Nowhere has the history of sovereign debt been more dramatic than in Latin America and the Caribbean, where dependence on often unpredictable international capital flows poses risks that have often resulted in financial and economic grief. The nineteenth century, starting from the immediate aftermath of the independence of the Latin American republics, was characterized by ambitious schemes of government borrowing for infrastructure development and other public projects. All too often these waves of borrowing culminated in market panics, debt crises, and sovereign defaults, notably in the 1830s, 1850s, 1870s, and 1890s (and in the twentieth century as well, in the 1930s and the 1980s). These crises left Latin American and Caribbean countries shut out of international financial markets and often were highly disruptive to their economic development. From the very start, then, sovereign debt has been a mixed blessing for Latin America and the Caribbean.

The current phase of Latin American sovereign debt history started in the early 1990s, when the Brady exchanges converted external bank debt into bonds, established emerging market bonds as an asset class, and opened the door for the re-emergence of bond markets as a source of external finance. This marked a return to the first era of globalization of 1880–1914, when Latin American sovereign bonds were a major component of a thriving global market for financial instruments issued by emerging economies. The current phase is evolving: Brady bonds have been almost entirely retired well ahead of their maturity dates, and domestic debt markets have started to emerge as the venue of choice for sovereign finance. This shift in the composition of public debt has led some observers to conclude that Latin American countries are gradually reducing their vulnerabilities, on the grounds that

foreign-currency-denominated external debt held by international investors is more sensitive to global factors than domestically issued debt held by resident investors.

More generally, current trends point to a steady enlargement of the menu of financing sources and instruments. With the acceptance by increasingly sophisticated investors of contingent provisions in bond covenants—of clauses indexing returns to a country’s rate of growth or the occurrence of a natural disaster, for example—it will become easier for prudent governments to manage higher levels of debt. That these higher levels apply to prudent governments should be emphasized, because the proliferation of instruments (derivative securities, for example) also increases the scope for things to go wrong. That said, the development of new instruments opens opportunities for debt managers to achieve superior combinations of expected cost and risk for any given level of public debt, relative to what was available before.

At the same time, the development and deepening of financial markets is also enhancing access to international financing by private borrowers from emerging economies, including those of Latin America. Traditionally, foreign borrowing was done by governments. During the early 1980s, more than two-thirds of the stock of all debt owed by Latin American and Caribbean countries to private international lenders was debt incurred by the public sector. By 1990, that share had risen to nearly 90 percent. (See Figure 14.1, which shows analogous figures for East Asia for purposes of comparison.) But in recent years, private firms and banks have been drawing finance from international markets in record amounts. While domestic bond markets are still overwhelmingly dominated by public issuers, private sector issuers are beginning to gain access (see Chapter 7). There are visible signs, in other words, of the development of economically significant corporate bond markets, although the small size of many Latin American firms still limits their access to bond finance.

Policy analyses of public debt management in Latin America and the Caribbean should be placed in this context. The fact that a growing range of private sector entities are now able to access debt finance, both at home and abroad, implies a diminished need for public borrowing. The fact that others can borrow on domestic and foreign markets weakens the argument that the government must borrow for them, whether to finance investment in infrastructure and productive capacity or for the purpose of smoothing consumption spending across good and bad times.[1]

This suggests that globalization and financial market development create two forces that influence the role of public debt, one expanding it, and the other reducing it. The availability of a wider range of instruments and a broader investor base increases the scope of the sovereign to borrow safely and finance its operations. Conversely, the private sector’s growing access to financial markets diminishes the traditional role of the state as intermediary for such financing. These two opposing tendencies operate with different degrees of intensity in different countries. Imagining the role of public debt in the twenty-first century therefore requires analyzing these dynamic forces and developments in individual countries in more detail.

WHY DO SOVEREIGNS BORROW?

The history of debt crises in Latin America underscores the risks involved in sovereign borrowing. Is government debt a threat to a country’s financial stability? Should governments therefore give up borrowing altogether, or limit their debt to such small amounts that any risk of financial instability is ruled out? A judicious answer to these questions should acknowledge that the risks of sovereign borrowing should be balanced against the benefits, that is, the functions that governments fulfill and whose execution requires resorting to debt. Chapter 1 discussed three primary economic justifications for government borrowing: (1) to redistribute income from wealthier future generations to those currently alive, (2) to fund development projects, and (3) to fund policies aimed at smoothing the effect of business cycles and other shocks.

The first rationale for public borrowing, borrowing for redistributive purposes, should be assessed within the framework of the political motivations for government borrowing, discussed in Chapter 9. The existence of shortsighted politicians, together with the fact that future generations are not represented in the current decision-making process, creates a tendency toward excessive current consumption and hence overborrowing. A case in point is social security systems, which often run deficits, taxing future generations to cover the shortfall created by the level of benefits provided to current recipients under the retirement system. Moreover, the progressive development of private insurance and annuity markets will eventually weaken the case for a government role in retirement benefits, or at least create the opportunity for alternative systems that do not involve the buildup of future public liabilities. At a minimum, the presence of such markets should help in evaluating the actuarial fairness of retirement systems and the extent of redistribution from future to current generations that such systems involve.

