International Monetary Cooperation Since Bretton Woods

Debt Relief
More radical action seemed appropriate; and it could no longer do the systemic harm that might have resulted earlier. Since 1982, banks had tried to build up reserves against developing country debt. To the extent that national tax systems allowed write-offs to be offset against profits, the governments of creditor countries (or their taxpayers) slowly in fact helped to bail out the banks. Some banks also sold parts of their Third World portfolios at discounts on initially rather thin secondary markets. In 1985 the total sale of international debt was estimated at less than $5 billion, but in 1988 it amounted to some $30–40 billion.111 In some cases also debt was exchanged for equity participations. Chile in particular was an innovator in this kind of swap. The capital exposure (credits in relation to bank capital) of the nine largest U.S. banks to Latin America fell from the alarming 177 percent of 1982 to 84 percent by 1988.112 But even a default on this reduced figure would have clearly constituted a major threat to the U.S. banking system. By 1988, however, it had become clear that this was rather unlikely, Latin America would not move as a bloc, and there would be no debtors’ cartel or “debtpec.” The spirit of Cartagena disappeared. Two attempts in particular by debtor countries to alter the working of the system in the event actually underlined its new stability.
In Peru, Alan García Pérez had been elected President in 1985 on a nationalist economic platform. His government faced a clearly impossible situation. Debt interest and repayments due in 1985 amounted to $3.7 billion, more than Peru’s total anticipated exports ($3.1 billion).113 In his inaugural speech, the new President complained about the difficulty of exporting in view of the protectionism of the industrial countries, and then announced that he would limit debt service to 10 percent of export earnings. There would be no “intermediaries” in debt negotiations. “President Alan Gatcía is an elected President and must report only to 20 million Peruvians and not to the officials of an international organisation … At this time of difficulties and austerity, we will propose a policy that will require efforts for some time but that policy will no longer be imposed by the IMF.” He also appealed to other Latin American governments: “A united Latin American continent will succeed in making the wealthy countries recognise that they are also responsible for this crisis, and that they must lower interest rates, extend repayment periods, and maintain our exporrs’ prices … We welcome and support the consensus of Cartagena as a decisive step for unity.”“114 In fact, Peru’s action completely failed to shake the international system.
Brazil like Argentina had attempted a currency stabilization (the Cruzado plan); it failed to establish confidence, and in February 1987 President José Sarney announced a unilateral moratorium on interest payments on Brazil’s $108 billion outstanding debt. It produced no concessions, and it did not unduly harm the bankers. In 1988 Sarney admitted, “The fact is that we cannot destroy the international system.… We can scratch it, but it can destroy us.”115
Neither Brazil nor Peru seemed to hold out attractive models of the gains to be achieved by debtor default; and in this way, perhaps paradoxically, their actions increased rather than reduced the confidence of bankers in the operating and the permanence of the system. The Brazilian action in fact formed the backdrop to the announcement on May 20, 1987, by the new chairman of Citicorp, John Reed, that he had increased the bank’s loss-reserves by $2.5 billion, and that he hoped to sell one third of the bank’s developing country portfolio. Citicorp’s action was followed almost immediately by a number of smaller New England banks, as well as by appeals from bank regulators not to dispose of debt on the existent but weak and small secondary market.
