Copies of this paper and those written by the Center for International Environmental Law (CIEL) and the Third World Institute (Uruguay) in the same series are also available in Spanish by request.  Please contact the authors for more information.

EXECUTIVE SUMMARY

INTRODUCTION

I.   How Might the FTAA Encourage Unregulated Capital Mobility?

II.  Capital Mobility and Financial Crisis

III. The Need for Capital Controls and "Speed Bumps"

• Some Country Experiences with Capital Controls   
•     1. Chile
•     2. Colombia
•     3. The Use of Capital Controls in Response to the Asian Crisis   

RECOMMENDATIONS FOR FTAA INVESTMENT NEGOTIATORS

Appendix.  Chart 1.  Net Private Capital Flows to Developing Countries of the Western Hemisphere 1991-1999

REFERENCES

    Recent experience with international capital markets suggests that it is not an appropriate time to negotiate a new commercial agreement which would limit signatory countries' abilities to regulate the flow of capital across international boundaries.

    But if the investment rules that emerge from negotiations on the Free Trade Area of the Americas (FTAA) are based on investment rules in the North American Free Trade Agreement (NAFTA) and the Multilateral Agreement on Investment (MAI) - as it now appears that they will be - the FTAA would do just that. Provisions in the NAFTA and the MAI on the definition of investment, transfers, and national treatment would severely restrain the ability of governments to regulate even short-term speculative capital flows, in order to prevent or manage a financial crisis.

    The recent Asian crisis shows the effects which may result from sudden capital outflows. In East Asia, there was a net reversal of private international capital flows to the region of $105 billion-- from a net inflow of $92.8 billion in 1996 to a net outflow of $12.1 billion in 1997. This amounts to about 11 percent of the GDP, before the crisis, of the combined economies of South Korea, Indonesia, Malaysia, Thailand and the Philippines. This is a massive and highly destabilizing reversal of international capital flows, and it does not appear to be the result of changes in the real, underlying economies of the region.

    Financial liberalization measures (often adopted under pressure from the OECD, IMF, or Washington) enacted by many Asian countries in the years prior to the 1997 crash bear much of the responsibility for the region's financial crisis. The result, made worse by IMF austerity measures, was a regional depression, increased unemployment and poverty, and increased political and social tensions that may persist for years to come.

    In Mexico, most of the capital inflows in the four year period leading up to the peso crisis were comprised of so-called "hot money" - mainly short-term bank lending and portfolio investment. At the end of 1994 there was a shift in investors' sentiment, resulting in large scale capital flight. In 1995, Mexico experienced a net capital outflow of US $15 billion, as compared to a net inflow of $31 billion in 1993. The impacts of this reversal on the Mexican economy were severe: the economy contracted sharply, unemployment rose, and poverty rates soared.

    Some countries have attempted to protect themselves from the harmful effects and instabilities inherent in a world of increasingly unrestricted capital flows. As recent experiences with capital controls in Chile, Colombia, Malaysia, Hong Kong, Taiwan, and China demonstrate, efforts to limit capital mobility are not inconsistent with orderly and robust economic growth and development. To the contrary, using capital controls can be a highly effective strategy in a country's efforts to return to economic growth after a financial crisis, or to prevent excessive short-term international borrowing, for example, from building up in the first place. However, if the deregulatory model (based on the NAFTA and the MAI) for international investment is followed, these and other types of controls would be mostly prohibited by the FTAA.

    A number of leading economists have entered the debate over global capital flows and the need for more national controls. Along with Columbia's Jagdish Bhagwati, MIT's Paul Krugman and Harvard's Jeffrey Sachs, World Bank Chief Economist Joseph Stiglitz has repeatedly spoken out about the dangers of further capital account liberalization. Moreover, prominent Latin American economists Manuel Agosin and Ricardo Ffrench-Davis have conducted extensive research into the role of capital flows leading up to the Mexican crisis, and the effectiveness of Chile's capital controls. They argue that governments should adopt "speed bumps," reserve requirements, and other types of capital controls to shield their economies from the destabilizing effects of short-term capital flows.

