Lessons from Argentina and Brazil.

Calomiris, Charles W.

What have we learned from the sovereign debt crises in Argentina and Brazil, and what can the United States and the International Monetary Fund do, if anything, to repair the damage, and to avoid similar problems elsewhere?

Policy Lessons

I would emphasize five policy lessons:

* First, in emerging market countries (EMs), monetary policy--or, what amounts to the same thing, exchange rate policy--is often constrained by the need to finance government spending, which underlies the eventual collapse of the exchange rate.

* Second, even well-regulated banking systems are highly vulnerable to the risks of fiscal imbalance.

* Third, the IMF needs to stop intervening to prevent sovereign defaults when they are necessary.

* Fourth, EM debt capacity cannot be captured adequately by the ratio of sovereign debt to GDP. Export growth, and hence the need to follow through on trade reform, is just as important a fundamental determinant of debt repayment as discipline over government spending.

* Fifth, "contagion" among sovereign debtors is selective.

Fiscal Imbalance and Monetary Collapse

Unlike the United States or the European Union, where an independent central bank determines monetary policy, in most EMs, the policies of central banks are often determined by arithmetic--the arithmetic that requires debts to be monetized, because that is the only way that they can be repaid. When government debt grows too fast, the government is unable to repay debt service with future taxes, and the government forces debt monetization to occur. That problem is at the core of every exchange rate collapse of the recent and distant past. Typically, exchange rate depreciation precedes debt monetization because the markets anticipate the inevitable monetization that will occur.

Sometimes, fiscal imbalance does not show itself in government accounts. That was true of Brazil in the 1970s, which used off-balance sheet spending to disguise its fiscal imbalance (Brazil often ran an official fiscal surplus 'alongside high inflation in the 1960s and 1970s). Anticipated banking bailouts (which have been costing upward of 20 percent of GDP in the "twin-crises" countries of the past two decades) are the most frequent source of fiscal imbalance in recent crises. But Brazil and Argentina reached their current fiscal difficulties and weak currencies largely in the "old-fashioned way"--by failing to rein in measured government spending programs.

In Argentina, government spending grew substantially in the final years of the Menem administration, despite the crescendo of criticism of the debt run-up and the visible need to reform the infamous "coparticipation" system that hampered fiscal reform. And that debt was almost entirely denominated in hard currency, despite the lack of adequate growth in exports. The fiscal side of the liberalization cycle in these and other countries seems to follow a familiar path: liberalization and privatization result in new revenues for government and ebullient expectations about future growth in GDP, government revenues, and exports; market confidence in reform lowers the cost of accessing foreign capital for both the private sector and the public sector; EM governments cannot resist running deficits, but the fiscal imbalance grows and eventually catches up with them.

Initially, the response to this fact is denial, with assistance from multilateral lenders (and not only at the IMF--Ricardo Hausman was described by Walter Molano, a prominent market analyst of Latin American debt markets, as the lead salesman for Argentine government debt in the mid-to-late 1990s, when Hausman was chief economist at the Inter-American Development Bank (IDB). Then, the IMF "programs" grow in size, along with the anti-growth tax hikes that the IMF insists upon in return for providing "stability." At this point, debt yields rise and market "analysts" become largely political forecasters: "Will this debt swap provide a short-run profit for me? Will the IMF give us an exit in the not-too-distant future, and is the current yield high enough to bet on that exit?" Economists who refer to long-run arithmetic are dismissed. Investment banks' "research" departments cooperate with the masquerade because not doing so means that they will be cut off from millions in underwriting revenue from the new debt offerings or debt swaps. When the collapse comes, the IMF shakes its head about how unstable markets are, how irrational investors are--all the more reason, of course, to increase IMF footings.

Economic "emergence"--the combination of industry privatization, trade liberalization, price stabilization, and financial deregulation--would work much better if sovereigns did not see market optimism about their private-sector prospects as an opportunity to ramp up their expenditures. Imagine how much better off the people of Brazil and Argentina would be today if their governments had not been able to borrow in the international bond market during the 1990s.

What were government officials thinking? Wouldn't even a self-serving politician or bureaucrat do better in the long run by waiting until after growth had succeeded before increasing government expenditures? The problem is twofold. First, politicians do not have long-time horizons. That failure ultimately must be seen as a political failure of democracy in EMs. Second, local bureaucratic interests can be impervious to change even when politicians try to cut spending. That was particularly true in Argentina and Brazil, where spending was often decided at the local level but paid for at the national level.