Instead, banks began to work out new agreements that included an element of debt relief or reduction. The reform proposals could now emanate from bankers rather than from politicians and academics. In 1987, Mexico worked with Morgan Guaranty Trust on an exchange of bank debt at a discount for 20-year bonds collateralized with U.S. government zero-coupon bonds. In the fall of 1988 the Speaker of the Deutsche Bank, Alfred Herrhausen, publicly called for a partial cancellation of commercial debts,116 a course he had been advocating in private for over a year. The following year, the Deutsche Bank proposed debt relief through debt-equity conversions, as well as auction-style buybacks by debtor governments with funds made available by creditor governments and multilateral institutions, and concentrating on those sectors of the debtor economies whose financial rehabilitation was most needed in order to restore operational efficiency.117
In informal discussions in 1987, at the IMF and World Bank Annual Meetings, the Japanese Finance Minister Kiichi Miyazawa proposed the conversion of debt to equities and the establishment of a global “risk insurance” against noncommercial risk. The multilateral institutions would take on a role as clearing centers for debt-equity swaps.118
When Nicholas Brady, James Baker’s successor as U.S. Treasury Secretary, proposed in a speech of March 10, 1989 to the Bretton Woods Committee Conference a plan for voluntary debt reduction, he was systematizing but also extending initiatives that had in practice already been undertaken. According to his proposal, a quarter of IMF and World Bank policy-based lending was to be used for discounted debt buybacks. The Fund and the Bank together could provide $20–25 billion. An additional $10 billion would be provided by Japan. But, apart from this, the commercial banks and the Fund were to be disengaged from each others’ commitments. Banks would be given a variety of options, including accepting lower interest rates or selling claims to the debtor at a discount. As a result, the plan was usually described as “menu-based.” The IMF would “lend into arrears”: its lending need not be dependent on an agreement of the debtors with the commercial banks (although there would usually have been some negotiation). (In fact, already in October 1987 an IMF stand-by arrangement with Costa Rica had been concluded that did not include a requirement for an agreement over arrears to private bankers.) In effect, the linkages of 1982 were now partly undone. The Fund no longer was to act as a “policeman for the banks.” These proposals were formalized and accepted as a new debt strategy at the meetings of the IMF Interim and Development Committees in April 1989.
Three negotiations were concluded in 1989 under the revised debt strategy (for Mexico, the Philippines, and Costa Rica), another three in 1990 (for Morocco, Venezuela, and Uruguay). In 1992 another operation was conducted with Nigeria, and agreements concluded with Argentina and (in principle) with Brazil, as well as with Mozambique and Niger. The Brazil deal was only finalized in 1994. These conversions were conducted in conjunction with IMF programs. The IMF agreements were a condition for the sale of U.S. Treasury bonds to collateralize the long-term conversion bond option. Only in the Brazilian case was there no parallel IMF package. Brazil’s Brady deal as a result was interpreted as signaling the partial withdrawal of the Bretton Woods institutions and the “end of the debt crisis.”119
The largest so-called Brady deal, for Mexico, can be used as a model to illustrate the features and possibilities of negotiated debt reduction. The options for banks included swapping old debt for 15-year “discount bonds” at 100:65, or a 1:1 exchange for “par bonds” with a reduced interest rate of 6.25 percent. Banks that chose neither of these options would agree to provide “new money” in order to compensate for the increase in value of remaining debt as debtor countries became more creditworthy. The new bonds would be transferable, could not be rescheduled by the debtor, and the principal would be collateralized with U.S. zero-coupon bonds (that is, the Mexican authorities held the U.S. bonds in an amount corresponding to the principal due). The principal was guaranteed, along with 18 months of interest, through the use the special funds provided by the World Bank ($2.06 billion), Japan ($2.05 billion), and the IMF ($1.64 billion) together with a direct guarantee from the Mexican government. There was an additional provision for “value recovery,” involving higher payments on the bonds if Mexican oil revenue exceeded a threshold. Altogether $49 billion debt was rescheduled, with a total relief amounting to $12–13 billion, from which should be deducted the $7 billion cost of providing collateral for the new bonds. Measured as a proportion of Mexican GDP, this may not appear as a large amount of relief (2.7 percent of GDP).120
It helped, however, quite decisively in eliminating the phenomenon of “debt overhang,” in which the level of debt was so high as to discourage new capital inflows. This phenomenon was only widely recognized as an obstacle to recovery at a relatively late stage in the international debt crisis. At a certain point, debt becomes so large that much higher levels of taxation are required to make service payments. This reduces the likelihood of countries growing out of the debt crisis, and thus also lowers the incentives for debtors and creditors to reach agreement. A certain measure of debt forgiveness, as a result, could change the climate of negotiations completely. After negotiations on debt reduction had begun, paradoxically the value of debt on secondary markets rose dramatically.121 For the original 15 Baker Plan countries, it increased from an average of around 30 in 1989 to above 50 by the beginning of 1992.122 As a result of the return of confidence, capital markets revived, and new bond issues (but not new bank lending) began. Already in 1989, Mexico had raised $1.4 billion through bond issues; in 1991 it took $4.1 billion. For developing countries as a whole, capital raised on the bonds market increased from $25.1 billion to $40.8 billion over this period.123 Flight capital also began to return. An estimate for Mexico is $5.7 billion in 1989 and $5.8 billion in 1990.124 Direct investment responded to the establishment of a more favorable economic environment. An essential element in the new confidence was the market-based or choice-oriented debt reduction of the Brady Plan, and the absence of a compulsory write-off of the kind advocated by many analysts in the mid-1980s.