The second rationale, namely, development borrowing, seems self-evident. There can be little doubt that developing countries would benefit from an increase in investment in human and physical capital of such a magnitude that it cannot be financed entirely from current taxes. Indeed, there is evidence that infrastructure bottlenecks are particularly serious in Latin America and the Caribbean (see Chapter 10). But does this justify public borrowing and spending? Although many infrastructure projects have a clear public-good content, the private sector can now provide more of the investment needed to finance them.[2] Indeed, the data indicate that private infrastructure projects have increased substantially over the last 20 years, although their growth has not fully offset the observed decline in public investment. This shortfall provides an immediate explanation for the persistence of infrastructure bottlenecks.[3] Another area in which the government should retain an important role is human capital investment. Here, there are problems that private capital markets are unlikely to solve. In practice, it simply is not possible, for example, for poor households, lacking negotiable collateral, to finance an education by borrowing against their offspring’s future earnings.[4]

While the development of private capital markets has also weakened the argument for government spending to smooth the impact of business cycles and other shocks, the case remains strong. In downturns, households and firms experience reductions in the value of their tangible collateral, limiting their ability to borrow to smooth consumption and investment. Moreover, when contemplating whether to borrow, they have no reason to take into account the aggregate (macroeconomic) effects, which accrue to the country as a whole in the form of an externality. Large shocks, such as those associated with natural disasters and financial rescue operations, also require abrupt increases in government spending. Such increases cannot be financed out of current revenues; it is desirable to smooth over long periods the impact of such events on tax rates, which implies a prolonged period of public indebtedness. The problem here is that the credit risk premium on borrowed funds often rises sharply in bad times, precisely when the country is most in need of financing. The increase may be sharpest on global markets, which is where it makes the most sense to seek financing during a domestic recession. Thus, instead of borrowing more during periods of weakness, emerging economies often must cope with a more limited and more costly supply of finance.[5]

SOVEREIGN DEBT IN THE TWENTY-FIRST CENTURY

The conclusion that emerges is that there are still legitimate reasons for governments to borrow, although imperfections in political systems and financial markets pose the risk that the recourse to sovereign debt may be used poorly or unduly restricted. To limit the vulnerabilities that result from sovereign debt and maximize its economic value, policy should be directed at three targets:

• Setting essential controls on the flow of debt. This implies creating a fiscal policy framework that ensures that debt stays within sustainable levels.

• Appropriately managing the inherited stock of debt. This involves using a combination of debt instruments to minimize vulnerability to a debt crisis and to lessen the constraints imposed by debt on monetary and fiscal policies while keeping the cost of debt service at acceptable levels.

• Improving the international financial environment in which these decisions occur. This entails reforms to the global financial environment to make sovereign borrowing safer.

CONTROLLING THE FLOW OF DEBT

What can countries do to better manage the ongoing process of debt accumulation? Domestic reform should start with measures to ensure that governments borrow for the right reasons (i.e., for tax smoothing, high-return infrastructure investment, or socially desirable intergenerational redistribution, as described above). In contrast, borrowing on an ongoing basis to, inter alia, pay the salaries of redundant civil servants is not sound practice.

Political motivations and electoral considerations can distort borrowing decisions, however, as discussed in Chapter 9. Political and procedural reforms and greater fiscal transparency can help to limit problems associated with such distortions. A large empirical literature now shows that more centralized fiscal procedures that leave less autonomy to spending ministries are conducive to better fiscal outcomes. Federal fiscal systems that limit vertical fiscal transfers from the central government to subnational governments similarly limit the scope for the latter to spend now and demand additional transfers from the center later. Finally, political systems that produce majority governments or stable coalitions not prone to excessive turnover encourage politicians to adopt reasonably long horizons when making fiscal decisions.

A mechanism for ensuring that debt policies are not distorted by political influences is to rely on fiscal rules that impose limits on unwarranted use of fiscal expansions. The most common fiscal rules are automatic stabilizers and fiscal targets. Automatic stabilizers are taxes and transfers that adjust over the business cycle. Progressive income taxes are a good example: income tax revenues are higher when incomes are higher. Automatic stabilizers have a number of advantages over discretionary tax changes. For one, symmetrical automatic stabilizers do not give rise to a deficit bias. Symmetry implies that the increase in revenues relative to expenditures during expansions is more or less equal to the reduction in revenues relative to expenditures during contractions. In contrast, the temptation to raise spending during bad times may not be matched by the desire or ability to cut it in good times when countercyclical stabilization is undertaken on a discretionary basis. Of course automatic stabilizers can be used only by countries that can access resources during bad times. In order to do this, countries need either to have continuous access to the international capital markets or to accumulate resources in a stabilization fund.

Fiscal targets, including legally mandated balanced budgets and deficit caps, are included in some of the fiscal responsibility laws that have been adopted in many Latin American countries over the past decade, and they figure prominently in Europe’s Stability and Growth Pact. These fiscal policy rules differ in the measure of fiscal performance that they involve, in whether they involve a strict ceiling or simply a target, and in their provisions in case targets are missed or special circumstances arise. The range of performance indicators includes the budget deficit, debt levels, and public spending at various levels of government. Some of the rules allow for margins around the target or for time averaging to provide an opportunity to make up for shortfalls, and many allow for departures in case of international crisis or natural disasters. At the same time, the laws provide for stiff financial or judicial sanctions for noncompliance (see Kopits, 2001).

