit to discriminate between actual progress made in financial market integration and other factors affecting the co-integration of financial prices such as increases in the speed of transmission of information.
2.5 Puzzles in international financial economics
The evidence available so far seems to point towards increases in the degree of integration of financial markets over recent years. This has contributed to reducing, but not completely removing, the puzzles constituted by the so-called home bias in portfolio investment and the high correlation between saving and investment within an economy with no or limited barriers to cross-border financial flows.
The home bias puzzle was identified by French and Poterba (1991), who noted the low proportion of foreign stocks in portfolio investment compared with the optimal levels suggested by the trade-off between risk and return. Investors seem to totally or partly ignore the benefits associated with the diversification of the sources of risks, as may be achieved through a wider geographical coverage. As emphasised for example by Obstfeld and Rogoff (2000), a possible explanation for home bias is the existence of various forms of transaction costs for cross-border purchases and sales of financial instruments. However, Tesar, L., and I. Werner (1995) point out that such transaction costs are unlikely to be very high since turnover in foreign stocks is larger than for domestic stocks. In contrast, Hasan and Simaan (2000) focus on information asymmetries. They show that home bias in portfolios of stocks can be reconciled with theory in the presence of uncertainty regarding the estimation of returns and risks on foreign stocks, provided such uncertainty is higher than for domestic stocks. In some countries, a further explanation for home bias is the existence of restrictions on the range of instruments which can be held by certain categories of investors, such as quotas on foreign country exposures of investment funds for example.
Another puzzle is the one identified by Feldstein and Horioka (1980). Feldstein and Horioka remarked that, over the long run, developments in domestic saving and investment rates were more closely related than theory would suggest. In other words, the financing of investment and the placement of savings seem not to take full advantage of international financing and placement opportunities. Amongst the various explanations which have been suggested to elucidate the Feldstein- Horioka puzzle, many are similar to the explanations envisaged for the home bias puzzle. For example, Obstfeld and Rogoff (2000) argue that transaction costs can help explain not only the home bias puzzle, but also the Feldstein-Horioka puzzle as well as other puzzles in international economics such as the low cross-country correlation of growth rates in consumption.
As pointed by several authors, however, there is evidence that, although they still exist, the puzzles have become less pronounced over recent years, as net cross-border financial flows have increased. In particular, international banking activity has increased steadily in recent decades as large banking institutions developed activities at the global level (see Figure 3). In addition, both the amount outstanding of securities issued in foreign currencies and the amount outstanding of securities held by foreign investors increased rapidly over recent years. As a result of these developments, there seems to be a progressive shift in the average composition of portfolios by currency and issuer, towards patterns more in line with those suggested by the theory.
2.6 The integration of financial markets in the euro area
The introduction of the euro has created the second largest financial market in the world. This is on the one hand an outstanding example of further regional integration and, on the other hand, also a contribution to the globalisation of financial markets.
As pointed out by Noyer (2000), in the context of the programme of Economic and Monetary Union, the processes of economic activity became more intertwined in the European Union, which naturally led to the development of more cross-border financial interconnections. Furthermore, financial flows between the euro area and the rest of the world increased rapidly over recent years. When measured as the sum of net direct investment and net portfolio investment, net financial flows between the euro area and the rest of the world increased from EUR 71 billion in 1997 to EUR 168 billion in 1999.
The integration and increased globalisation of financial markets in the euro area are taking place against the background of price stability and sounder public finances.
When inflation is low and expected to remain low and subject to limited variations over the medium term, the prices of financial assets incorporate little inflation risk premia, in contrast to a situation of high or uncertain inflation and where the central bank lacks credibility. Moreover, with the introduction of the euro financial prices are no longer affected by intra-area foreign exchange risk premia. As both the foreign exchange risk premium and the inflation risk premium become relatively less important as a determinant of financial prices within the euro area, other factors such as credit risk can play a more important role in the price formation mechanism. In order to deal with this evolution, financial market participants active in the euro area have stepped up their assessment of the credit quality of the security issuers. This development may be described as the development of a "credit risk culture". Ultimately, this results in a more efficient allocation of financial resources, which is beneficial for economic growth over the longer term.
In the run-up to the Stage Three of EMU the governments of the European Union have made efforts to reduce their deficits and their debt ratios so as to put them in line with the requirements specified by the Maastricht Treaty. As was intended by the drafters of the Treaty, this development contributed to achieving sounder fiscal policies, which are going to be more conducive to economic growth over the longer term and make it possible for automatic fiscal stabilisers to work more effectively when needed. A further effect of the reduction of public deficits and debt, all other things being equal, is the freeing of financial resources for use by the private sector.
In this context, the introduction of the euro has acted as a catalyst to further integration of financial markets in the euro area, although this process is still far from complete particularly in some specific fields.