    In short, compelling evidence suggests that it is an inappropriate time to negotiate a treaty that makes it easier for capital to flow uninhibited across national boundaries. But the investment chapter of the proposed Free Trade Area of the Americas (FTAA), if based on the NAFTA/MAI investment model, would accelerate the pace of investment liberalization and capital mobility in the Western Hemisphere. It would restrict the ability of governments to shield themselves from the volatility of global financial markets. At a time when the control of international capital flows is the subject of heightened controversy and debate, the measures currently promoted for inclusion in the FTAA would prohibit most capital controls and similar regulatory measures. Before negotiating any such treaty, we should take stock of the impact of speculative capital flows as well as international capital mobility more generally. It would be unwise and counter-productive to negotiate a treaty that ignores recent experience with deregulated capital flows.

INTRODUCTION

    The events of the last several years have highlighted the risks and complications of increasing international capital mobility. The role of highly volatile capital flows in precipitating the Mexican peso crisis (1995), and more dramatically the Asian, Russian, and Brazilian financial crises of the last two years, has caused a major shift in the thinking of economists. As recently as two years ago, the majority of the profession believed that increasing liberalization of capital flows, like the opening of economies to freer trade, was inherently beneficial. But today most economists would qualify any such statement considerably, and there has been a broad expansion in the number and types of capital controls that economists would be willing to consider in order to avoid or ameliorate the types of financial crises that we have recently witnessed.

    However, policy makers have been slow to adapt to these changes in economic analysis and circumstances. They have continued, for the most part, to pursue through commercial agreements and other venues, measures that would increase the liberalization of capital flows.

    The International Monetary Fund asked its executive board in September 1997 to complete an amendment to the IMF's Articles of Agreement that would make liberalization of capital movements one of the mandates of the Fund. The Multilateral Agreement on Investment (MAI), an investment liberalization treaty that was under negotiation at the OECD until December 1998, would have similarly mandated, through its broad definition of investment and language on transfers, liberalization of signatory countries' capital accounts.⁽²⁾ The MAI, in turn, was modeled on Chapter 11 of the North American Free Trade Agreement (NAFTA), which contains a broad definition of investment, as well as language on financial transfers which effectively precludes most capital controls. The Free Trade Area of the Americas (FTAA) is the next venue where capital account liberalization will probably be mandated. If, as is anticipated, the investment rules in the FTAA are based on those in the NAFTA and the proposed MAI, the agreement could severely limit the ability of signatory governments to employ capital controls or otherwise protect themselves from the dangers of financial crises and destabilizing capital flows. It is therefore worth examining these investment rules in the light of recent economic events and analysis.

I. How Might the FTAA Encourage Unregulated Capital Mobility?

    The FTAA's Negotiating Group on Investment has as its mandate "to establish a fair and transparent legal framework to promote investment through the creation of a stable and predictable environment that protects the investor, his investment and related flows, without creating obstacles to investments from outside the hemisphere."⁽³⁾ The negotiating group is in the process of drafting the FTAA's investment chapter primarily on the basis of the investment provisions the North American Free Trade Agreement (NAFTA), but it will likely also draw upon the MERCOSUR treaty and the MAI. The Group will also use the recommendations submitted by the Business Forum of the Americas, a coalition of multinational corporations from the United States and South America.

    The relevant sections in each of the agreements cited above relating to a nation's ability to impose capital controls are the definition of investment, and provisions on national treatment, as well as transfers. The language in the FTAA on these issues will have the most impact on nations' abilities to regulate capital inflows and outflows.

