This section describes the remarkable growth of international capital flows and places it in historical perspective.
Changing Volume and Composition of Capital Flows
No reader will be oblivious to the explosive growth of capital flows, especially private capital flows to developing countries, during the 1990s. Net flows to developing countries have tripled from roughly $50 billion a year in 1987–89 to more than $150 billion in each of the three most recent calendar years.7 Private flows now dominate official flows: where private flows accounted for only 3 percent of developing countries’ domestic investment as recently as 1990, as of 1996 they accounted for fully 20 percent (World Bank, 1997, p. 9). Just 10 countries receive about 80 percent of net private flows to developing countries, and the top 14 account for 95 percent, but increasing openness to international financial transactions is nonetheless general: few developing countries are unaffected.8
To understand the forces creating increasing international capital flows, the reader should recall that the phenomenon is not limited to developing countries. Indeed, the cross-border transactions in bonds and securities of the industrial countries are growing even faster. Those of industrial economies have grown, from some $50 billion in 1979–82 to nearly $550 billion in 1993–96. If much of the discussion focuses on developing countries, this is because it is there that the problems that can be created by volatile international capital flows have been most dramatic.
Also important is the extent to which the channels for private capital flows have changed. For many years commercial bank lending accounted for nearly two of every three dollars of private capital flowing to developing countries. It has now been overtaken not only by foreign direct investment but also by portfolio capital flows (Table 1). Today, bonds and equities account for more than a third of total private capital flows to developing countries, whereas bank lending accounts for less than a third (see Figures 1 and 2). These developments reflect technological change, which has reduced the cost of issuing and trading securitized financial instruments; privatization, which has created a population of profit-oriented companies in which it is attractive to invest; far-reaching deregulation of financial markets in key industrial countries in the 1980s and early 1990s, which played an important role in allowing developing countries to raise capital in the form of bonds and equity; the growth of institutional investors like pension funds and mutual funds with an appetite for foreign securities; and macroeconomic and trade reform in developing countries, which has rendered emerging markets more attractive to investors seeking to diversify internationally.
Portfolio and Direct Investment Flows
(Billions of U.S. dollars; annual averages)
Sources: Bisignano (1994); and IMF, Balance of Payments Statistics, Part 2.Portfolio and Direct Investment Flows
(Billions of U.S. dollars; annual averages)
| Gross Outflows | Gross Inflows | Net Inflows | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1973–78 | 1979–82 | 1983–88 | 1989–92 | 1993–96 | 1973–78 | 1979–82 | 1983–88 | 1989–92 | 1993–96 | 1973–78 | 1979–82 | 1983–88 | 1989–92 | 1993–96 | |
| Industrial countries | |||||||||||||||
| Direct investment | 28.6 | 46.9 | 88.2 | 201.3 | 259.6 | 17.9 | 36.6 | 69.3 | 141.9 | 173.0 | -10.7 | -10.3 | -18.9 | -59.4 | -86.6 |
| Portfolio investment | 11.8 | 35.0 | 126.5 | 274.6 | 436.4 | 24.4 | 51.0 | 139.1 | 343.0 | 549.9 | 12.6 | 16.0 | 12.6 | 68.4 | 113.5 |
| Developing countries | |||||||||||||||
| Direct investment | 0.4 | 1.1 | 2.3 | 10.4 | 19.2 | 5.0 | 14.6 | 15.5 | 37.8 | 106.4 | 4.6 | 13.5 | 13.2 | 27.3 | 87.2 |
| Portfolio investment | 5.5 | 17.8 | -5.1 | 10.3 | 19.2 | 1.3 | 3.1 | 4.0 | 27.5 | 95.9 | -4.2 | -14.7 | 9.1 | 17.2 | 76.7 |
Portfolio and Direct Investment Flows
(Billions of U.S. dollars; annual averages)
| Gross Outflows | Gross Inflows | Net Inflows | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1973–78 | 1979–82 | 1983–88 | 1989–92 | 1993–96 | 1973–78 | 1979–82 | 1983–88 | 1989–92 | 1993–96 | 1973–78 | 1979–82 | 1983–88 | 1989–92 | 1993–96 | |
| Industrial countries | |||||||||||||||
| Direct investment | 28.6 | 46.9 | 88.2 | 201.3 | 259.6 | 17.9 | 36.6 | 69.3 | 141.9 | 173.0 | -10.7 | -10.3 | -18.9 | -59.4 | -86.6 |
| Portfolio investment | 11.8 | 35.0 | 126.5 | 274.6 | 436.4 | 24.4 | 51.0 | 139.1 | 343.0 | 549.9 | 12.6 | 16.0 | 12.6 | 68.4 | 113.5 |