Already in the run-up to the introduction of the euro, but especially since 1 January 1999 , there has been a marked decline in transaction costs in many areas. In foreign exchange, interest rate and equity markets, bid-ask spreads have generally declined in 1999 compared to 1998, sometimes very significantly. Furthermore, trading activity expanded, particularly in some segments such as the private bond markets. As a result of the increase in trading activity, market participants were able to carry out large transactions within a short period of time more easily and with smaller costs. Hence, in 1999 not only bid-ask spreads but also other indicators pointed towards an increase in the depth and liquidity of financial markets in the euro area. Bayle, Santillán and Thygesen (2000) consider structural developments in the money and bond markets of the euro area in detail. While highlighting how much progress has already been made, they also point towards areas where the pace of integration has been slower. This includes in particular the repo market, where a set of technical factors such as market infrastructure and remaining differences in the tax regimes applicable to repo transactions seem to be mainly responsible for delays in further market integration. Considering retail banking markets, Kleimeier and Sander (2000) argue that integration remains incomplete so far despite the harmonisation of the regulatory framework in the European Union. However, as pointed by Noyer (2000), there are some indications which suggest that integration has also been progressing somewhat in retail banking markets, although slowly.
Despite the relatively slow pace of progress made in some areas, Danthine, Giavazzi and von Thadden (2000) argue that the integration of financial markets in the euro area will, over the longer term, lead to an improvement in the efficiency of the functioning of financial markets in the euro area. This conclusion should be welcomed, although one should not overlook the fact that there are a number of areas in the financial markets of the euro area where integration is still not complete. Danthine, Giavazzi and von Thadden show that, in the process of integration of financial markets at area-wide level, the breadth, depth and liquidity of financial markets in the euro area will improve. Eventually, this will contribute to an increase in economic welfare in the euro area resulting from a better allocation of financial resources.
Furthermore, it can be argued that the improvement in the efficiency of financial markets in the euro area will be beneficial also for foreign borrowers and investors, who are able to access the financial markets of the euro area and take advantage of their breadth, depth and liquidity. The financial markets of the euro area, even if they are not yet fully integrated at area-wide level, are larger and more open than any of the markets denominated in the predecessor currencies of the euro.
Several recent developments are symptomatic of the high degree of openness of the financial markets of the euro area.
Over recent months, amounts outstanding in euro-denominated debt securities increased in all issuing sectors, but the rate of increase was more pronounced in the private sector than in the public sector. Obviously, significant changes in the relative share of the various sectors occur only progressively over time. However, it is noteworthy that issuance of euro-denominated debt securities by non residents, although small relative to the total, increased particularly rapidly over recent months (see Figure 4). As a result, the euro-denominated segment of international bond markets has taken up a much larger role than that heretofore played by the predecessor currencies of the euro. However, in this process investors located in the euro area have remained, as before, the main investors active in the euro-denominated compartment of the international bond market. Detken and Hartmann (2000) argue that the euro has become an important placement currency for international bonds virtually from the beginning, but that the use of the euro as an international investment currency is likely to develop more progressively over time. These developments tend to show that, by virtue of its openness as well as of its breadth, the bond market of the euro area has become an important component of international bond markets and can be expected to develop further in the coming years.
The high degree of openness of euro area financial markets is also apparent in the pattern of structural developments recently seen in stock markets. In 1999 and in the first half of 2000, there was a vast array of initiatives taken by stock exchanges in view of forming alliances or merging activities. These initiatives concerned not only stock exchanges located within the euro area but also exchanges located in other countries, notably the United Kingdom and the United States . The process of structural change currently underway in stock markets reflects not only factors specific to the euro area but also other factors, among which the increased economies of scope and scale which can be achieved by stock exchanges using new information and telecommunication technologies.
2.7 Spreading of the use of financial derivatives
Transactions in so-called "derivative" products have underpinned much of the increase in gross financial flows over recent decades (see Figure 6). With derivative products, financial market participants can exchange promises related to future developments in financial asset prices.
As highlighted by Issing and Bischofberger (1995), an important distinguishing feature of derivative products is that they can be used to separate risk exposures into various sub-components. Consider for example the case of an investor holding a fixed coupon bond, who enters into a swap contract whereby he pays the fixed coupon to a third party in return for an interest rate payment indexed on money market rates. Thanks to the swap, the investor is able to hedge his interest rate exposure while retaining the credit risk exposure related to the possibility of default on the part of the bond's issuer. As a result, the investor has unbundled his interest rate and credit exposures, enabling him to manage the two types of risk separately. It should be noted, however, that the reduction in interest rate exposure comes at a price, which comprises the cost of the swap transaction as well as additional credit risk exposure related to the possibility of default by the third party on its obligations related to the swap.