To be sure, there are also costs associated with such measures. Rules are rigid; such is their nature. Under extreme circumstances, such as an unusually severe recession, a financial crisis or a natural disaster, it may be desirable for stabilization purposes to cut taxes or increase public debt by more than would be appropriate in a typical downturn. Some rules do include “escape clauses” to provide for such contingencies. But this may raise problems of its own. Politicians inclined to use public spending to advance their re-election prospects will be tempted to cite an unanticipated contingency justifying a discretionary increase in spending whenever an election approaches. This problem can be ameliorated by assigning responsibility for declaring the existence of a relevant contingency to an independent, extrapolitical body but, in practice, it is difficult to do this. Nevertheless, rules can be designed to be more responsive to current economic conditions. In this regard, the Chilean rule, described in detail in Box 9.3, is a step forward because it targets a structural measure which adjusts the actual budget balance for the state of the economy and the price of copper, a mineral export that contributes substantially to the country’s fiscal position. Similarly, the Brazilian fiscal responsibility law sets limits on debt accumulation and contingent liabilities with some well-defined escape clauses related to the state of the economy (see Box 9.4).

A further problem is that a major component of debt accumulation is the result of contingent liabilities (or “skeletons”) and balance sheet effects that are not recorded in the traditional measure of the fiscal deficit that is the subject of the pertinent fiscal policy rule (see Chapter 3). One way around this problem may be to set a ceiling on public debt rather than on the government’s deficit. However, contingent liabilities can derive from unfunded obligations of the sovereign (such as unfunded pension obligations), implicit obligations for servicing the debts of subnational governments, and implicit responsibility for the liabilities of public enterprises, banks, etc. This means that it may be difficult for a government to ignore these liabilities even if there is no room to accommodate them if the debt rule is respected. And it may be equally hard to impose a rule on the volume of contingent liabilities. The magnitude of these contingent liabilities tends to be difficult to estimate—this is their nature, since they are contingent, after all. In this regard, enhancing the transparency of fiscal policy and making the budget as comprehensive as possible, which is desirable under all circumstances, can be especially valuable. Special interests pushing for bailouts with narrow benefits but widely dispersed social costs will find it more difficult to do so when fiscal policies are formulated in the light of day and additional expenditures cannot be easily hidden as off-budget activities. Transparency will generally strengthen the operation of market forces, whereby interest rates and credit ratings will more accurately provide an assessment of fiscal sustainability. More intense market discipline will in turn put pressure on authorities to refrain from creating too many skeletons.

Rules-based institutions can also help countries manage volatile revenue flows resulting from commodity exports, either through taxation or by direct ownership of the natural resource. Commodity stabilization funds have in fact been widely used for some time in Latin America and the Caribbean (Engel and Meller, 1993). The idea is to save resources in good times and use them in bad times, which is a sound principle. However, actual experience with stabilization funds has not been entirely happy. Many stabilization funds have been expropriated (in other words, their rules were changed and their assets spent prematurely, and they ended up stabilizing very little). Moreover, even when a stabilization fund is working as envisaged, the government can go on a spending binge involving the central budget during a commodity price bonanza, in effect offsetting the savings accumulated by the stabilization fund. This is a problem not with the concept of stabilization funds per se, but with the design of many stabilization funds that have been implemented in the past, and with the broader fiscal institutional framework within which they operate (see Box 14.1).

MANAGING THE STOCK OF DEBT

Even when countries have good policies in place to control deficits, management of inherited stocks of debt poses several challenges. High economic volatility and low policy credibility are often more serious issues in Latin American and Caribbean economies than in other emerging markets, let alone advanced economies (see IDB, 1995). The volatility of fundamentals (GDP, terms of trade, exchange rates, tax revenues) in the region has been linked to factors ranging from limited diversification of the economy to a narrow tax base. Partly as a result, political processes in Latin America and the Caribbean tend to be less effective and transparent, detracting from the credibility of economic policies and the confidence of resident and international investors.[6] Latin American and Caribbean countries are often especially dependent on foreign borrowing, but volatility and a tendency toward investor panic makes access to foreign markets unreliable. The implication is that maintaining reasonably low public deficits is not enough to eliminate the possibility of a debt crisis.[7] Creating a debt structure that makes public finances less vulnerable to shocks is also essential in the Latin American and Caribbean context.

Gaining credibility requires creating and enhancing the perception that public debt is not a liability to be serviced in good times and restructured in bad times, but rather an obligation that will be serviced under all reasonable circumstances. Budgetary reform that raises the likelihood that a country’s debt will be limited to prudent levels can enhance this perception. So too can the creation of a domestic investor class that holds the government’s debt and is likely to be less prone to herding than international investors. Such an investor group could also become a strong political constituency in favor of responsible fiscal policies and dependable debt service.