A. Definition of Investment

    Trade agreements and investment liberalization treaties have generally defined investment broadly in recent years. The NAFTA, in article 1139, defines investment to include enterprises, equity and debt securities, loans, income and profits, and real estate or other property, tangible or intangible. The draft MAI defines investment as "every kind of asset owned or controlled, directly or indirectly, by an investor..." This includes property, shares and stocks, claims to money, loans, and intellectual property rights.⁽⁴⁾

    These definitions are very broad and include not only Foreign Direct Investment (FDI), but also portfolio investment and other forms of short-term lending. The most volatile kinds of flows included in these definitions are short-term financial credits to banks and large domestic firms, short-term deposits by non-residents in the domestic financial system and purchases of stocks and bonds by non-residents. These types of transactions are more likely to be for the purpose of taking advantage of interest rate differentials or to get quick capital gains. It is these types of short-term and especially speculative capital flows that are most likely to exacerbate a crisis when they are sharply reversed. (See Chart 1, Appendix, for a depiction of trends in direct and portfolio investment in developing countries in the Western Hemisphere). On the other hand, foreign direct investment, which generally includes a controlling interest and often involves the creation of tangible assets such as the construction of factories, is more likely to remain invested for a longer time.⁽⁵⁾ When economists argue for the benefits of investment liberalization, it is generally FDI that they rely on to make their case.

    Because the definition of investment in the NAFTA and the MAI are broad enough to cover portfolio investment and other forms of short-term lending, the liberalizations and protections of Foreign Direct Investment (FDI) in these agreements will apply equally to speculative portfolio flows. Of particular concern are the provisions liberalizing the transfers of capital relating to investment and the national treatment protections of investment.

B. Transfers

    The NAFTA mandates the unrestricted transfer of capital relating to investments including principal, profits, dividends, interest, and capital gains.⁽⁶⁾ The investment chapter of the FTAA, based on other treaties and proposed language, may include a broadly worded section on transfers as well, without distinguishing between short-term flows and other forms of investment. If the Business Forum of the Americas' recommendations are adopted, the FTAA may go beyond even existing language on transfers to "design effective measures for free transfer and beyond-border transactions" to ensure that no controls are applied to investment entering or leaving the host country.⁽⁷⁾

C. National Treatment

    National treatment - the requirement that foreign investors be treated no less favorably than domestic investors, regardless of the circumstances - is another fundamental principle of investment liberalization treaties; it is a provision of the NAFTA and the proposed MAI. Because of the broad definition of investment, this provision could be used to protect highly mobile foreign capital from actions taken by national governments in the interest of protecting their economies. National treatment provisions would permit a foreign bank or other investor to argue that it has the same rights as domestic investors and banks under international law.⁽⁸⁾

    The recommendations of the 1998 Business Forum of the Americas included broadening the definition of national treatment. The Forum advocated the extension of national treatment to taxation and suggested that the FTAA "extend uniform non-discriminatory national treatment to capital originating from countries outside the FTAA."[Emphasis Added]⁽⁹⁾ Under such an arrangement, investors from Europe and Asia, for example, would be entitled to the same treatment rights as investors from within the hemisphere. As will be shown in the examples below, national treatment provisions, as applied to portfolio and other short-term investments, risk increasing the vulnerability of national economies to financial turbulence and its harmful economic and social effects.

    The provisions on the definition of investment, transfers, and national treatment that the FTAA is likely to include would almost certainly create major obstacles to any government seeking to regulate short-term speculative capital flows, even to prevent or manage a financial crisis.⁽¹⁰⁾ At the same, the potential dangers of short-term capital inflows are well-documented in the recent economic literature on the subject. The crises in Asia and Mexico provide a useful introduction to the story of unchecked capital movements.

II. Capital Mobility and Financial Crises

A. The Role of Capital Flows in the Asian Crisis

    Prior to the onset of the Asian financial crisis, there were few mainstream challenges to the maxim that the deregulation of international capital flows was in the best interests of everyone. This principle seemed almost as well established as the theory of comparative advantage with regard to trade, and was able to benefit from a sort of "proof by association" with the latter. This is in spite of the fact that the reasoning of the trade theory does not apply to capital flows,⁽¹¹⁾ not to mention the very limited usefulness of the theory of comparative advantage itself to any kind of economic development strategy.