In principle, derivative instruments can be contrived for any risk exposure at any maturity. A main prerequisite is that there exists a creditworthy counterpart who is willing to agree on contractual terms, including in particular the modalities of observation of the contingencies foreseen in the exchange. In the above example, the future path of the money market rates is the main contingency foreseen in the swap contract. One of the factors that play a key role in the development of markets for derivative instruments is the ability of market participants to assess the creditworthiness of counterparts. A further factor is the ability of market participants to identify a commonly agreed source for the observation of contractually defined contingencies, including in particular some sources of prices for the financial instruments underlying the derivative contract. This will be facilitated by the existence of a deep and liquid market for the underlying financial instruments. Under these conditions, derivative products can contribute to enlarging the set of "promises" available for exchange. As argued by Issing and Bischofberger (1995), derivative products can thus contribute to making financial markets more complete.
Because of the need of reliable price sources for derivative instruments, derivative markets will function well only if the markets for underlying instruments also function well. Conversely, as argued for example by the Bank for International Settlements (1999), the depth and liquidity of the markets for the underlying instruments can benefit from the existence of a well functioning market for derivative products.
In the process of integration of financial markets at the global level, activity on the markets for derivative instruments has played a much more important role over recent decades than it did over the preceding period of close financial integration, before 1914. It can be conjectured that advances in information systems and telecommunications technology played a key role in explaining the increased role of derivative market activity. These advances made it possible not only to exchange more information more rapidly, but also to verify the accuracy of information in a more reliable manner, a feature which is of importance because such information provides the basis against which due payments are calculated according to derivative contracts.
In summary, the evidence available so far seems to indicate that a process of increasing globalisation of financial markets is underway. Although net financial flows are not as large or stable as they were in the period from the mid-1870s to 1914, by the 1990s gross financial flows have become much larger. Although there seem to remain sizeable limitations to the globalisation of financial markets, as indicated by the persistence of puzzles such as home bias or the Feldstein-Horioka puzzle, it would seem that the barriers to internationalisation are progressively lowered within the limits imposed in particular by remaining information asymmetries. In particular, data on both international banking activity and international security holdings seem to point towards a clearer cross-border orientation of financial flows. Although derivative markets may be subject to inefficiencies as well, when they function well they can enhance the completeness of financial markets, thus contributing to more efficiency in the allocation of financial resources.
As I have pointed out on previous occasion, the globalisation of financial markets is better seen as a process rather than as a fixed state of affairs. As mentioned by Alan Greenspan, "there are no signs as yet that the globalisation process is about to stop or, excluding the one-off effect of the introduction of the euro, even slow down perceptibly in the immediate future". Starting from this observation, in the remainder of this article I would like to examine the benefits and risks associated with the globalisation of financial markets. In addition, I will also discuss the role that monetary policy can play in this context, particularly as concerns the euro area.
3. What are the benefits and risks associated with the globalisation of financial markets, and what are the implications for monetary policy?
A short description of the economic benefits associated with the globalisation of financial markets is proposed by Obstfeld (1994), who writes that, "in theory, […] individuals gain the opportunity to smooth consumption by borrowing or diversifying abroad, while world savings are directed to the world's most productive investment opportunities". A practical implication of the theory is that globally integrated financial markets provide more flexible ways of both financing current account deficits and recycling current account surpluses. Moreover, the free play of market mechanisms should tend to ensure that both borrowers and lenders do not knowingly take excessive risks.
Additional benefits from the globalisation of financial markets include the more rapid spreading of technological advances, financial innovation as well as, more generally, financial performance to the various parts of the globe.
In a global financial market, technological advances in payment, settlement and trading systems as well as in financial information systems can be made available to all market participants instantaneously. For example, in the 1990s certain large financial institutions published, at no or limited cost, the new methods of calculation which they had developed for measuring their market risk exposures as well as, later, their credit risk exposures. This contributed to a relatively rapid spreading of the new risk measurement technologies amongst numerous financial institutions which needed to upgrade their risk assessment systems, including in particular financial institutions located in the euro area.
Another source of benefit from the globalisation of financial markets is the spreading of financial innovation. Irrespective of their location of residence, financial market participants may, provided they can demonstrate sufficient creditworthiness, make use of new financial products as soon as these instruments start being marketed in the global financial marketplace.
More importantly, in open financial markets the entry of foreign financial institutions into domestic financial markets can bring sizeable benefits, as increased competition can help to enhance efficiency in the financial sector. Berger, DeYoung, Genay and Udell (2000) provide an analysis of changes in efficiency in the financial sector following the entry of foreign financial institutions. They show that efficiency gains remain limited, which presumably reflects the difficulties associated with the acquisition of local financial information. However, they also show that US financial institutions seem to be able to cause gains in efficiency when they enter local financial markets, which seems to reflect the positive impact of increased competition and of the transfer of corporate expertise.