Using Contingent Contracts

Even with the strongest willingness to honor debts, the probability that an extreme adverse shock will tip the balance toward renegotiation would still be greater in an emerging economy. This is where well-designed debt management policies can improve the trade-off between the risk of debt crises and the cost of sovereign finance. As highlighted in Chapter 13, the structure of public debt contributes to the burden the debt imposes as importantly as the level of debt itself. In Latin America, in particular, that structure is often biased towards foreign-currency-denominated instruments. In this case, a real exchange rate adjustment has a powerful impact on the most widely used indicator of sustainability, the debt-to-GDP ratio, and in most cases on actual measures of how burdensome debt service is. But a number of other factors, such as commodity prices and other real shocks, or exogenous contagion and panics, can also turn debt sustainability indicators around very quickly (see Chapter 12).

This creates an argument for introducing into debt contracts contingencies with equity-like features that allow for more efficient sharing of this volatility.[8] These would be instruments that offer lower payoffs during bad times and higher payoffs during good times, which should make them safer for investors and would afford governments the opportunity to manage their fiscal policy stance better over the business cycle. Interest payments can be indexed to commodity prices, the terms of trade, or the rate of growth of GDP. While indexing to the price of a commodity has been the traditional recommendation—and still makes the most sense in some cases—emerging economies are diversifying, and a debt contract indexed to the country’s growth rate is likely to be applicable to a broader set of countries nowadays (see Anderson, Gilbert, and Powell, 1989; Borensztein and Mauro, 2004; Caballero and Panageas, 2006; Hausmann and Rigobón, 2003; and Eichengreen and Hausmann, 2005, for discussion of different forms of indexed debt). Under such a contract, when commodity prices drop or growth rates slow, the burden of debt servicing on the government will decline, as investors will share part of that debt-servicing burden with the government.[9] Chapter 13 illustrated how making use of such provisions can reduce the volatility of a country’s debt-to-GDP ratio and effectively reduce the probability of a debt crisis. The opportunities for sovereigns to make use of a broader set of debt instruments have increased significantly in recent years, as noted above.

Another option is to obtain contingent coverage directly from international financial markets, through the use of derivative contracts. An example would be the case of a commodity producer subject to fiscal shocks due to fluctuating commodity prices. Such a country can reduce uncertainty by using futures, forwards, and options markets for the commodity. In practice, however, there are problems with this approach. First, many futures and options markets lack depth and liquidity and therefore offer only limited scope for insurance. The lack of markets is more acute in respect to events such as fluctuations in tourism revenue, hurricanes, and other natural disasters. Fortunately, financial market innovation is increasing the scope for using this type of market coverage as insurance, as in the case of the recent operation by Mexico securing earthquake insurance for three at-risk geographical areas (see Box 14.2). Second, contracts aimed at isolating countries from external shocks are likely to be very large and complicated and may present significant demands in terms of management, and it may be difficult to allow traders sufficient leeway to operate in the markets while ensuring that their trades and risk taking are aligned with the objectives of the government.

Finally, obtaining some form of market insurance, either through derivative contracts or through indexed debt, must also surmount a more fundamental obstacle. By its very nature, any such device implies a cost that must be paid during good times. This is analogous to paying an insurance premium and takes the form of losses in a futures or option contract or high coupon payments on debt. As these contracts are relatively complex, such losses can be easily misunderstood and become politically costly. This creates little incentive for politicians to enter into large-scale contracts of this type, especially for myopic politicians, considering that the cost is likely to be paid up front but the payoff from the insurance may accrue only years later.

Currency-Maturity Trade-Offs

Another trade-off that often arises in the context of dedollarization of sovereign debt (that is, the shift from foreign- to local-currency-denominated instruments) is between currency and rollover risk. Debt denominated in local currency is often placed at short tenors, largely because of steep currency premiums—the result of lingering fears of inflation, which has long been a concern in Latin America—that make long-term local currency borrowing excessively costly (see Chapter 13). Furthermore, when credibility can be regained only gradually, governments should avoid locking in high risk premiums in long-term bonds. If the menu is limited to long-term, fixed rate foreign currency debt and short-term domestic currency debt, it makes sense for the issuer to maintain a diversified portfolio. Fixed rate foreign currency debt insulates the issuer from sharp fluctuations in interest rates on local currency instruments, while short-term domestic liabilities protect the issuer against a sharp increase in the debt burden when the domestic currency depreciates. Inflation-indexed instruments provide an alternative that can help improve the terms of this trade-off. It may be possible to issue long-term inflation-indexed instruments at moderate cost, as investors are protected from the risk of unexpected inflation. But governments may have been wearied by past experiences in which financial indexation spearheaded widespread indexation of wages, pensions, subsidies, and so on and created a situation of stubborn inflation and inflexibility of relative prices. Still, some countries have been successful in using indexed financial instruments widely without perceptibly worsening inflation persistence.

Since recent experience has pointed to currency fluctuations as an important source of vulnerability, Latin American governments have reacted by favoring local currency debt over dollar debt.[10] But trading one risk for another (in this case, currency risk for rollover risk) is not a panacea. If the next shock to the region’s economy is a rise in local currency funding costs rather than a fall in exchange rates, concentrating exposure on the maturity side may prove not to have been a prudent bet. A large investor base for debt that is denominated in the domestic currency at fixed nominal rates and reasonably long maturities does not yet exist. Interestingly, it would appear that foreign investors are more interested in these types of instruments, as such investors are less troubled by a history of inflation (see Chapter 7). But at the same time, these investors may be highly sensitive to changes in credit quality or less favorable prospects in short-term returns, which means that market access may be unreliable for Latin American sovereigns.