    The debate over the effects of deregulated international investment intensified when the economies of South Korea, Indonesia, Malaysia, Thailand, and the Philippines, and others in the region were hit by a financial crisis that subsequently developed into a regional depression. Although the policies of the IMF helped transform the financial crisis into a crisis of the underlying real economy, even pro-globalization economists have noted that the financial liberalization of these countries was a major proximate cause, if not the major cause, of the onset of the crisis. Jagdish Bhagwati, one of the world's leading international economists and the Economic Policy Adviser to the Director-General of the GATT (1991-93) has noted that "the Asian crisis cannot be separated from the excessive borrowings of foreign short-term capital as Asian economies loosened up their capital account controls and enabled their banks and firms to borrow abroad. . . it has become apparent that crises attendant on capital mobility cannot be ignored."⁽¹²⁾

    What was so striking about this case is that it was truly "the intrinsic instability in international lending"⁽¹³⁾ that pushed these countries to the abyss. Most important was a net reversal of private international capital flows to the region of $105 billion-- from a net inflow of $92.8 billion in 1996 to a net outflow of $12.1 billion in 1997. This amounts to about 11 percent of the GDP, before the crisis, of the combined economies of South Korea, Indonesia, Malaysia, Thailand and the Philippines.⁽¹⁴⁾ This is a massive and highly destabilizing reversal of international capital flows, and it does not appear to be related to the workings of the real, underlying economies of the region.⁽¹⁵⁾

    Financial liberalization measures enacted by many Asian countries in the years prior to the 1997 crash bear much of the responsibility for the inability of governments to control the massive net reversal of capital flows out of the region in 1997. For example, the Korean government began to relax its control over financial flows in the early 1990s. In order for South Korea to join the OECD in 1996, it had to liberalize its financial sector even further.⁽¹⁶⁾ The deregulation of financial transfers provided for in NAFTA and the proposed MAI would, most importantly, 1) increase the ability of affected economies to rapidly acquire accumulations of short-term debt and 2) remove from national authorities the ability to enact capital controls when the resulting panic sets in.

    Other explanations of the onset of the Asian crisis, especially the media sound bites about "crony capitalism" or "inefficient" industrial organization do not have much evidence to back them up. Indeed, when people, including some economists, refer to the Korean economic system as "inefficient" it is not clear what they mean. The most obvious economic meaning would be that resources were allocated inefficiently, so that the economy (and therefore living standards) did not grow in accordance with its full potential. The South Korean economy grew at a per capita rate of 7.2% over the last thirty years, one of the highest rates of economic growth in the history of the world. It is certainly possible that it could have grown even faster, but no one has presented an economic argument as to how this might have been accomplished.

    All this is not to say that there were no problems accumulating in these five Asian countries. There was a build up in domestic bank lending in all of the countries except Indonesia, where firms increased their borrowing directly from foreign banks. Real exchange rates did appreciate noticeably - about 12% for South Korea and 25% for the other four countries - as large international capital flows poured in. But other countries have had much larger real appreciations without the kind of currency collapse that these countries underwent. And it should be stressed that these weaknesses as well as the current account deficits were very much tied to the liberalization of capital flows that took place in the preceding years.

    Radelet and Sachs⁽¹⁷⁾ have also examined the effect of international shocks, such as the devaluation of the Chinese yuan in 1994, the increased competition from Mexico, and the overcapacity in particular industries such as semiconductors. The combined effect of these influences does not appear to account for what happened.

    It is therefore difficult to escape the conclusion that the instability caused by recent international financial liberalization bears the primary responsibility for the onset of the crisis. The reversal of capital flows amounting to eleven percent of the regional GDP was a result of "herd" behavior, with foreign and domestic investors alike stampeding for the exits for fear of being caught with greatly depreciated local currency and assets.

    The logic of such panics is fairly straightforward. In the Asian crisis, it began with the fall of the Thai currency, then soon spread to other countries. With a high level of short-term international debt, a depreciation of the domestic currency increases the cost of debt service. Everyone needs more domestic currency to get the same amount of dollars for debt service, and the selling of domestic currency to get those dollars or other "hard" currencies drives the domestic currency down further. It does not take much to set off a panic, especially if the central bank does not have a high level of foreign currency reserves relative to the short term debt. These reserves shrink further as more and more investors convert their domestic currency and domestic assets into dollars. Foreign lenders refuse to renew their short-term loans (or, as in South Korea, are able to even call them in early), and the downward spiral continues.