In the last three decades, while more progress was made in the internationalisation of financial markets, there have been numerous occurrences of financial crises. Some of these financial crises exerted disruptive and protracted effects on the economy, both locally and globally. As a result, economic policy makers together with market participants have devoted considerable attention to the issue of how best to avoid financial crises. In particular, in April 1999, a grouping called the Financial Stability Forum (FSF) was created with the objective of promoting international financial stability through information exchange and international co-operation in financial supervision and surveillance.
In economic terms, the risks to financial stability may be seen as mainly arising from market inefficiencies. As pointed out earlier, the home bias and Feldstein-Horioka puzzles would seem to suggest that inefficiencies remain sizeable in international financial markets, presumably because of information asymmetries as well as transaction costs. In broad terms, these inefficiencies can manifest themselves in various ways, including externality and co-ordination problems. Externalities would seem to occur in particular with the setting up of market infrastructures such as payment and settlement systems. Co-ordination problems would seem to occur mainly in the form of principal- agent problems affecting the relationship between borrowers and lenders because of information asymmetries or information verification problems.
As a remedy against financial crises, some commentators have proposed to reinstate some restrictions on capital flows in certain specific cases. An example often cited in this regard is that of Chile , where short-term capital inflows were penalised between 1991 and 1998 through the imposition of minimum reserve requirements. Certain observers, considering the performance of the Chilean economy and its resilience to the crises of recent years, have remarked that this system of penalisation of short-term capital inflows may have been effective in preventing crises and promoting economic growth in Chile . However, as shown by Edwards (1999), a careful analysis of the evidence does not permit to confirm any positive impact of the penalisation of short-term capital inflows in the case of Chile.
Another suggested remedy is the proposal originally made by James Tobin to tax foreign exchange transactions or, more generally, all short-term financial flows. Proponents of the Tobin tax argue that it would help reduce financial market volatility while costing little to financial market participants, particularly for long-term transactions. A forceful argument which has been made against the proposed Tobin tax is that it would be difficult to implement. Other arguments against the Tobin tax are more closely related to economic theory. The Tobin tax would lead to an increase in transaction costs. Although the increase may be small relative to the transacted amounts, it could nevertheless amplify the market inefficiencies already observed within the current framework, which seem to partly arise from even low transaction costs and are apparent in particular the home bias and Feldstein-Horioka puzzles. Hence, it would seem difficult, in view of the present state of economic research, to provide a clear-cut case in favour of the Tobin tax.
Let me come back to the two above-mentioned manifestations of financial market inefficiencies which may lead to financial crises: externalities and information problems.
Externalities are especially consequential in the development of the infrastructure of financial markets, including in particular payment and settlement systems. In many cases, economic agents are not able to endogenise the benefits associated with well functioning payment and settlement systems, such as the ability to process transactions with a large number of potential counterparties and in conditions of high safety. As a result, economic agents may content themselves with sub-optimal payment and settlement systems unless further incentives are provided to ensure that the existing systems are upgraded. It has now become widely accepted that the most secure systems for processing large value payments are real-time gross settlement systems. For the euro, the payment system operated by the European System of Central Banks, TARGET, falls into the category of real-time gross settlement systems. Over recent years, the setting up of a more secure infrastructure for the settlement of foreign exchange transactions has become a more prevalent issue. Ways were being sought to enhance the foreign exchange settlement infrastructure so as to minimise the risk of payment system gridlock of the type occasioned by the failure of the Bankhaus Herstatt in 1974. Building on the recommendations published by the Committee on Payment and Settlement Systems of the Group of Ten central banks, several private initiatives were launched over recent years to try and set up a new, more secure, foreign exchange settlement system. As from 1997, some of the promoters of these endeavours joined their efforts in the so-called Continuous-Time Linked Settlement Bank (CLS Bank). According to Kahn and Roberds (2000), the system designed by the CLS Bank should permit to achieve a substantial reduction in foreign exchange settlement risk with limited costs and in a sufficiently safe and secure manner.
Over recent months, concerns of a nature similar to those posed by foreign exchange settlement risk have arisen following the development of the electronification of trading systems, which has proceeded at a remarkable pace in particular in the foreign exchange markets. Persaud (2000) notes that, on certain days in 1998 and 1999, there were some unusually sharp moves in foreign exchange rates over very short periods of time. Persaud suggests that this so-called gapping behaviour may be attributed to the adaptation of the market to the new infrastructure. As the electronification of financial markets progresses further, it is obviously important that the providers of electronic trading platforms adopt sufficiently safe procedures, so as to limit the risk of system failures, including the risk of gridlock in trading activity in case of market turbulence.