Vulnerability of Banks in Emerging Markets

Argentina was one of the boldest and most successful reformers of bank regulation during the 1990s. By the late 1990s, the banking system had achieved (1) substantial foreign entry by European and North American banks, (2) substantial privatization of loss-generating provincial banks, and (3) real reform of deposit insurance, capital regulation, liquidity regulation, and other prudential regulation and supervision, which resulted in bank solvency, stability, and private market discipline over bank risk taking. These reforms were impressive, and were the result of years of hard work by Roque Fernandez, Pedro Pou, and their staffs at the Banco Central (Calomiris and Powell 2001).

But that very success produced a plum (in the form of banking system liquidity and net worth) that was ripe for government picking. Domingo Cavallo opined at a conference in 2001 that the regulatory system in Argentina was too good, that banks were forced to maintain too much liquidity and capital. He "fixed" that "problem" by forcing changes in the Banco Central's personnel and rules to reduce banks' liquidity, and more important, he used those freed-up banking system resources to absorb ever more government debts by forcing banks to "participate" in new government financing schemes. Ultimately, we learned in Argentina that when sovereigns are at the end of their rope, they'll use all the power at their disposal--including financial sector supervisory and regulatory powers--to address their short-term needs, irrespective of the long-term consequences for the financial system or the economy. That Cavallo would destroy the Argentine banks was predictable as early as April 2001 (Calomiris 2001a, 2001b, 2001c) because he was unwilling to recognize the necessity of default and debt restructuring. His example was imitated recently by President Chavez of Venezuela, who encouraged runs on Venezuelan banks to penalize them for being unwilling to buy as much government debt as he wanted them to buy.

Sovereign Defaults May Be Necessary

The IMF played the role of facilitator of Cavallo's doomed plans in 2001 because Stanley Fischer believed that debt default would necessarily produce an exchange rate collapse, and so the IMF decided to take a very poor odds bet on Cavallo's "plan." Fischer understood that depreciation was not a way out for the highly dollarized Argentine economy, and so he was willing to do almost anything to avoid depreciation. Fischer was right to want to avoid depreciation, but his resistance to debt renegotiation was an error in economic reasoning. The best chance for Argentina in 2001 was default. If the IMF and the G7 had supported default by Argentina, if default had been accompanied by credible government expenditure reform and trade liberalization, and if the IMF had offered liquidity assistance to support the currency board during the debt restructuring, then Argentina could have written down its debt and avoided currency collapse. (Supporting sustainable fixed exchange rates, after all, is the original mandate of the IMF, isn't it?). The combination of credible expenditure reform and sovereign debt reduction would have obviated the long-run need for depreciation. A likely return to the precrisis trend growth rate in labor productivity would have avoided the need for long-run real exchange rate adjustment via deflation.

A reverse Dutch auction, combined with the use of "exit consents"--preferably with an IMF-established floor on debt values, as proposed by Lerrick and Meltzer (2001)--probably would have resulted in a speedy restructuring process by minimizing holdout problems and avoiding lengthy discussions with formal creditor committees.

This approach might not have been politically feasible, of course, but we will never know that because it was never tried. A "share-the-pain" program for Argentina in early 2001 that would have simultaneously reduced debt service, liberalized trade, and reformed and reduced expenditures might have been able to attract political support internally and externally. Both foreign creditors and domestic residents surely would have been better off under this approach, and many creditors I spoke with recognized the advantage of quickly reducing the amount of debt by late 2000.

Part of the IMF's opposition to this proposal, of course, reflected concern about the size of the liquidity support that they might have had to provide and the risk of loss to the IMF. Those are legitimate concerns, but it is hard to see how the alternative policy that they pursued--which placed the IMF and other official creditors at risk of losing more than $20 billion--was a better bet, even from the narrow perspective of the IMF's risk of loss.

Importance of Trade Liberalization

A fairly reliable sign of deterioration in EM debt prices, which was illustrated by both Argentina and Brazil, is the attempt by officials of the finance ministry or the central bank to point to their moderate debt-to-GDP ratios as indicators of their low risk of default. The Argentine finance minister argued throughout 2001 that Argentina could repay its debts, and that it was unfairly being singled out by critics. His "proof" was that at a roughly 50 percent ratio of national government debt-to-GDP, Argentina was in the middle of the pack of sovereign debtors. True, and also not relevant, for two reasons. Markets look at trends in expenditure, which forecast future debt levels, not just at current levels. The ramping up of government spending after 1995, and market doubts about the political feasibility of constraining spending (doubts that, in the event, proved correct) were more relevant than the current debt ratio.