    Other economists have found that the inherent instability of international financial markets was a major cause of previous financial crises, including Mexico's in 1994.⁽¹⁸⁾ And Radelet and Sachs' statistical analysis of recent crises in emerging markets found that the most important predictor of crisis was the ratio of short-term international debt to the country's foreign exchange reserves.⁽¹⁹⁾ In other words, these countries became vulnerable to panic-induced capital outflows, as well as runs on their currency, because of a build-up of short-term international borrowing.

    This build-up of short-term international borrowing was thus a direct result of the financial liberalization that took place in the years preceding the crisis. In South Korea, for example, this included the removal of a number of restrictions on foreign ownership of domestic stocks and bonds, residents' ownership of foreign assets, and overseas borrowing by domestic financial and non-financial institutions.⁽²⁰⁾ Korea's foreign debt nearly tripled from $44 billion in 1993 to $120 billion in September 1997. This was not a very large debt burden for an economy of Korea's size, but the short-term percentage was dangerously high at 67.9% by mid-1997.⁽²¹⁾ For comparison, the average ratio of short-term to total debt for non-OPEC less developed countries at the time of the 1980s debt crisis (1980-82) was twenty percent.⁽²²⁾

    Financial liberalizations in the other countries led to similar vulnerabilities. Thailand created the Bangkok International Banking Facility in 1992, which greatly expanded both the number and scope of financial institutions that could borrow and lend in international markets. Indonesian non-financial corporations borrowed directly from foreign capital markets, piling up $39.7 billion of short-term debt by mid 1997, eighty-seven percent of which was short-term.⁽²³⁾ On the eve of the crisis the five countries had a combined debt to foreign banks of $274 billion, with about sixty-four percent in short-term obligations.

    In sum, it is clear that international financial liberalization played a significant role in bringing about the Asian financial crisis, from which the region has yet to recover. We now turn to the Mexican peso crisis, and recent experience with international capital flows in the Americas.

B. The Mexican Peso Crisis and Latin America's Experience with Liberalized Capital Flows

    Latin American nations have become accustomed - for better or for worse - to the boom-and bust cycles that have come to exist in a world of growing transnational capital flows. In the 1970s and early 1980s, many countries in the region received large inflows of foreign capital. At the same time, countries in the region undertook liberalization of their financial sectors and relaxed or eliminated foreign exchange regulations. The inflows dried up in the 1980s as the debt crisis discouraged lending to the region. After the debt crisis subsided, however, Latin America again received large capital inflows beginning in 1991. In the 1990s, foreign capital inflows have reached between 5 and 10 percent of GDP in Argentina, Brazil, Chile, and Mexico.

    During the periods when capital inflows were plentiful (from 1976-1981 and 1991-94), macroeconomic imbalances emerged towards the ends of the cycles. Throughout both periods, currencies appreciated and current account deficits grew in many countries. As a result, some of these countries became very vulnerable to the actions of external creditors. The creditors in turn became sensitive to even the slightest bit of "bad news."⁽²⁴⁾

    This was the case in the lead-up to Mexico's financial crisis in December 1994. Access to external capital grew very quickly in the early 1990s - in 1991, 1992, and 1993 annual net capital inflows represented more than 8% of Mexico's GDP.⁽²⁵⁾ These large capital inflows supported an overvalued peso as well as chronic current account deficits. These factors combined to make the economy very vulnerable to external shocks.