Such a possibility had indeed been held out as a route to the re-establishment of external viability even in the early stages of the debt crisis. At the end of 1983, de Larosière had explained that “foreign direct investment could play a far greater role in resource transfers than in the past. Not only does such investment avoid creating an overhang of debt, it often facilitates the transfer of technology and skills and is directly tied to productive capital formation.”125 In fact, for the major debtor countries, foreign direct investment had fallen off with the uncertainties surrounding the onset of the general debt crisis, and only revived again with policy reform in the late 1980s, with major inflows to countries that had pursued reform most vigorously. (See Figure 12-3.) At the same time, portfolio investment also surged after 1990, encouraged by the process of privatization in capital importing countries, and creating incentives for further and more radical privatization. Again, a relatively small number of countries accounted for the major part of the flows: from 1989 to 1993, two thirds went to five countries (Argentina, Brazil, Korea, Mexico, and Turkey). Roughly half of the private flows in the period 1990–94 took the form of portfolio investment, while half came as foreign direct investment. In this way, the risk shifted from a concentration in the public to a dispersion over the private sector, and the “de-accumulation” of risk has been a major factor in overcoming the debt crisis.126
Figure 12-3. Foreign Direct Investment: Argentina, Brazil, Korea, Mexico, and Turkey
(In billions of U.S. dollars)

Source: World Bank, World Debt Tables.
A further contribution to the solution of the debt crisis came from the lowering of international interest rates. In the late 1980s, it was estimated that a 2.5 percentage point drop in interest rates would save Latin America as a whole $6 billion a year, or a resource saving of 8 percent of imports. Such a change would clearly improve economic performance. Another calculation showed a 2.5 percent increase in LIBOR as reducing growth rates in developing countries by 0.5 percent.127 In practice, three-month LIBOR fell from an average of 10.25 percent in April 1989 to 3.19 percent by May 1993 (after which date it began once more to rise). This reduction was less a result of any calculated plan designed with the debt issue in mind, or even of the greatly improved international policy coordination of the later 1980s, than a consequence of world economic circumstances. Discrepancies in national policies, such as existed at the beginning of the decade, are a cause of increased uncertainty, and can result in higher interest rates as governments attempt to compensate investors for that uncertainty. Increasing convergence, providing that it was accompanied by fiscal consolidation, might be expected to produce lower rates. Above all, it was the effect of the recession in the industrial countries that drove down interest rates, alleviated the debt crisis, and thereby paradoxically increased the prospects for faster development in the rest of the world.
By the beginning of the 1990s, the complete pessimism that had characterized the debt discussion as recently as 1987 and 1988 had faded. Countries had discovered that sound policies would eventually reopen international capital markets. The mechanism of the Brady Plan, with its offer of a multiplicity of choices, had helped to make it clear that there was no longer one single phenomenon of “developing country debt.” Instead, as in domestic lending, there existed a whole range of differing qualities of investment. The financial experience of the 1980s also made it clear that other forms of lending—for instance, against real estate in Texas, or New England, or California or New York, Tokyo, and London—could be quite as risky as lending to countries whose development potential held out at least a possibility of rapid growth. After the property fiasco of the early 1990s, the British property developer Martin Landau (whose collapsed firm required the largest single write-down in the history of banking) was quoted as saying: “Banks are very lazy. They won’t lend it to farmers, they don’t understand manufacturers, they won’t lend to the third world, so they come back to lending to property. Bankers are definitely not very bright people.”128 Then, by the mid-1990s, another public sector debt problem appeared. Persistent fiscal imbalances, analogous to those of Latin America 15 years earlier, and the scale of future liabilities as a consequence of social security, health, and pensions commitments to an aging population, led to worries about the stability of major industrial countries, and produced dramatic upheavals in the bond markets.
Banks indeed should have learned a new lesson by the end of the debt decade: anything could be risky. In fact, the developing country debt crisis—unlike the property boom and bust in the United States, the United Kingdom, Japan, and Scandinavia, and then, with a slight delay, in France and Germany—did not bring down a single major financial institution. New or alternative forms of capital movement offered a more stable alternative to short- or medium-term commercial bank credit. After the mid-1980s new financial flows began to developing countries. Direct investment, which had remained through the period of the debt boom and bust at rather low but constant levels, increased dramatically, year by year, after 1986.