Turning to the second manifestation of financial market inefficiencies, information problems - or coordination problems - occur when borrowers and lenders are not able to exchange all the information that would be required in order to conduct transactions in an efficient manner. Consider for example the so-called adverse selection problem. In order to avoid possible credit losses, lenders with limited information on the creditworthiness of borrowers will only be able to offer relatively high interest rates on loans. This will tend to attract the less creditworthy borrowers, which may result in credit rationing for the best borrowers. More generally, when the financial soundness of borrowers is low and information asymmetry is high, there may be sizeable inefficiencies in financial markets. In such cases, the release of information revealing the true creditworthiness of borrowers may result in a disorderly withdrawal of funds when lenders realise that they do not have sufficient capital to buffer possible credit losses. The risk of such a scenario materialising is obviously particularly high when the modalities of corporate governance are such that the managers of firms do not abide by minimum standards of honesty and transparency.
These problems can explain why, when poorly managed, the liberalisation of financial markets can augment the risk of occurrence of financial crises. As discussed in a recent Economic Issues publication of the IMF, financial institutions may react to more competition with imprudent lending, or they may be unable to deal with the information problems which are magnified in the transition period. Bernard and Bisignano (2000) show that information problems play an important role in the international inter-bank market. They argue that the problems can be so severe as to make it necessary to introduce a subsidy to market participants, in the form of implicit guarantees of support in case of failure. However, Bernard and Bisignano (2000) show that this appears to have led to moral hazard, resulting in excessive risk taking which may have contributed to amplifying the financial market turbulence of the autumn of 1998.
These information problems can be remedied if financial market participants endeavour to measure risk exposures as adequately as possible and constitute capital buffers which are adequate in order to overcome possible losses.
Obviously, risk exposures should be determined according to several possible degrees of severity of circumstances. For normal circumstances outside periods of market stress, the so-called Value-at-Risk indicator can provide reliable measures of risk exposure. However, prudence requires that financial market participants also evaluate their risks of losses in periods of extreme market stress, so as to be able to gauge the extent to which their capital buffers would be consummated in such circumstances. The recommendations made by the Basle Committee on Banking Supervision specify how market risk exposures should best be calculated. As concerns credit risk exposures, the Basle Committee is in the process of preparing a new set of recommendations dealing with more accurate and varied measurement methods.
In the capital adequacy framework which has progressively been elaborated by the Basle Committee, an important role is played by the requirement that large financial institutions disclose information about their risk exposures and their capital buffers. When such requirement is in force, financial institutions with low capital buffers will find themselves being scrutinised more closely as they will be seen as potentially less creditworthy, which will provide incentives for a prompt upward adjustment of their capital buffers. In addition, the publication of information on the assessment of risks by large financial institutions will positively contribute to the mechanism of formation of a consensus amongst financial market participants about the true state of creditworthiness of borrowers in the economy. Hence, the disclosure of information about risk exposures and capital buffers can help both to enforce discipline amongst financial market participants and to reduce information asymmetry amongst borrowers and lenders.
It would therefore seem to be very important, in order to fend off financial crises, to have both a solid financial infrastructure and sound financial institutions which publish information about their risk exposures and capital buffers. As suggested by Grenville (1999), the existence of such features in financial markets such as those of Australia and Hongkong may help to explain why these economies were able to weather the episodes of financial turbulence in East Asia in 1997 and 1998 well, despite close trade links with the economies in crisis.
A further protection against the effects of financial turbulence, which has been suggested by Alan Greenspan, is the existence of a spare tyre, in the form of strong banks for the case when capital markets seize up, and in the form of deep and liquid capital markets for the case when banks seize up.
3.3 Implications for monetary policy
A situation in which the international financial system is based on the free functioning of market forces can only be considered as beneficial for the economy if financial markets function in an efficient manner.
Monetary policy can provide two main contributions to financial market efficiency, thereby also contributing to economic growth over the medium term.
First, by firmly maintaining price stability, monetary policy will provide a strong anchor for the formation of expectations regarding future developments in consumer prices. When inflation is low and expected to remain low and subject to limited variations over the medium term, the prices of financial assets need not incorporate as high inflation risk premia as in a situation of high or uncertain inflation. As the inflation risk premium becomes relatively less important as a determinant of financial prices, other factors such as credit risk can take on a larger role in the price formation mechanism. Ultimately, this results in a more efficient allocation of financial resources. (Obviously, there are other beneficial effects of price stability. In particular, by reducing instability in relative prices, price stability contributes to a better functioning of the price formation mechanism, thus enhancing the quality of price signals.)
Second, monetary policy can also contribute to market efficiency by reducing, to the extent possible, the occurrence of surprises related to monetary policy decisions. In such a way, monetary policy would avoid creating any unnecessary volatility in long-term interest rates, which would contribute to the quality of price formation on interest rate markets. The occurrence of surprises can be minimised by providing financial market participants as well as the public with a clear, detailed and honest description of the monetary policy strategy followed by the central bank. In addition, it is useful to publish, on a regular basis, the analysis of economic developments made by the central bank so as to ensure that the public is able to ascertain in detail what the views of the central bank are about the risks to price stability.