This suggests that eliminating (or decreasing) the existing trade-off between currency and maturity will also require the development of liquid, well-functioning bond markets for domestic currency instruments that are underpinned by a stable investor base. Domestic institutional investors, such as pension funds, are increasingly forming the core of such an investor base in many countries. By the nature of their liabilities to beneficiaries—and also as a result of direct regulation—pension funds are naturally stable, dedicated investors in domestic bond markets. As policies and institutions in the countries of the region gain credibility and inflation fears continue to recede, the core investor base will grow broader. A better debt structure will, in fact, make the policy framework in these countries sounder and itself contribute to gains in credibility on price and exchange rate stability. Thus, the strategy for gaining access to long-term, fixed rate, high-credit-quality, domestic-currency-denominated debt should be based on these two elements: improving credibility through sound policies and developing local bond markets.

Managing Rollover Risk

Even when countries are in a sound position in terms of debt sustainability, they may face liquidity problems. Countries need to roll over maturing debt and cover their annual financing needs, and this can become virtually impossible in the event of a sudden stop in global financial markets. Moreover, a liquidity crisis can trigger more fundamental insolvency problems by causing a large exchange rate depreciation, a recession, and/or bank failures. When debt is denominated in foreign currency, only the accumulation of a large stock of international reserves can protect a country from potential liquidity crises. In recent years, emerging economies, especially those in Asia, have accumulated vast international reserves (Figure 14.2). The accumulation of reserves in Latin America has been one of the least extensive in all of the regions, but still significant. While in some cases, notably Middle Eastern oil exporters, the level of reserves goes well beyond what may seem necessary from a financial stability standpoint, for many emerging economies the main purpose of accumulating international reserves is crisis prevention.

But the accumulation of international reserves is expensive. Reserves are held in safe liquid assets so that they can be mobilized when there is a need to intervene in the foreign exchange market, either to avoid wide fluctuations in the exchange rate under disorderly market conditions or simply to smooth out the effect of temporary shocks. But safe and liquid assets such as U.S. Treasury bonds carry low interest rates. For emerging markets, the spread of their own debt over the yield on U.S. treasury bonds can be significant. Self-insurance thus entails a “cost of carry” that the government has to pay in excess of the return on liquid foreign assets to finance the purchase of excess reserves, namely, the sovereign risk premium, which for most Latin American countries—as opposed to Asian countries—tends to be large.[11]

Self-insurance is, by its own nature, an inefficient strategy. Any car driver would recognize the efficiency of buying a car insurance policy instead of saving and stashing away millions of dollars for possible liability claims before owning a car. Essentially along these lines, some authors have proposed ways of improving on this self-insurance strategy, for example, by investing reserves in assets that are negatively correlated with country risk (as opposed to high-grade foreign currency assets) or, in the case of commodity exporters, through the use of derivatives (Caballero and Panageas, 2005, 2006; Rigobón, 2006). This is a sound strategy assuming that there exist assets or commodity derivatives with a reliable correlation with country risk and a sufficiently liquid market. A number of obstacles may have to be overcome to implement such a strategy, however, including the already noted political cost of paying the insurance fee (in this case, the derivative losses when the economy is in a good state). But the case for this strategy is easier to make in countries where the volume of reserves amply exceeds what may be needed purely for the purpose of ensuring stability in the foreign exchange market.[12]

Self-insurance through the accumulation of reserves has a further drawback. Any readily available pool of public resources is subject to political capture. In other words, reserves could be spent before the “rainy day.” This has been highlighted in the literature on stabilization funds but applies more generally to any type of public savings.

An alternative strategy would be to obtain liquidity insurance from the private financial markets, for example, in the form of credit lines that can be activated if there is an incipient sudden stop, as measured by an increase in spreads or some other variable. In fact, such credit lines were implemented for a handful of countries, including Mexico and Argentina (see IMF, 1998). Private liquidity insurance, however, faces some serious challenges. In particular, private lenders want to reduce their exposure when conditions deteriorate and can effectively undo in other markets the loans that are activated through the liquidity insurance agreement.

REFORMING THE INTERNATIONAL FINANCIAL ARCHITECTURE

The financial crises of the 1990s resulted, at least in part, from market imperfections that led to “herd” behavior by investors, contagion, and panics resulting in self-fulfilling liquidity runs (Calvo, 2005b). This suggests a role for the international community, and the international financial institutions (IFIs) in particular, in implementing initiatives to limit the consequences of instability in international financial markets. Efforts in this direction gained momentum after the Asian and Russian crises of 1997–1998 and continue to evolve as the global financial system continues to pose challenges.

The traditional role of the IFIs has been to provide financial and policy support when a country facing a currency or financial crisis requests it. But it is clear that the best way of minimizing the costs of crises is to avoid them in the first place. While this depends on prudent fiscal policies on the part of the countries themselves, the IFIs’ role is to help minimize the risks that arise from global financial markets, mainly rollover and contagion risks. In addition, the IFIs have been working to eliminate major obstacles that emerge in connection with the resolution of debt crisis events.