| Developing countries | |||||||||||||||
| Direct investment | 0.4 | 1.1 | 2.3 | 10.4 | 19.2 | 5.0 | 14.6 | 15.5 | 37.8 | 106.4 | 4.6 | 13.5 | 13.2 | 27.3 | 87.2 |
| Portfolio investment | 5.5 | 17.8 | -5.1 | 10.3 | 19.2 | 1.3 | 3.1 | 4.0 | 27.5 | 95.9 | -4.2 | -14.7 | 9.1 | 17.2 | 76.7 |
Net Capital Flows to Developing Countries
(Billions of U.S. dollars; period averages)
Source: IMF, World Economic Outlook.1Short- and long-term trade credits; loans (including use of IMF credit); currency and deposits; and other accounts receivable and payable.Net Capital Flows to Developing Countries
(Billions of U.S. dollars; period averages)
Source: IMF, World Economic Outlook.1Short- and long-term trade credits; loans (including use of IMF credit); currency and deposits; and other accounts receivable and payable.Net Capital Flows to Developing Countries
(Billions of U.S. dollars; period averages)
Source: IMF, World Economic Outlook.1Short- and long-term trade credits; loans (including use of IMF credit); currency and deposits; and other accounts receivable and payable.
Selected Developing Countries: Structure of Capital Account
(Average for 1993–97)
Source: IMF, World Economic Outlook.Note: Countries are ordered according to the share of net portfolio investment. Other investment is a residual category that includes all financial transactions not covered under direct investment, portfolio investment, or reserve assets (e.g., trade credits, loans, currency and deposits, use of IMF credit, etc).Selected Developing Countries: Structure of Capital Account
(Average for 1993–97)
Source: IMF, World Economic Outlook.Note: Countries are ordered according to the share of net portfolio investment. Other investment is a residual category that includes all financial transactions not covered under direct investment, portfolio investment, or reserve assets (e.g., trade credits, loans, currency and deposits, use of IMF credit, etc).Selected Developing Countries: Structure of Capital Account
(Average for 1993–97)
Source: IMF, World Economic Outlook.Note: Countries are ordered according to the share of net portfolio investment. Other investment is a residual category that includes all financial transactions not covered under direct investment, portfolio investment, or reserve assets (e.g., trade credits, loans, currency and deposits, use of IMF credit, etc).Historical Precedents and Comparisons
It is commonly asserted that this situation, in which financial markets are globally integrated, debts are securitized, and holdings are widely disbursed, is unprecedented. With equal frequency it is rebutted that there was an earlier period of equally extensive financial globalization, namely, the 40 or so years before 1914.9 In some respects, as detailed in Appendix I, it is true that the financial integration of the pre-1914 era remains unsurpassed.10 Interest differentials on government bonds were slim even by current standards, and the volume of net capital flows, scaled by GDP, remains unmatched even today.
In other respects, however, present-day financial markets are more closely integrated than those of the past. An important difference is that the range of financial instruments that are traded internationally is broader today. In the nineteenth century, international financial transactions were dominated by claims on governments, railroads, and mining companies; today, in contrast, a broader range of securities are traded internationally, including claims on manufacturing and service-sector companies. In addition, whereas bonds (and debentures) were more important than equity claims before 1914, the two types of instruments contribute in roughly equal proportion to international investment today. Finally, foreign direct investment, which is now undertaken primarily by multinational firms in a wide variety of manufacturing and service sectors, was undertaken then by freestanding companies that specialized in resource-extracting activities.11
These differences underscore the importance of information asymmetries and changes over time in the mechanisms for overcoming them. When information is highly asymmetric, investors will concentrate on projects and companies whose assets are tangible and whose operations are most transparent. Hence the disproportionate concentration of nineteenth-century investors first on government securities and then on railways, canals, ports, mining companies, and public utilities.