According to the statute of the ECB, the primary objective of the ECB is to maintain price stability. The statute of the ECB thereby provides not only a clear mandate for monetary policy, but also an anchor for market expectations.
In the context of the globalisation of financial markets, there are two main challenges to be faced by monetary policy.
A first challenge is that, in its communication with the public, monetary policy should take into account the rapidity with which information is transmitted across the world. The increased speed of transmittal of information means that the reactions to monetary policy decisions can be rapid and widespread. Decisions therefore have to be explained immediately after they are taken, in a concise manner. This is one of the reasons why the President and the Vice President of the ECB present and explain monetary policy decisions in a press conference which follows every second meeting of the Governing Council of the ECB. On these occasions, the President and the Vice President also lend themselves to the questions of the journalists. Obviously, the monthly press conference does not reduce the need for a more in-depth analysis of current developments, which is published in the Monthly Bulletin of the ECB.
The second challenge is that monetary policy should be able to take into account structural changes to the economic environment which occur as a result of the globalisation of financial markets. As highlighted by Okina, Shirakawa and Shiratsuka (1999), financial integration must be considered together with economic integration resulting from increased cross-border trade in goods and services, although the two forms of integration may not always progress at the same pace. In any case, economic interdependence will increase as the international orientation of the economy augments. In particular, the relevance of various economic variables for assessing risks to price stability will change with the development of more globalised markets. Moreover, the pattern of transmission of the monetary policy stance to the economy may also change and, in the process, gain in complexity.
The monetary policy strategy of the ECB is well equipped to deal with such changes in the structural relationships which are at work in the economy. One of the reasons for this is that the monetary policy strategy of the ECB was designed in such a way as to be able to cope with the change of regime which would be occasioned by the changeover to the euro. The first pillar of the monetary policy strategy will continue to provide a framework of reference to assess developments in monetary aggregates, which have a stable and predictable relationship with inflation over the long run. Within the second pillar of the strategy, a wide range of indicators is monitored carefully so as to carry out a broadly- based assessment of the outlook for future price developments. Clearly, some indicators gain heightened relevance when financial markets are more global. For example, the analysis of financial flows can come to play a more important role in the overall assessment.
4. Concluding remarks
Globalisation is a process which is not yet finalised and will lead to a world in which countries and economic areas become more and more interdependent. In some quarters, the unfolding of the process of globalisation has given rise to the view that central banks are becoming increasingly "powerless". As I have shown, the challenges posed to monetary policy by the increasing globalisation of financial markets are of two orders. First, the reactions to monetary policy decisions can be rapid and widespread. Second, monetary policy needs to take into account structural changes to the economic environment, in the form of more interdependence amongst economies. These challenges require clear and honest communication on the part of central bank and a robust as well as encompassing monetary policy strategy which is able to cope with structural economic change. In the context of globalised financial markets, it is of the utmost importance that monetary policy pursues as its primary objective the maintenance of price stability. This is the best contribution that monetary policy can make to economic growth over the longer run and will contribute to efficient and stable financial markets.
The interdependence of economies also has consequences in other fields of economic policy apart from monetary policy. Economic policy objectives in general should be pursued with a stability orientation, so as to limit the risk of creating excessive uncertainty particularly in financial markets. In particular, fiscal and tax policy must be conducted prudently and with a forward-looking orientation, so as to avoid the sanction of abrupt adjustments which market forces can impose on unsound or erratic policies.
In financial markets, the process of globalisation opens access to borrowing, lending and investing world-wide. In addition, it plays a central role in disciplining policies. When monetary and economic policy pursue stability-oriented objectives, our economies will be able to reap the full benefits of the globalisation of financial markets while containing the associated risks.
List of figures
Figure 1 Average of the absolute values of current account balances in major economies
Figure 2 Test of the hypothesis of co-integration between short-term real interest rates in
France , Germany , the United Kingdom and the United States
Figure 3 International bank lending, by maturity, sector and location of borrower
Figure 4 Euro-denominated debt securities by issuing sector
Figure 5 Activity on the euro-denominated interest rate swap market
Figure 6 Activity on over-the-counter and exchange traded derivatives markets
Annex: [pdf, 228kb]
 In this respect, the present communication can be seen as an extension, into the domain defined by the concept of "globalisation", of the task of developing a "vocabulary" to describe and interpret monetary policy, as suggested by B. Winkler (2000) in "Which kind of transparency? On the need for clarity in monetary policy-making", ECB working paper no. 26, August.
 Section 2 of Bordo, M., B. Eichengreen and J. Kim (1998), "Was there really an earlier period of international financial integration comparable to today?", NBER working paper no. 6738, September.