Furthermore, an EM whose public and private borrowing is largely in dollars must not only achieve domestic fiscal balance (slow growth of government deficits relative to tax collections), it must also satisfy an external constraint requiring that future hard currency debt service be paid with net export receipts. Countries cannot run external Ponzi schemes, depending on ever-growing future capital inflows to pay preexisting hard currency debt service: eventually exports have to grow relative to imports to pay those debts. Argentina and Brazil never achieved the export-to-GDP ratios that they should have achieved, and that they could have achieved, if domestic politics had not constrained trade liberalization and labor reforms. In Argentina, even if export growth had been higher, the lack of discipline over domestic public finances probably would have produced default in any case. The lesson, of course, is that trade liberalization can help to stabilize economies. The practical difficulty, again, is the lack of a long-time horizon in domestic politics. That produces both the debt run-up and the resistance to removing protectionist policies that prevent export growth. Free trade may benefit everyone in the long run, but in the short run it can pose political costs for liberalizers, which discourages trade liberalization.

Contagion Is Selective

It is worth remembering that this lesson, like the preceding four, is not news. After all, Chile had largely avoided the fallout from the Mexican crisis of 1994-95. And Singapore did not collapse during the Asian crisis of 1997. But file lack of uniformity in sovereign yield changes in the wake of the Argentine collapse, and the subsequent decline in Brazilian debt values has made it especially clear that contagion in the wake of one country's crisis is not a result of mass hysteria but rather of fundamental changes in markets that affect some countries more than others. Contagion reflects three kinds of channels that link other countries to crisis countries: export and import links (either through competition among exporters, as in the fallout for Indonesia and Malaysia after the Thai collapse, or import market linkages, as in the case of Uruguay's reaction to Argentina's crisis), direct financial links between the crisis country and other countries (which were of some importance for explaining the transmission of the Russian crisis to Brazil, the Argentine crisis to Uruguay, and the Asian crisis to Korea), and indirect, general-equilibrium financial links associated with global portfolio rebalancing (which explain much of the problems experienced by Brazil and Argentina in the wake of the Mexican crisis, and of Brazil in the wake of the Russian crisis).

Indirect, portfolio rebalancing links are an international form of what is sometimes called a "credit crunch" or a "capital crunch" when it occurs within a domestic banking system. Financial institutions (banks, hedge funds, and others) rely on creditors as sources of their funding, and those creditors have limited tolerance for increases in default risk. When institutions suffer large losses in one category of their assets (e.g., Mexican bonds), they must restore the low default risk on their own debts by selling other risky assets. Countries that issue risky debt (like newly liberalized Argentina and Brazil as of 1994) are especially vulnerable to asset sell-offs that seek to restore financial institutions' creditworthiness in response to losses elsewhere.

The selective nature of contagion is important because it implies that countries can, to a large extent, control their own destinies through export market growth and diversification, and by placing limits on the growth of government debt during the early stages of liberalization. Furthermore, domestic bank regulation can play an important role in reducing vulnerability to either internal or external shocks. Insolvent Korean banks' willingness to invest in high-risk Indonesian securities, for example, was largely a reflection of the weakness of Korean bank regulation. And the weakness of the Korean banking system, which was known to be insolvent before the Asian crisis, further fueled capital flight once the crisis spilled into Korea. Of course, implementing the key policy lesson--that banking systems must be strengthened--is extremely difficult for the same reason that it is hard to limit government expenditures or to ensure continuing trade liberalization. Weak domestic bank regulation is a convenient means to channel hidden taxpayer subsidies (via the banking system "safety net") to banks and borrowers taking excessive risks at taxpayers' expense. Those favored few are given special access to subsidies through their access to bank charters and bank credit. The iron triangle of banks, government, and large firms that control the political process and the allocation of credit through the financial system (a system that is often referred to, in shorthand, as "crony capitalism") is the reason for resistance to financial sector reforms that would limit risk taking.

In summary, the lessons of the Argentine and Brazilian crises are not new, nor is there much likelihood that the recent crises will increase the chances that these lessons will be learned in Brazil, Argentina, or in other countries. The central problem, both before and after these crises, has not been a lack of available information from which to draw useful policy conclusions, but rather, the lack of domestic political will within EMs to implement policies that avoid financial crises. The multilateral institutions (the IMF, the IDB, and the World Bank) on balance have not played a helpful role in Argentina or Brazil, and surely global political economy is part of the explanation for that fact. Still, it would be wrong to place the primary blame for failure on the IMF, the IDB, or the G7.