Similarly, when asymmetric information aggravates problems of agency and corporate control, investors will prefer debt instruments, the returns on which are less closely tied to management actions than is the return on equity. By supplementing shareholders’ stakes with nonvoting debt, relatively concentrated equity stakes diminish the collective action problem for equity holders seeking to monitor management actions and to overcome the principal-agent problems (i.e., problems of incentive compatibility between owners and managers) created by asymmetric information. Hence the preference of pre-1914 investors for debt rather than equity.12
Finally, the foreign direct investors’ reliance on freestanding companies rather than multinationals before 1914 can also be understood in terms of the risks created by asymmetric information. As Wilkins (1998) puts it, freestanding companies “were structured to solve the problem posed earlier: business abroad was risky; it was hard to obtain adequate and reliable information about firms in distant lands; returns were unpredictable; but there were clearly opportunities abroad; a company organized in the source-of-capital country, with a responsible board of directors, under source-of-capital law, to mobilize capital (and other assets) and to conduct the business in foreign countries could take advantage of the opportunities, while reducing transaction costs by providing a familiar conduit” (p. 13, emphasis added).
Growth of International Financial Flows in Perspective
The greater breadth of international financial integration today can be understood, at least in part, in terms of the diminution of information asymmetries. Here recent changes in information technology have played an important role. Computer links enable investors to access information on asset prices at minimal cost on a real-time basis, while increased computing power enables them to calculate correlations among asset prices and between asset prices and other economic variables. The decline in telephone tariffs, the advent of satellite television and the Internet, and other improvements in communications technologies enable them to follow developments in far-distant countries and companies much more efficiently than before.13 Important information asymmetries remain, but it is far easier for, say, a U.S. resident to obtain information about the balance sheet, business strategy, and regulatory environment in which, say, an Argentine utility company operates than it was several years ago. This environment encourages institutional and individual investors to hold an increasingly broad range of foreign securities rather than limit their purchases to that relatively narrow subset of securities issued by governments and corporations with the most tangible assets and transparent operations.
At the same time, new technologies make it increasingly difficult for governments to control international capital flows.14 The fact that computerized trading makes it easy to create synthetic positions in derivative markets, for example, complicates the task of monitoring or attempting to control or tax foreign exchange transactions. Technologies facilitating the transfer of funds between financial centers (and to offshore corporate and financial subsidiaries) and the growth of nondeliverable currency forward contracts make it increasingly difficult for individual countries to effectively enforce capital controls (see Box 2).
The liberalization of capital markets, domestically and internationally, is an ongoing process, with profound implications for the kinds of policies that governments will find it feasible and desirable to follow.
Role of Policy in Facilitating Portfolio Capital Flows
Technological progress, by reducing transactions costs, may have encouraged international financial flows, but the explosive growth of cross-border borrowing and lending would not have been possible without the simultaneous liberalization of national policies on international capital flows. To be sure, the relaxation of statutory controls on portfolio capital flows is itself a response, at least in part, to these same technological forces, which require controls to become more invasive and, hence, more distortionary if they are to retain their effectiveness. At the same time, other factors have contributed to the relaxation of controls, notably macroeconomic stabilization and the trend in many parts of the world toward the deregulation of economic activity and greater reliance on market-based resource allocation.