Rollover Risk

To strengthen crisis prevention, the international community needs to implement plans to prevent or mitigate sudden hard currency liquidity shortages. Credit facilities to prevent liquidity runs that can turn into debt crises are referred to as country insurance facilities.[13] A country insurance facility would consist of a liquidity window that lends in the short term to eligible countries at predetermined interest rates—in much the same way as the central bank, acting as lender of last resort, lends to domestic financial institutions. Since rollover risk (that is, uncertainty about access to sources of finance) is the main aspect driving liquidity runs, the availability of liquidity with certainty is a strong deterrent to the start of a self-fulfilling run.

Although facilities of this type have been recognized to be the best response to liquidity and contagion risks,[14] there are also some implementation difficulties that must be worked out. A commonly voiced concern is the potential for moral hazard. This concern relates to the possibility that a government could adopt risky policies with high short-term political rewards once it has secured a country insurance line. Indeed, moral hazard is an issue that comes up in relation to any insurance contract. As with private insurance, there are also mechanisms for dealing with moral hazard issues in the case of countries. The facility can avoid moral hazard by applying appropriate eligibility conditions, based on triggers that are exogenous to the assisted country (such as international interest rates or natural disasters), or a policy prequalification condition. The latter condition could determine the volume of resources to which a country has access based on consistent and transparent indicators of the soundness of the country’s policies. An alternative mechanism would be the requirement of a commitment by countries not to borrow in international markets at above a predetermined spread, which would put an early stop to risky “borrow and spend” runs (Cohen and Portes, 2006). Eligibility rules may also face problems related to the governance of institutions managing the credit facility (Powell and Arozamena, 2003). For example, a declaration by the institution which is providing insurance that a country has become ineligible (perhaps because of a deterioration of fundamentals) may lead to a market run and precipitate a crisis.

As a partial response to the liquidity risk, some emerging economies have started to develop regional country insurance schemes. These typically take the form of regional swap agreements under which participating countries can borrow from other members on short notice for limited periods of time. These agreements include the North American Swap Agreement (NAWA), the Chiang Mai Initiative (CMI), and the Latin American Reserve Fund (FLAR, after its Spanish name, Fondo Latinoamericano de Reservas). NAWA, set up in April 1994 among Canada, Mexico, and the United States, provides 90‑day renewable collateralized loans. CMI, launched in May 2000 by the 10 members of the Association of Southeast Asian Nations (ASEAN) plus China, Japan, and Korea, involves bilateral currency swap arrangements.[15] FLAR originated in 1978 as the Andean Reserve Fund (FAR) and was expanded to include all interested Latin American countries (with the name changing to FLAR) in 1988. Currently, FLAR includes six countries (Bolivia, Colombia, Costa Rica, Ecuador, Peru, and Venezuela) and has a size (as measured by the capital subscribed by its members) of $2.1 billion.

While these arrangements are close in spirit to a multilateral country insurance facility, in that they offer immediate access to short-term liquidity while avoiding the hedging problems that may arise with private insurers, their effectiveness is hampered by their limited (albeit growing) size and, in the Latin American case, by the absence of a large country with reliable access to dollar liquidity in the arrangement—a factor that severely reduces the scope to leverage central bank resources without raising borrowing costs and, as a result, the size of available credit lines.

These problems notwithstanding, there seems to be scope for regional insurance in Latin America, potentially leveraging the liquidity support that could be provided by a country insurance facility. FLAR has been able to leverage its capital to some extent by funding itself in international markets at interest rates below those of the participating countries (Figure 14.3). This implies that this regional arrangement entails a lower insurance cost relative to what member countries would have to pay individually. A similar effect is observed in the rates paid by the Andean Financial Corporation (CAF, for Corporación Andina de Fomento), a regional development bank, which suggests that there may be efficiency gains associated with this type of arrangement, stemming from risk pooling or from the perception that these facilities enjoy a preferred creditor status.[16]

Absent large creditor countries, the IFIs could have a potentially important supporting role in such arrangements by guaranteeing, under certain conditions, the debt placed by the multilateral insurance fund or even contributing resources to the pool (Cordella and Levy Yeyati, 2006b).[17] This could significantly leverage the size of the fund at a reasonably low cost, while keeping the intended countercyclical pattern of IFI lending. In other words, the IFIs would shift from making loans to offering guarantees in good times when alternative sources of finance are abundant, and back to making loans in recessions when funds become scarce (given that the insurance fund has already been built up). Alternatively, the IFIs could go further to provide a global arrangement along the lines of the existing multilateral insurance schemes, which would be superior to a group of regional ones, because liquidity shocks tend to be regionally correlated, because neighboring economies are subject to similar risks, because neighboring economies trade with one another, and because contagion tends to have strong regional links.

Risk of Contagion

Contagion has escalated individual emerging market crises to regional or even global events in several past episodes (Chapter 5). While a well-implemented and fully credible country insurance mechanism could eliminate contagion episodes, regulators and supervisors could also play a role by putting in place mechanisms to limit the damage caused by disorderly markets. In many domestic exchanges, rules such as circuit breakers that suspend trading temporarily when price fluctuations become too large have been implemented to help prevent market failures from developing into full-fledged crises. There is no equivalent mechanism in the global market for sovereign debt.