    Most of the capital inflows to Mexico in the four year period leading up to the crisis were comprised of so-called "hot money" - mainly short-term bank lending and portfolio investment. In 1993, portfolio inflows alone amounted to 7.7% of GDP, with FDI inflows amounting to just 1.4% of GDP (See table 1). In the four-year period leading up to the crisis, Mexico's net external debt grew by US $92 billion, of which only US $24 billion was Foreign Direct Investment (FDI).⁽²⁶⁾

    Mexico was especially hard hit by the inflow of short-term capital in the 1990s because of its laissez-faire approach to capital inflows. Beginning in the mid-1980s, Mexico's capital account was dramatically liberalized. Before 1988, foreign portfolio investment had been strictly regulated.⁽²⁷⁾ After 1988, due to the lifting of capital account restrictions, there was a surge in foreign portfolio investment. The liberalization was part of a larger effort to liberalize the economy (including trade liberalization, privatization of state-owned enterprises, and removal of restrictions on foreign ownership). Many of these policy changes were required in order for Mexico to accede to the GATT, become a member of the OECD, and join the NAFTA, all of which Mexico did in the early 1990s.⁽²⁸⁾

    The crisis was precipitated in 1994 largely as a result of the U.S. Federal Reserve's doubling of short-term interest rates in the U.S. from February 1994 to February 1995, which caused Mexico's bond market to become relatively less attractive to investors. Investors grew even more wary when the Zapatista uprising in Chiapas and the assassination of leading presidential candidate Luis Donaldo Colosio suggested political instability. The result: market sentiment changed, and both foreign and domestic financial capital stampeded for the exits.

    The reversal of capital flows was drastic. In 1995, Mexico experienced a net capital outflow of US $15 billion, as compared to a net inflow of $31 billion in 1993. In just the fourth quarter of 1994, there was a net outflow of $5.5 billion of portfolio investment, and Mexico was facing a $17.7 billion (4.2% of GDP) balance of payments deficit at the end of the year. Foreign exchange reserves fell from $24.9 to $6.1 billion, and the Mexican peso collapsed, losing a third of its value in the last 10 days of 1994, and half of its value by the end of 1995.⁽²⁹⁾

    The impacts on the Mexican economy were drastic, although they were certainly exacerbated by the IMF austerity policies - including enormous interest rate hikes and budget cuts - that followed. The economy contracted by 6.6% in 1995 and investment dropped 30%. As interest rates soared, companies (especially small businesses) failed, debt burdens became unpayable and layoffs ensued.

    By mid-1995 official unemployment had doubled as compared to two years earlier. From 1994 to 1995, there was a 20% increase in the proportion of people living in extreme poverty.⁽³⁰⁾

    Economic growth resumed in 1996, but much of the crisis-induced decline in living standards persisted or worsened for the bulk of the population. By 1997 poverty levels had barely come down. Real wages for manufacturing workers plummeted by 39% from 1994-1997,⁽³¹⁾ even as manufacturing employment grew.

    As in the Asian economic crisis, it is difficult to separate the effects of the financial liberalization, the ensuing crisis and currency collapse, and the IMF austerity policies.⁽³²⁾

    Some observers have laid the blame on major policy errors, in particular the maintaining of an over-valued currency for the years prior to the crisis.⁽³³⁾ Others, such as Robert Blecker have made a strong case that there are fundamental contradictions in the whole growth and development model pursued by Mexican governments since the early 1980s, not only in their exchange rate policies, but in macroeconomic and other policies which thwarted its attempt to harness foreign trade and investment as the engine of economic growth (Blecker 1996).

    Nonetheless it is clear that the Mexico's financial liberalization played a key role in enabling not only the reversal of capital flows that precipitated the crisis, but the imbalances that were involved (e.g. the overvalued currency and swelling trade deficit-- $28.8 billion in 1994, from a $4.2 billion surplus in 1987).