The relaxation of statutory restrictions on international capital flows has been a gradual but persistent process in the industrial countries. (The figure in Box 3 shows changes over time in the prevalence of various restrictions on international capital flows.15) As Appendix II details, international financial transactions were tightly controlled immediately after World War II. Significant liberalization then occurred from 1950, with the creation of the European Payments Union, through the middle of the 1960s. In the second half of the 1960s, progress slowed as policymakers were forced to grapple with growing payments imbalances among the industrial countries and pressures on their Bretton Woods exchange rate parities. From the late 1970s, liberalization accelerated again. European countries then took major steps toward opening their capital accounts following the creation of the European Monetary System at the end of the 1970s and the adoption of the Single European Act in the second half of the 1980s. (The first of the two panels in Box 3 shows sharp reductions in the prevalence of restrictions on capital account transactions around the time of these two events.) Restrictions on capital account transactions in the industrial countries then fell to very low levels in the first half of the 1990s. Thus, while the pace of liberalization has varied over time, the direction has been uniform—toward progressively greater capital account convertibility.16
Developing countries have also adopted policies conducive to increased international capital flows, although progress has been less uniform and less unidirectional. (Again, see Appendix II.) In some regions, progress toward capital account liberalization was reversed in the early 1980s at the time of the debt crisis. In particular, there was an increase in the prevalence of restrictions on capital account transactions in Latin America, where external debt problems were most severe.17 However, the reimposition of controls was not sustained. Restrictions on capital account transactions began to decline again at the end of the 1980s, once the highly indebted Latin American countries put the worst aspects of the debt crisis behind them, with the help of the Brady Plan and through significant shifts in the stance of their macroeconomic and structural policies, and once markets in the industrial countries evinced a renewed willingness to undertake lending to developing countries. Thus, even countries that have shown a willingness to reimpose controls in times of crisis have sought to return to the path of capital account liberalization and restore their international capital market access when conditions permit.18
Financial liberalization frequently has been accompanied by efforts to strengthen prudential regulation. Quirk and Evans (1995) document that countries that have liberalized their capital accounts generally have made concurrent efforts to strengthen prudential supervision and regulation, especially in the area of the foreign exchange risk assumed by financial institutions. In most of the countries they considered, the bulk of the reforms to improve prudential standards took place before or together with capital account liberalization, although some countries (Costa Rica, Guyana, Indonesia, Jamaica, and Peru, for example), strengthened prudential measures only after adopting capital account convertibility. In countries where banking sector problems worsened after rapid liberalization of the capital account (Argentina, Costa Rica, Latvia, and Venezuela, for example), difficulties often reflected the tendency for liberalization to magnify preexisting weaknesses in banks’ balance sheets. In particular, liberalization in a context where banks were already weak encouraged gambling for redemption and facilitated asset stripping.19
Three Observations on Policy
First, capital account liberalization has gone hand in hand with domestic economic and financial liberalization. Periods of external financial liberalization have also generally been periods of domestic financial deregulation. The removal of certain controls on capital account transactions by the leading European countries in the 1950s was part of the general trend toward progressively greater domestic financial liberalization as countries emerged from the highly regulated and controlled wartime environment. External financial liberalization in East Asia and Latin America in the 1970s was part of a broader move away from economic regulation, import substitution, and governmental guidance of economic development. Capital account liberalization in Europe in the 1980s was a corollary of broader efforts to reduce regulatory burdens and to create a single European market free of statutory barriers to cross-border financial flows. Similarly, the resumption of capital account liberalization in Latin America at the end of the 1980s and its acceleration in other parts of the developing world were part of a broader trend toward economic and financial deregulation and heightened reliance on market forces.20
Nondeliverable Forwards
An example of how financial innovation can increase the incentives to evade capital controls, and the ease of doing so, involves nondeliverable forward contracts. The Mexican peso crisis of 1994 and subsequent volatility in emerging market currencies created a demand for financial products offering investors the ability to hedge against exchange rate risk in emerging markets. One response has been the introduction of a variety of future contracts. In particular, nondeliverable forward contracts have assumed a significant role. These are offshore contracts for currencies for which there are capital controls restricting forward market activity, or for which an onshore forward market does not exist.
Typically, offshore markets for nondeliverable forwards are in New York and London, although there is also an active Asian market in Singapore and Hong Kong S.A.R. nondeliverable forwards require no physical delivery of currencies. At the inception date, parties set a price for the contract. At maturity, the contract is settled on the basis of a rate indexed to the underlying currency, so that only the difference between the agreed price and the realized price is exchanged. In the presence of capital controls intended to drive a wedge between domestic and international interest rates, nondeliverable forwards may contribute to evasion to the extent that, by offering a way to hedge against exchange rate risk, they increase the appeal of foreign credit obtained through delayed, anticipated, or misinvoiced trade payments.