 Flandreau, M., and C. Rivière (1999), "La grande 'retransformation'? Contrôles de capitaux et intégration financière internationale, 1880-1996", Economie internationale, no. 78, p. 11-58
 An alternative way to measure cross-border financial flows is to consider the sum of net direct investment and net portfolio investment. This measure is analysed in Section 2.6 below.
 Bloomfield, A. (1963), "Short-term capital movements under the pre-1914 gold standard ” , Princeton Studies in International Finance, no. 11.
 Obstfeld, M. (1994), "International capital mobility in the 1990s", CEPR discussion paper no. 902.
 Bordo, Eichengreen and Kim (1998), op. cit. In order to approximate "ex ante" real interest rates, the authors use an estimate of expected inflation based on an auto-regressive model of inflation. The results are claimed to be robust to alternative methods for estimating ex ante real interest rates.
 Forbes, K., and R. Rigobon (1999), “ No contagion, only interdependence: measuring stock market co- movements", NBER working paper no. 7267, July
 The measure proposed by Chen and Knez compares the distance between the sets of "stochastic discount factors" implicit in the time path of financial prices with known future payoffs. Stochastic discount factors can be seen as a generalisation of the concept of discount factor, which permits to estimate the present value of payments occurring with certainty at future dates. Chen and Knez show that, broadly speaking, when the distance between implicit stochastic discount factors is equal to zero, there are no opportunities of arbitrage between the two markets. See Chen, Z., and P. Knez (1995), "Measurement of market integration and arbitrage", Review of financial studies, vol. 8, no. 2, p.287-325.
 Ayuso, J., and R. Blanco (1999) "Has financial market integration increased during the nineties?", Banco de España, documento de trabajo no. 9923
 Stressing the limited power of statistical inference based on stochastic discount factors, R. Kan and G. Zhou (1999) caution against the imprudent use of such methods for portfolio choice and performance evaluation, while recognising its usefulness for addressing certain questions not answered by traditional asset pricing theory. See Kan , R., and G. Zhou (1999), "A critique of the stochastic discount factor methodology ” , The Journal of Finance, vol. 54, no. 4, p.1221-1248, August.
 French, K., and J. Poterba (1991), "Investor diversification and international equity markets", American Economic Review, no. 81, p.222-226, May. A survey of the literature on home biases in equity and consumption may be found in Lewis, K. (1999), "Trying to explain home bias in equities and consumption", Journal of Economic Literature, vol. 37, no. 2, p.571-608, June. An early description of home bias problems can be found in Solnik, B. (1974), "Why not Diversify Internationally rather than Domestically?", Financial Analyst Journal, July.
 Obstfeld, M., and K. Rogoff (2000), "The six major puzzles in international macroeconomics: is there a common cause ?", NBER working paper no. 7777, July.
 Tesar, L., and I. Werner (1995), "Home bias and high turnover", Journal of International Money and Finance, no. 14, p.467-492.
 Hasan, I. , and Y. Simaan (2000), "A rational explanation for home country bias", Journal of International Money and Finance, no. 19, p.331-361.
 Feldstein, M., and C. Horioka (1980), "Domestic savings and international capital flows", Economic Journal, no. 90, p. 314-329, June.
 Obstfeld, M., and K. Rogoff (2000), op. cit. In a discussion of this paper, C. Engel shows that the explanation suggested by Obstfeld and Rogoff (2000) may have counterfactual implications, including a possibly too close link between developments in real interest rates and current account balances. See Engel, C. (2000), "Comments on Obstfeld and Rogoff's 'The six major puzzles in international macroeconomics: is there a common cause?' ", NBER working paper no. 7818, July.
 See for example Flandreau and Rivière (1999), op. cit.
 Noyer, C. (2000), "The development of financial markets in the euro area", speech delivered at the Royal Institute of International Affairs, June.
 Data source: ECB (Section 8 of the "Euro area statistics" section of the Monthly Bulletin of the ECB).
 Duisenberg, W.F. (1998), "Remarks delivered at a conference on 'Monnaie unique etfiscalité de l'épargne: quelle Europe financière?' ", September.
 Bayle, M., J. Santillán and C. Thygesen (2000), "The impact of the euro on money and bond markets", ECB occasional paper no. 1, July.
 Kleimeier, S., and H. Sander (2000) "Regionalisation versus globalisation in European financial market integration: evidence from co-integration analyses", Journal of Banking and Finance, no. 24, p.1005-1043.
 Noyer, C. (2000), op. cit.
 Danthine, J.-P., F. Giavazzi and E.-L. von Thadden (2000), "European financial markets after EMU: a first assessment", CEPR discussion paper no. 2413, March.
 This development contributed to an increase in activity on the euro-denominated interest rate swap market (see Figure 5).