The last thousand years of human history indicate that there is no means to sustained economic growth or to long-term political reform other than the establishment of a competitive and internationally open economic system (Jones 1988, Landes 1999, Maddison 2001, Calomiris 2001d). The primary constraint against progress in this direction, now and in the past, has been resistance to such reforms by entrenched rulers and special interests that do not see short-term advantages from improving the long-term lot of the average citizen. In this sense, there is nothing new to be learned from the crises in Argentina and Brazil.

What We Can Do

What can we in the United States, through G7 policies, the IMF, and other agents, do to encourage developing country emergence, growth, and stability? Trade policy is the most important single influence we can have. The effect of liberalizing our own barriers to EM imports would be far and away the most important source of "aid" the G7 could provide EMs (IMF and World Bank 2001).

How can IMF, World Bank, and regional development bank policies be improved? I continue to believe that the solutions advocated by the Meltzer Commission (IFIAC 2000), on which I served, are the best path forward, although they do not constitute a complete set of reforms. The emphasis in our report was to focus each of the multilaterals on separate areas of policy (with the IMF focusing narrowly on liquidity support, not bailouts, and the other agencies providing aid to support institution building, poverty alleviation, and global public goods), to provide more flexible and effective means for channeling such assistance (e.g., through grants in many cases, rather than loans), and to establish credible means of holding multilaterals accountable for their performance. The World Bank is moving in this direction, and the IMF has made some changes in its lending practices that are consistent with our recommendations, but overall, the multilaterals are still muddling through without the proper focus, accountability, or mechanisms necessary for achieving desired results.

Is there a need for the IMF to use its authority to change the rules of sovereign workouts and to become a bankruptcy process facilitator or arbiter to assist in the resolution of sovereign debt crises? The good news is that policymakers in Washington are finally realizing that "participation" in losses by private sector participants and speedy debt restructuring are important to restore growth and to avoid excess risk taking by private investors in anticipation of IMF bailouts (the "moral-hazard" overlending problem). But having finally arrived at the conclusion that the IMF should avoid bailing out insolvent sovereign debtors or prolonging the resolution of unsustainable debt burdens, the IMF now is leaping to the conclusion that a formal debt resolution process must be established by statute and that it should manage debt resolutions. Neither of these conclusions is warranted, at least for now. A better policy would be to reform IMF intervention, and see whether private market creditors and sovereigns can sort out their differences sufficiently well without the IMF. That contractual approach would rely on existing techniques and institutions, perhaps with some contractual modifications going forward (like the adoption of collective action clauses in bonds), once the IMF has adopted a more limited role in debt crises.

Opponents of the IMF bankruptcy proposal worry about several possible problems. First, the IMF is "conflicted" because it is a creditor. It has become quite possible that Argentina and Brazil will default on IMF loans (the Argentines have even announced that they would do so before using their remaining reserves to repay the IMF), and long-term losses to the IMF are now quite possible, if not likely. Some private creditors regard that as a positive development because it might limit their own losses. Some academics also regard that possibility favorably because it would chasten the IMF and possibly discourage it from continuing to promote bailouts. The IMF, of course, regards its own lending as properly senior to all private and Paris Club (bilateral government) debts. But, regardless of one's view on these matters, surely the IMF is in a conflicted position, and is not the appropriate entity to influence decisions about whether and how much official creditors should share the pain of debt restructuring.

Second, it is likely that a bankruptcy process would entail substantial discretionary authority for the arbiter of that process and that could create a new political "football" within an already highly politicized organization (the IMF), despite IMF claims that it would only handle procedural matters and not make important disrectionary decisions. And it would be hard to imagine an alternative bankruptcy arbiter that would be immune to political capture. Would a system that would favor some countries and creditors in its bankruptcy proceedings, for political reasons, be a desirable alternative to the current already politicized system? That is a hard question to answer, but thinking about it certainly leads one to look for other alternatives.

Third, establishing a bankruptcy process may add costs to sovereign debt spreads. That is not entirely a bad thing if you view EM sovereign overborrowing as a key problem to be avoided (as I do). Still, it would be preferable not to add costs unnecessarily, but instead, to reform sovereign debt markets by eliminating IMF bailouts, which would provide adequate incentives for creditors to take a long-term view of sovereign risk and limit the supply of credit to EM sovereigns.