Does this mean a return to the “anarchy of the market”? What can prevent a repetition of past harmful cycles of indiscriminate overlending followed by panic? Two sorts of solution can be offered: a regulatory one, dealing on an international level with commercial banking practice; and a systemic one, addressing the world economy as a system in which government policies, whether appropriate or inappropriate, are in constant interplay with a broad spectrum of commercial judgments.
The outbreak of the 1980s debt crisis encouraged a renewed interest in financial regulation across frontiers. One important U.S. initiative extended the regulatory debates that had taken place in the course of the 1970s. It was now thought that banking supervision should be carried out internationally in a more systematized and open way than the tentative, informal, and secret coordination and exchange of regulatory experience carried out in the framework of the Basle Committee on Banking Regulation and Supervisory Practices (which had begun functioning in 1974). Part of the pressure came from domestic opinion in the United States, and from the widespread fear that a solution to the international debt problem would involve a bailout of the banks. The consequent reluctance of the U.S. Congress in 1983 to increase IMF quotas (as part of the Eighth General Review) could only be overcome through a legislative measure that extended control of banking. The International Lending Supervision Act (November 1983) gave U.S. regulatory authorities the legal right and the obligation to “cause banking institutions to achieve and to maintain adequate capital by establishing minimum levels of capital.” It also included a request for American regulators to pursue a convergence of international bank capital standards. Subsequent negotiations, first with the United Kingdom, then with Japan and the other G-10 members, produced a set of standards embodied as the Basle Accord of 1988.129
The Accord required an 8 percent ratio of recognized capital to credit-weighted risk exposure by the end of 1992. Five different categories of relative riskiness were established: loans to official borrowers from the OECD, as well as to countries with an arrangement with the IMF through the General Arrangements to Borrow (which, in practice, added Saudi Arabia to the OECD list) were given a zero risk weight; loans to non-OECD governments had a 100 percent weight. Like any arbitrary categorization, this system encountered obvious objections: would it not, for instance, act as a deterrent to developing country borrowing, but also to borrowing by industrial corporations, and would it not make lending to industrial country governments too attractive? There exists both a scope and a demand for the evolution of a more sophisticated and sensitive system

and imports were overinvoiced. The total short-term outflows from Argentina during 1982 were estimated at $5 billion. Already in August it became clear that Argentina could not sustain this radically inflationary course; the monthly cost of living increase was 160 percent; and, on August 25, 1982, the Treasury Minister directly responsible for the emergency program resigned. The course of inflation now proceeded without interruption. Indeed, some of the measures taken as part of a preparation for a transition to democracy, such as the lifting of restrictions on trade union activity, helped to perpetuate an inflationary momentum and a balance of payments crisis.
The ratio of scheduled debt service to exports exceeded 100 percent for 1982 as a result of the collapse of Argentine trade; and for 1983 and 1984 interest payments on the foreign debt involved an estimated 8 percent of GDP. These figures were based on the information then available, while in fact the total extent of the Argentine liabilities was very hard precisely to determine. Official figures gave the sum of public and private debt registered at the end of 1982 with the Central Bank as $37 billion. But it also became cleat in the course of 1982 that this was an underestimate and that the military regime had borrowed heavily between 1976 and 1982 through overseas branches of the Banco de la Nación.
At the November 16, 1982 meeting with bankers at the New York Federal Reserve, where the Mexican crisis undoubtedly provided the main theme, the IMF’s Managing Director also attempted to set out what would be the Argentine financing requirements. The current account deficit for the last quarter of 1982 and for 1983 would amount to some $2 billion. An amount of $1.8 billion would be available from the IMF; the bankers would have to put up $3 billion, of which de Larosière estimated that half could be financed through liquidating arrears and unwinding swaps involving Argentine businesses. There would be a short-term bridging loan from the banks of $1.1 billion, representing 5 percent of the banks’ exposure to Argentina, and the provision of new money amounting to 6.8 percent of exposure.
The program accompanying the SDR 1,500 million IMF stand-by arrangement approved on January 24, 1983 involved the reduction of the public sector net borrowing requirement to 8 percent of GDP in 1983 and 5 percent in 1984, with an increase in revenue financed