    The Mexican crisis also had an impact on the rest of the region. Annual GDP growth was zero for Latin America in 1995 and unemployment rates rose that year in Argentina, Costa Rica, Paraguay, Uruguay, and Venezuela. The crisis had a particularly large impact in Argentina and Uruguay, where GDP fell 5.0% and 2.3%, respectively, in 1995. Argentina also opened up its capital account in the early 1990s and pursued a laissez-faire approach like Mexico, which at least partially explains that country's heightened sensitivity to the Mexican crisis, in comparison with other economies in the region.⁽³⁴⁾

    Not all Latin American economies were affected as severely as Argentina by the Mexican crisis. Chile and Colombia, for example had different policies towards large inflows of foreign capital. Rather than adopting complete financial and capital-account liberalization, these countries maintained various kinds of controls and restrictions on short-term inflows. In 1995, for example, the Chilean economy grew by 10.6% and Colombia by 5.8%; it is possible that their capital controls helped insulate them from the fallout of the Mexican peso crisis.

III. The Need for Capital Controls and "Speed Bumps"

    In light of the experience with recent financial crises, which ensued largely because of sudden reversals of international capital flows, a growing number of economists have begun to consider the need for capital controls. In fact, some countries in Asia and Latin America already have such controls in place, despite the general opposition to such measures by the United States, the IMF, and private financial institutions. As recent country experiences with capital controls demonstrate, these measures are not inconsistent with orderly and robust economic growth and development. To the contrary, capital controls can be a highly effective tool in a country's efforts to return to economic growth after a financial crisis, or to prevent excessive and volatile capital inflows from building up in the first place. If the deregulatory model (i.e., based on the NAFTA and the MAI) for international investment is followed, however, these controls would be prohibited by the FTAA's investment provisions.

A. Some Country Experiences with Capital Controls

1. Chile

    In Latin America, Chile is the most frequently cited example of a country whose application of capital controls protected its economy from speculative capital inflows. In the early 1990s, as large amounts of capital entered the country, authorities adopted a range of policies to regulate the inflow of capital. In June 1991, a non-interest bearing reserve requirement of 20% was established on external credits. This meant that foreign investors had to place sums equal to 20% of their investment into a zero-interest account with the Central Bank, with the requirement that these reserves be maintained at the Bank for a minimum of 90 days. A stamp tax (at an annual rate of 1.2% on operations of up to one year) that had formerly only applied to domestic loans was applied to external credits as well. As upward pressures on the currency grew in May 1992 (mainly due to growing capital inflows), authorities raised the reserve requirement to 30%. In October, the central bank increased the length of the reserve requirement to one year regardless of the loan's maturity.⁽³⁵⁾

    With these measures, Chile sought to protect itself from the destabilizing effects (e.g., excessive currency appreciation) of capital inflows by discriminating against the least desirable and volatile inflows. Discrimination against short-term flows is inherent in the reserve requirement. Because the requirement is fixed regardless of the term of the credit, the cost of the requirement decreases with the duration of the investment. For example, with the reserve requirement at 30% and an international interest rate (LIBOR) at 5%, the cost of the reserve requirement to an investor who borrows abroad to invest in Chile for a one-month period is 29%; for a two-month period, the cost falls to 13.5%. At the same rate, the cost of borrowing falls to 2.1% for a year-long investment term and to .2% for a ten-year term.⁽³⁶⁾

    Another important feature of Chile's policy is its flexibility. As seen, the requirements were phased in - but they were also designed to move up or down, depending upon the amount of funds coming in at a given time. An example of this flexibility was demonstrated as the current account deficit grew in 1998 and investors became less interested in Chile (due to a steep drop in the price of copper, its chief export, and the effects of decreased demand in Asia). As a result, Chilean authorities reduced the reserve requirement from 30 to 10 percent. Even with the decline in the reserve requirement, the policy still served as a barrier to the most speculative and volatile forms of foreign capital inflows.⁽³⁷⁾ Since September 1998, Chile has temporarily dropped the reserve requirement to 0% - based on the fact that there are currently no speculative, short-inflows. But the authorities have reserved the right to re-establish the reserve requirement when necessary.

Table 1. Net inflows of foreign capital to Chile, Colombia and Mexico, 1991-1994, as a percentage of GDP

Source: Agosin and Ffrench-Davis 1996. Data from Central Bank of Chile, Banco de la Republica (Colombia) and CEPAL.

58. Krugman, October 5, 1998.