Fourth, there are alternatives to the IMF bankruptcy proposal that avoid most or all of its shortcomings and achieve most or all of its objectives (especially the important objectives of coordinating the multiplicity of sovereign claims, and avoiding holdout problems during renegotiation). Alternatives to bankruptcy entail creditor-debtor renegotiation without any third-party involvement. The alternative contractual approach might be enhanced by greater use of majority voting clauses or, alternatively, a reliance on existing contracts and procedures combined with innovative renegotiation strategies (reverse Dutch auctions with exit consents) to overcome holdout problems and consolidate disparate debts.

It is beyond the scope of this paper to review in detail the relative merits of the contractual approach, except to note that there is every reason to believe, based on logic, evidence, and judicial precedents, that the contractual approach could work, even without mandating majority voting clauses (see Buchheit and Gulati 2009 for a history of workout arrangements that suggests various options). The chance for success of this approach would be increased if the IMF provided bona fide liquidity assistance to sovereigns during debt workouts (as distinct from IMF bailouts). Lerrick and Meltzer (2001) have outlined a simple, creative, and promising approach to providing such liquidity assistance and have argued that doing so not only would provide incentives for swaps (by giving bidders liquid markets in which to trade), but also would make it harder for "vulture" funds to acquire controlling positions in "orphan" debt issues. Simulations of this renegotiation process involving real market participants have been very encouraging.

It is also possible that it would be desirable to encourage or even require the inclusion of collective action clauses (CACs) in future sovereign debt offerings. The argument for the efficiency of CACs is that having them in place would resolve many of the legal uncertainties posed by the current use of exit consents and other similar devices. Of course, arguing for the efficiency of a contractual feature is not the same as arguing that it should be required. If it is so efficient, then why must it he mandated?

The most plausible argument for requiring governments to adopt CACs is that doing so would have a positive effect on IMF behavior (an argument frequently made by IMF officials with respect to the Sovereign Debt Restructuring Mechanism, or SDRM). The IMF would be more likely to avoid counterproductive bailouts if its bureaucrats, and the G-7 finance ministers who control them, were able to see a clear and viable alternative to bailouts in the form of an orderly restructuring process. The presence of CAGs would make that alternative clearer. Of course, that possibility would not necessarily appeal to either creditors or to sovereign debtors, who might prefer IMF bailouts to the economic and political hardships that accompany debt write-downs. That preference would lead the market to avoid CACs precisely because they would be socially beneficial.

Under the assumption that CACs would add some cost to debt issues (an argument that has yet to be proven), there are additional arguments in favor of public policy to encourage or require CACs.

First, if EM rulers are myopic (as I argued above), then they will tend to prefer cheap credit today over credit that is cheaper in the long run. That is, they will undervalue the contingent benefits to their citizens of CACs (which lower costs of renegotiation), particularly if they fear that such clauses will raise market yields (there is significant concern on the parts of the government of Mexico, Brazil, and others that this would occur). Even if CACs are efficient, if they are costly, myopic sovereigns may not choose to adopt them voluntarily.

Second, to the extent that CACs would limit negative externalities across borders (so-called contagion effects) their benefits would not be fully internalized by issuers and their creditors, which might also lead to an underuse of socially desirable CACs. (1)

Third, there is an important distinction between potentially beneficial IMF assistance in establishing new rules for collective action by creditors, and inappropriate IMF incursions into restructurings, in the form of its proposed SDRM. In particular, the IMF should not be in the business of establishing statutes that effectively change ex ante contracts agreed upon by debtors and creditors. I understand that the existence of debts without CACs may prove inconvenient during the transition period toward a new set of rules, but in my view that inconvenience (even if it produces one or two cases of protracted delay in debt restructuring) is well worth preserving the basic sanctity of contracts. And, there are far better ways of dealing with that transition problem (e.g., encouraging the swapping of old non-CAC debt for new CAC debt).

Fourth, we should recognize that we have little empirical basis for the belief that a contractual approach, even without majority voting clauses, would not work better. Do we really know that vesting classes of creditors with veto powers over restructuring plans (as the IMF proposes to do in its SDRM mechanism) would resolve holdout problems faster than the exertion of raw power by sovereigns in debt swaps? I do not believe that it would. Sovereigns have much more power to

force agreement with a restructuring plan than do private debtors. Some point to the Argentine crisis as evidence that informal renegotiation under current rules cannot work. That is a strange argument since the Argentines, under IMF tutelage, resisted trying that approach. What we have learned over the past several crises is that even under the current international policy regime (which includes IMF bailouts and excludes bona fide IMF liquidity assistance) there have been some successes in informal renegotiation (notably, in Ecuador).