Although it may not be feasible for an international institution to act as a global regulator, there are proposals that could provide circuit-breaker-type benefits to limit contagion effects. For example, Calvo (2005b) proposes the creation of an Emerging Market Fund (EMF) aimed at stabilizing an emerging market index, such as the JPMorgan Emerging Markets Bond Index Plus (EMBI+). The fund would be endowed with G3 debt instruments and, in the event of a disturbance, could limit contagion by making a credible commitment to buy bonds from the emerging markets that are not at the center of the crisis. The EMF could thus slow down or even stop a generalized collapse in the asset class, preventing fire sales from sending the wrong signal to investors. According to Calvo, the fund would not try to fight trends but only intervene in special circumstances. Action could be triggered only by a financial meltdown, defined as a drop in the index of more than a certain percentage relative to a moving average. If the initial drop reflected a change in fundamentals and prices did not recover from the initial drop, the moving average would decline over time, and the EMF would sell its emerging market bonds and revert to holding only G3 bonds. The same thing would happen if the intervention was successful and prices recovered to the precrisis level. In both cases, the EMF would have negligible holdings of emerging market bonds in tranquil times. Calvo shows that creating such a fund would require less than 1 percent of the public debt of G3 countries and could even be profitable, as long as the majority of the crises were indeed due to contagion and not deterioration of fundamentals.[18]

Crisis Resolution

Even in the best-planned system, train wrecks sometimes happen. When sovereign debt crises occur, there are no well-established procedures for restructuring debts and restoring financial normalcy.[19] Restructuring typically occurs through a bond exchange offer in which new bonds are exchanged for the existing debt. Because exchanges are voluntary, there is always a fraction of bondholders that do not accept the offer. The value of the claims of these holdout investors creates legal uncertainty and litigation. Consequently there has been considerable debate over proposals to establish a statutory mechanism to adjudicate defaulted claims, such as the Sovereign Debt Restructuring Mechanism (SDRM) proposed by the IMF.

While there has been disagreement in the international community regarding the desirability of an SDRM, there is consensus on the desirability of a more modest initiative to include collective action clauses (CACs) in bond covenants (see Eichengreen and Portes, 1995). CACs provide for changes in the payment terms of a bond if a supermajority of bondholders—usually 75 percent—accept the changes. This automatically solves holdout problems by binding in dissidents. CACs have become commonplace in emerging markets’ global bond issues since Mexico pioneered them in 2003. Recent CACs have also included “aggregation” clauses that permit a supermajority of bondholders—typically 85 percent—to restructure all outstanding bonds and binding the minority to accept the write-down. This solves the problem that the simple version of a CAC applies only bond by bond, and many sovereigns have issued tens, or even hundreds, of bonds.

While CACs have become standard in new bond contracts, older bonds still in circulation do not include them. In most cases, it will take many years for the stock of outstanding bonds to mature and be replaced by new instruments containing CACs and aggregation clauses. This means that holdout uncertainty will not disappear quickly. The lack of collective action created by holdouts has not been a major impediment to recent sovereign restructuring operations, although a large mass of holdout claims remain unresolved, and the outcome of ongoing litigation and possible new legal strategies by holdouts may change the situation again (see Sturzenegger and Zettelmeyer, 2005b).

New Financial Instruments

The international community can help to improve debt management by supporting the development of new markets and new instruments to allow countries to minimize the risks of sovereign borrowing, keep the costs of borrowing at moderate levels, and improve the cyclical timing of fiscal policy.

Developing Local Currency Bond Markets

There is broad consensus that it is desirable for the universe of debt instruments issued by a government to include a significant share of domestic currency debt. As noted above, Latin American countries have started to change the structure of their debt in response to this recognition.

The IFIs could accelerate this process by helping to increase the scope of available local currency instruments at home and abroad. One option is for them to enlist their own market liabilities. Already the multilaterals have begun issuing bonds denominated in emerging economies’ currencies, although often the objective has been mainly to minimize their own borrowing costs. (Box 14.3 comments on the experience of the IDB in this regard.)

By borrowing in local currencies, the IFIs could also support the development of markets for such instruments. One of the main factors limiting a country’s ability to issue external debt in its own currency is the small size of the market. While the largest Latin American economies, Brazil and Mexico, for example, may not be seriously affected, the currencies of many emerging markets are considered “exotic” and carry substantial liquidity premiums (Eichengreen, Hausmann, and Panizza, 2005a). An ambitious proposal along these lines is to create a synthetic unit of account that pools currency risk from a large and diversified group of emerging economies, and to have the international financial community take steps to develop liquidity in this unit (see Eichengreen and Hausmann, 2005).[20]

Some observers have noted the influence of sovereign (credit) risk in the underdevelopment of local currency markets and have pointed to the IFIs’ bonds in exotic currencies as a way to decouple sovereign risk from currency risk. This is implicit in the previous proposal, where the IFIs are seen as the first issuers of bonds in a basket of currencies. But it is even more critical for resident (particularly institutional) investors, who are more naturally inclined to invest in their home currencies but may shy away from domestic assets for fear of default. Absent an international market in their domestic currencies, the offshoring that characterizes many non-investment-grade Latin American countries may lead to the dollarization of domestic savings for reasons unrelated to currency risk. It follows that IFI bonds in domestic currencies may find their main investor base among residents.