Thus, the causes and consequences of capital account liberalization cannot be understood in isolation from the forces behind these broader movements. Insofar as capital account liberalization coincides with more comprehensive economic and financial liberalization—indeed, to the extent that these initiatives are all responses to the same fundamental forces—it is extremely difficult to isolate the effects of capital account convertibility. As Appendix III explains, many financial crises have occurred soon after domestic financial markets have been deregulated and restrictions on international financial transactions have been relaxed. Is it then the domestic financial deregulation or the capital account liberalization that “caused” these crises? The fact that some banking crises occur while the capital account remains controlled and that others—for example, the Savings and Loan crisis in the United States—have not had a significant international dimension is a reminder that external financial liberalization cannot be assumed to be at the root of all these problems, although detailed analysis suggests that, in many cases, precipitous or badly executed external liberalization has magnified the adverse effects of domestic policy problems.
Second, the policy on portfolio capital flows has differed from the policy on foreign direct investment. The preceding discussion, like much recent attention, has focused on policy on portfolio capital flows and on the role of technology in eroding the effectiveness of controls designed to limit the magnitude of portfolio capital movements. But while the rapid pace of change affecting the financial services industry may be undermining the effectiveness of controls on portfolio capital flows, detecting and controlling inward foreign direct investment remain relatively straightforward. Many countries continue to regard foreign direct investment (the purchase by foreign investors of “national champions” in particular) as a sensitive issue that warrants some control. As of the end of 1996, of 184 countries surveyed by the IMF, 144 still maintained some form of controls on foreign direct investment: 13 of 23 advanced industrial economies and 131 of 161 developing countries.
Controls on foreign direct investment take various forms, including special procedures for approving investment applications, regulating the repatriation of profits, and restricting the liquidation of the initial investment. In industrial countries, such controls generally focus on “strategic sectors,” such as telecommunications and banking, while in developing countries they cover a broader range of industries. That said, in recent years a number of developing countries have taken important steps to liberalize inward foreign direct investment. Thailand, for example, lifted most restrictions on inward foreign direct investment in import-competing industries in the 1970s and deregulated foreign investment in export industries in the 1980s. Chile similarly liberalized inward foreign direct investment between 1985 and 1989.
Third, it is important to clearly distinguish between controls on inward and outward capital flows. A large body of literature (reviewed in Appendix IV) casts doubt on the effectiveness of controls designed to inhibit outward portfolio capital flows, especially when a devaluation is anticipated. Investors who expect a discrete change in the exchange rate in short order can look forward to substantial (annualized) capital gains if they succeed in evading controls on outflows and limits on taking short positions in domestic financial markets. Given this strong incentive, it is perhaps not surprising that all but the most comprehensive and repressive controls on outflows have had at best limited effects.
Measuring Capital Account Liberalization
Providing a quantitative measure of capital account liberalization is difficult. In particular, it is hard to quantify restrictiveness of existing capital control measures. The left panel of the figure presents several measures of official restrictions on international capital flows in industrial countries, and the right panel presents an overall index of capital account openness, derived from the annual IMF publication Annual Report on Exchange Arrangements and Exchange Restrictions. The left panel shows the fraction of countries that impose three types of restrictions on international capital flows: restrictions on payments for capital account transactions, surrender of export proceeds, and multiple exchange rates. The panel on the right provides a measure of capital account openness constructed by Quinn and Inclán (1997) by coding measures limiting international capital mobility according to their degree of restrictiveness (outright prohibition being the most restrictive, followed by quantitative restrictions—approval required, taxation, and no restriction). The figures clearly show the increase in the degree of capital market liberalization, especially during the last decade.