 According to Detken and Hartmann (2000), an international placement currency, or finance currency, is a currency widely used to issue debt securities, although the proceeds of the issuance may be converted into another currency by the borrower while investor demand may mainly arise domestically. An investment currency is a currency widely used by international investor to compose their portfolios. Hence, while placement currency use refers to the demand for credit, investment currency use refers to the supply of credit. See Detken, C., and P. Hartmann (2000), "The euro and international capital markets", ECB working paper no. 19, April.
 Issing, O., and K. Bischofberger (1995), "Derivate und Geldpolitik".
 Issing, O., and K. Bischofberger (1995), op. cit.
 Bank for International Settlements (1999), "Market liquidity: research findings and selected policy implications", report of a study group established by the Committee on the Global Financial System, May.
 Issing, O. (1999) "The ECB's monetary policy in the context of globalisation", speech delivered at the Weltachsen 2000 Conference, November.
 Greenspan, A. (2000) "Global challenges", remarks delivered at the Financial Crisis Conference, July.
 Obstfeld, M. (1994) op. cit., p.1.
 Berger, A., R. DeYoung, H. Genay and G. Udell (2000), "Globalization of financial institutions: evidence from cross-border banking performance", forthcoming in Brookings-Wharton Papers on Financial Services, vol. 3.
 The Financial Stability Forum was created in accordance with the conclusions of a report prepared by the then President of the Deutsche Bundesbank, Hans Tietmeyer, on the request of the Finance Ministers and Central Bank Governors of the G7 countries. The internet address of the Financial Stability Forum is http://www.fsforum.org .
 Edwards, S. (1999), "Crisis prevention: lessons from Mexico to East Asia ", NBER working paper no. 7233, July.
 For example: Tobin, J. (1996), "Prologue" in The Tobin Tax: Coping with Financial Volatility, Oxford University Press, p.x-xviii.
 A recent analysis of these issues can be found in De Grauwe, P. (2000), "Controls on capital flows", Journal of Policy Modeling, vol. 22, no. 3, May, p.391-405.
 Kahn, C., and W. Roberds (2000), "The CLS Bank: a solution to the risks of international payment settlement?", forthcoming in the Carnegie-Rochester Conference Series on Public Policy.
 Persaud, A. (2000) "The liquidity puzzle", Risk, p.64-66, June. For foreign exchange markets, an indicator of the global volatility level was proposed in Brousseau, V., and F. Scacciavillani (1999), "A global hazard index for the world foreign exchange markets", ECB working paper no. 1, May.
 Eichengreen, B., M. Mussa, G. Dell'Ariccia, E. Detragiache, G. M. Milesi-Ferretti and A. Tweedie (1999), "Liberalizing capital movements: some analytical issues", IMF, Economic Issues no. 17, February.
 Bernard, H., and J. Bisignano (2000), "Information, liquidity and risk in the international inter-bank market: implicit guarantees and private credit market failure", BIS working paper no. 86, March.
 The Value-at-Risk indicator can be contrived as a measure of bundled risks, covering the various classes of risks including foreign exchange, interest rate, equity and credit risk. Some authors have suggested to use Value-at-Risk as an indicator of aggregate risk exposures. For example, Hada and Sahel (1997) use Value-at- Risk to analyse the costs and benefits associated with a shift in invoicing currency when aggregate foreign exchange risk becomes too large. Blejer, M., and L. Schumacher (1998) use Value-at-Risk to try and estimate the ability of central banks to maintain exchange rate pegs. See Hada, F., and B. Sahel (1997), "Le yen, monnaie de facturation?", Revue d'économie financière, no. 44, December; Blejer, M., and L. Schumacher (1998), "Central bank vulnerability and the credibility of commitments: a Value-at-Risk approach to currency crises", IMF working paper no. WP/98/65, May.
 Grenville, S. (1999), "Financial crises and globalisation from an Australian perspective", remarks delivered at the Reinventing the Bretton Woods Committee Conference, July.
 Greenspan, A. (2000), op. cit. Some commentators have observed that many financial crises follow a common pattern, which involves contagion effects leading to a panic. In an illuminating analysis of how panics unfold, Dupuy (1992) argues that financial markets can make it possible to achieve efficient outcomes in the vast majority of cases, although the pattern of behaviour of economic agents is permanently characterised by a strong tendency to herding in the presence of uncertainty. See Dupuy, J.-P. (1992), Le sacrifice et l'envie: le libéralisme aux prises avec la justice sociale, Calmann-Lévy, especially p.318-329.
 For a more detailed analysis, see Issing, O. (1996), "Geldpolitik in einer Welt globalisierter Finanzmärkte", Aussenwirtschaft, 51. Jg., H. 3, p.295-309
 Okina, K., M. Shirakawa and S. Shiratsuka (1999), "Financial market globalization: present and future", Bank of Japan , Monetary and Economic Studies, December.
 See "Monetary policy transmission in the euro area", article published in the July 2000 edition of the Monthly Bulletin of the ECB.
 Issing, O.(1999), op.cit.