Contingent Bonds

Debt sustainability and risk sharing can be enhanced by instruments with equity-like features, which provide for lower payments in the event of adverse shocks like natural disasters, recessions, and commodity price busts, but these markets are grossly underdeveloped. To be sure, creating a market in such securities poses a number of challenges. New, original instruments may have shallow markets initially and command an illiquidity premium. Designing a new type of instrument is costly, creating a first-mover problem.

Markets in such instruments do not spring up spontaneously. Someone has to sink the costs of designing the new instrument, and someone has to be the first to issue in a nonexistent or illiquid market. In the past, official intervention has been instrumental in the development of pathbreaking financial instruments, such as the market for mortgage-backed securities in the United States. In the case of contingent bonds, the international community can provide technical assistance on instrument design and expected pricing. In the case of GDP-linked debt, for example, the international community could strengthen the quality and reliability of statistics by various means, enhancing their credibility for investors. As in the case of collective action clauses, the international community could help in the drafting of a model contract and resolve legal uncertainties (for example, questions about the legal standing of a GDP warrant relative to other sovereign instruments). It could provide guidance on the drafting of GDP link clauses to ensure the reliability and integrity of their application.[21]

A more ambitious idea would be for some of the international financial institutions to become the first issuers of an instrument of this type and sow the seeds of a market that countries themselves could tap later. The risk could be unloaded by swapping this instrument with the beneficiary country, although it can be argued that the international financial institutions already carry an equity-type risk in regard to their member countries, because the institutions will need to help these countries if the countries suffer an adverse shock. Alternatively, an international financial institution could issue a bond on an index of real variables of several countries, in a form analogous to the currency basket referred to above. In the same vein, the IFIs could guarantee contingent instruments, or at least the part of the instruments that is contingent. This might be seen as a subsidy to spur innovation in the market by compensating for novelty premiums, setup costs, and concerns about manipulation of certain indices (see Anderson, Gilbert, and Powell, 1989).

FINAL REMARKS

While it is trivially true that pushing the region’s debt to zero would eliminate Latin America’s vulnerability to debt crisis, this is neither feasible in the short run nor economically desirable. The central conclusion of this report is that, more than the level, it is the structure, namely the quality, of the debt issued by Latin American and Caribbean countries, and the inherent volatility of their economies, that makes the region prone to crises. While the specific design and parameters of a debt management strategy for the region’s countries would differ from case to case, some general principles are valid. Latin American and Caribbean countries should continue to shift their debt structures away from foreign-currency-denominated debt and into debt denominated in domestic currency. However, there are trade-offs that need to be considered carefully as they advance in this process. In particular, to avoid locking in excessive interest costs, countries sometimes need to issue instruments with very short maturities. Otherwise, vulnerability to a debt crisis—or to an inflationary outburst—will not disappear but only change its nature. The development of sound domestic bond markets, based on a core set of institutional investors and the use of inflation-linked instruments, can help improve the terms of this critical trade-off. Foreign currency debt will maintain a share in each country’s liabilities both because of the need to tap foreign investors and because the structure of revenues in the country may be partly related to foreign currency and thus make it advisable from a risk management perspective. Countries should explore more aggressively the use of contingent debt as a mechanism for obtaining insurance from foreign investors against adverse shocks such as recessions, commodity price collapses, and natural disasters.

Debt management is crucial in the volatile Latin American environment, but limiting the risks of sovereign finance also demands gaining the markets’ (and their citizens’) trust in the institutional and policymaking framework. In particular, the flow of new debt, namely, budget deficits, must be controlled to ensure both that the ability to borrow is not abused by the political leadership and that fiscal policy does not worsen economic fluctuations. Although design problems have impaired some experiences, fiscal rules and stabilization funds continue to be ideal mechanisms for underpinning a sound approach to fiscal deficit controls.

The current relatively benign global environment is partly due to better policies and safer debt management, but it heightens the risk that the international community will become complacent and needed initiatives will be postponed. Tranquil times are the best for discussing and introducing new initiatives aimed at reducing the vulnerabilities that still lurk in the global financial system.

In recent years, the international community has focused on the process of resolution of debt defaults, and progress is being made in this area, with the widespread introduction of collective action clauses in debt contracts. But progress has not been made in the area of crisis prevention, and available instruments were designed in a pre-financial-globalization era. In this area, the IFIs could contribute a great deal by designing workable credit facilities to prevent liquidity runs and self-fulfilling market panics and by supporting in various ways reserve-pooling arrangements by emerging market economies. The IFIs also have an important new role to play as facilitators of reforms aimed at limiting the risk of sovereign finance. The IFIs can promote the development of markets for local currency instruments and new contingent debt instruments in various ways. They can provide assistance with the design of those instruments, and they can help to overcome the externalities and start-up costs of new markets and attract new investors. Finally, the IFIs can change the nature of their own loans to member countries by offering a wide menu of domestic currency loans and contingent facilities and thus contribute to the dedollarization process.