The Capital Account in Industrial Countries
(Average across countries)
Source: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions.1 Based on IMF classification.2 Based on Quinn and Inclán (1997).The Capital Account in Industrial Countries
(Average across countries)
Source: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions.1 Based on IMF classification.2 Based on Quinn and Inclán (1997).The Capital Account in Industrial Countries
(Average across countries)
Source: IMF, Annual Report on Exchange Arrangements and Exchange Restrictions.1 Based on IMF classification.2 Based on Quinn and Inclán (1997).There is more empirical support for the notion that controls on capital inflows can have significant effects.21 When the motivation for capital movements is gradual portfolio diversification designed to attain a better trade-off between risk and return over time, for example, and not a scramble to avoid the large capital losses associated with an impending devaluation, it is more likely that controls and taxlike measures designed to influence the composition and volume of inward capital movements will have noticeable effects. The evidence is mixed. In particular, studies that distinguish the short run from the long run suggest that controls are likely to become progressively less effective the longer they remain in place, as investors and others discover new routes for circumventing them, leading the authorities in turn to broaden the coverage of existing restrictions to limit avenues for evasion.22 Still, a number of analysts conclude that measures adopted by countries like Chile, Colombia, and Israel to influence the level and composition of portfolio capital inflows have not been without effect.23
All figures are from the WEO database unless otherwise noted. The rise in gross flows, both to developing countries and generally, has been even more dramatic, increasing by 1,200 percent between 1983–88 and 1989–94. Johnston and others (1997) argue that gross flows are more informative than net flows for gauging the degree of and trends in capital market integration (because, for example, capital account liberalization may encourage domestic residents to diversify their portfolios by increasing their holdings of foreign securities at the same time as foreign investors, responding to similar incentives, step up their purchases of domestic securities, resulting in a significant increase in gross flows without necessarily a commensurate or even noticeable rise in net flows).
The fact that 140 countries account for a mere 5 percent of the private net capital flows to developing countries reflects in part the fact that their gross national products are relatively small.
A representative discussion of this era can be found in Folkerts-Landau (1997), pp. 234–38.
This discussion, like that in Appendix I, draws on Bordo, Eichengreen, and Kim (1998).
Freestanding companies were companies incorporated in the capital-exporting country for the express purpose of establishing a business presence in an emerging market. For details, see Appendix I.
This response can create agency problems of its own, given the problem of incentive compatibility that can arise between bondholders and equity holders. Thus, for example, management may tend to take on higher risk projects to benefit equity holders at the expense of debt holders. One interpretation of the historical experience is that this agency problem was less severe than the one described in the text.
Other technological advances, while not necessarily reducing information asymmetries, also encourage international investment. For example, increased computing power facilitates the construction and pricing of complex derivative securities, which allow investors to assume only those risks they wish to bear and to hedge the rest. Similarly, electronic trading allows investors to purchase small lots of securities at lower cost and thereby to diversify away concentrated risks. The declining cost of computing has also fueled the growth of the mutual fund industry, another channel through which small investors can diversify risks.
Even without advanced technology, it is difficult to distinguish transactions associated with the provision of trade credits and with the hedging of foreign exposures from purely speculative positions taken in anticipation of future exchange rate changes, complicating schemes to tax “hot money flows” while leaving productive foreign investment unaffected. See Folkerts-Landau and Ito (1995), p. 96.
Along with important warnings about the accuracy of these summary indicators.
With the exception of a temporary increase in the prevalence of multiple exchange rate practices following the breakdown of the Bretton Woods System of pegged but adjustable exchange rates.
Accompanied by a rise in the prevalence of measures requiring the surrender of export proceeds.
As noted in Appendix II, the same has been true of European countries that reimposed controls in response to currency crises in 1992–93. Johnston and others (1998) similarly note that although the pace of liberalization slowed in 1997, and some of the Asian countries most directly affected by the crisis reimposed a variety of restrictions on capital account transactions, the recent rise in volatility in emerging markets did not trigger a generalized resurgence of capital account restrictions.
The effects of which were supplemented and reinforced, by macroeconomic stabilization.
See Appendix IV and Folkerts-Landau and Ito (1995) for a detailed discussion of the experience with controls on capital inflows.
This point is emphasized in Johnston and others (1998).
Again, for details see Appendix IV.