Fifth, given the little we really know about how well renegotiations could work in a world without IMF promotion of bailouts or delays under any of the three proposed new regimes (contractual renegotiation without CACs, renegotiation in the presence of CACs, or a formal bankruptcy process), and given the irreversibility of establishing a bankruptcy process, it seems advisable to follow a cautious approach. That means trying the contractual renegotiation approach first, and contemplating far-reaching, hard-to-reverse changes (like an IMF bankruptcy process) only after it has been clearly demonstrated that contractual renegotiation along one or more of the lines suggested by Buchheit and Gulati (2002) and Lerrick and Meltzer (2001) really cannot work. A rush to adopt the SDRM might forestall beneficial private adaptations within the contractual approach that would otherwise occur. That is essentially the policy position taken by the four Shadow Financial Regulatory Committees of Europe, Japan, Latin America, and the United States (2002) in their joint resolution on SDRM.

A gradual approach to reform would also have the distinct advantage of displaying some modesty on the part of the IMF and the G7 about their ability to redesign the financial world on the basis of their theories and get it right. If only IMF bureaucrats could approach their work with the spirit of humility that Friedrich A. Hayek recommended:

 What we must learn is that human civilization has a life of its own,
 that all our efforts to improve things must operate within a working
 whole which we cannot entirely control, and the operation of whose
 forces we can hope merely to facilitate and assist so far as we
 understand them. Our attitude ought to be similar to that of the
 physician toward a living organism [Hayek 1960: 69-70].

(1) Even though sovereigns and market participants seem to fear that collective action clauses would substantially raise borrowing costs, that is not obvious. It may be that, by reducing costs of renegotiation, collective action clauses would even lower borrowing costs. That seems quite possible, especially since, in my view, the presence of such clauses would have little effect on the probability of a sovereign default.

References

Buchheit, L. C., and Gulati, G. M. Gulati (2002) "Sovereign Bonds and the Collective Will." Working Paper No. 34. Washington: Georgetown University Law Center (March).

Calomiris, C. W. (2001a) "Argentina Can't Pay What It Owes?" Wall Street Journal, 13 April.

--(2001b) "How To Resolve the Argentine Sovereign Debt Crisis." Working Paper. Washington: American Enterprise Institute (April).

--(2001c) "The Fastest, Fairest Way Out of the Argentine Debt Crisis," Wall Street Journal, 9 November.

--(2001d) A Globalist Manifesto for Public Policy. Occasional Paper 124. London: Institute for Economic Affairs (July).

Calomiris, C. W., and Powell, A. (2001) "Can Emerging Market Rank Regulators Establish Credible Market Discipline? The Case of Argentina, 1992-1999." In F. S. Mishkin (ed.) Prudential Supervision: What Works and What Doesn't? 147-96. Chicago: University of Chicago Press.

Hayek, F. A. (1960) The Constitution of Liberty. Chicago: University of Chicago Press.

International Financial Institutional Advisory Commission, IFIAC (2000) Report. Washington: Government Printing Office (March).

International Monetary Fund and World Rank Staff (2001) Market Access for Developing Countries" Exports. Washington: International Monetary Fund (April).

Jones, E. L. (1988) The European Miracle. Cambridge: Cambridge University Press.

Krueger, A. O. (2002) "Preventing and Resolving Financial Crises: The Role of Sovereign Debt Restructuring." Paper presented at the Latin American Meeting of the Econometric Society, San Paolo, Brazil, 26 July (www.imf.org/externa]/speeches/2002/072602.htm).

Landes, D. (1999) The Wealth and Poverty of Nations. New York: W.W. Norton.

Lerrick, A., and Meltzer, A. (2001) "Beyond IMF Bailouts: Default without Disruption." Quarterly International Economics Report, Carnegie Mellon Gailliot Center for Public Policy (May).

Maddison, A. (2001). The World Economy: A Millennial Perspective. Paris: Organization for Economic Cooperation and Development.

Shadow Financial Regulatory Committees of Europe, Japan, Latin America, and the United States (2002) "Joint Statement on Reforms in the Process of Restructuring International Sovereign Debt," 7 October (available at www.aei.org).

Charles W. Calomiris is the Henry Kaufman Professor in the Division of Finance and Economics at the Columbia University Graduate School of Business. He codirects the Project on Financial Deregulation at the American Enterprise Institute.