VI Debt Management
- International Monetary Fund
- Published Date:
- September 1995
After a period of consolidation in the late 1980s, many developing countries have witnessed a significant rise in external indebtedness during the 1990s. In the aftermath of the Mexican financial crisis and amid renewed concerns about the volatility of capital flows, external debt-management practices have received greater attention. Events leading up to the recent crisis in Mexico and during its aftermath illustrate the importance of the Government’s policies for managing external debt, as well as the need for sound macroeconomic policies. The sharp appreciation of the yen in early 1995 also brings debt-management issues into focus.
At the same time, a large part of the rise in the external indebtedness of developing countries during the 1990s has been associated with the activities of the private sector and public sector corporations. The more active role of these entities in tapping international capital markets for funding raises a number of additional issues for external debt management. Although private sector entities should be held accountable for their own debts, difficulties encountered by the government can negatively affect the ability of the private sector to access international financial markets or to service its debts, adding to the severity of a financial crisis. Alternatively, problems in the private sector can quickly pose significant challenges for macroeconomic policies and impose substantial fiscal costs. This may be the case especially if the banking sector has borrowed heavily abroad to fund its domestic lending, and important segments of the sector encounter difficulties in meeting their obligations. Failure of major nonfinancial corporations may also create important policy challenges owing to the adverse impact of such failures on the domestic economy and on the ability of the government and other domestic borrowers to access international financial markets. The potential negative externalities associated with private and public sector behavior require that the government pay close attention to the activities of nongovernmental entities with regard to external borrowing.
Management of Public Sector Debt
As noted by Fischer (1983), there is no completely satisfactory theory of government debt management; nonetheless, a review of the literature points to some key factors to be considered with regard to the maturity profile of the debt, debt indexation, and the currency denomination of the debt.
In an early systematic treatment of the choice of maturity structure for public debt, Tobin (1963) cites three main objectives and related strategies with respect to the choice of debt maturities: (1) minimization of interest costs; (2) economic stabilization (or support for fiscal and monetary policy in influencing the level and composition of aggregate expenditures); and (3) neutrality, meaning the minimization of effects of public debt financing on financial markets.
There are trade-offs among these objectives that governments have to take into account. For example, cost minimization might call for short maturities. However, the need to roll over public debt very frequently may have disruptive effects on financial markets, since uncertainty may be created about the government’s ability to refinance its debt, especially during periods of market turbulence (Giavazzi and Pagano (1990)). Such a policy may in the end raise costs, as market agents demand increasing interest premiums to refinance the debt. In contrast, a debt structure that produces a smooth repayment stream spread over a longer period may help to achieve neutrality of public debt with respect to financial markets, but it also will carry a higher interest cost.
Under a situation of certainty, a government would choose a debt maturity structure designed to meet the three criteria and to match the public debt-service profile with prospective revenue and financing. However, countries are subject to unexpected shocks that make debt management more complicated. Typically, public debt is used to smooth the impact of shocks to national income and public revenue. Although revenue is usually procyclical, spending on debt service could be countercyclical, and the government may be forced to borrow from the market at times when its financial situation may be perceived as weak.
In making decisions about the maturity structure of debt, the government has to be aware that such decisions will affect market perceptions about its solvency. In the context of a formal model of an open economy with free capital mobility, where the central bank follows a fixed exchange rate regime, the chance that the government will be able to withstand a confidence crisis is affected by the extent to which it has to appeal to the market at each given date to roll over its debt. This, in turn, depends on the amount of debt outstanding, its average maturity, and the time pattern of maturing debt (Giavazzi and Pagano (1990)). A confidence crisis can be avoided if the amount of debt outstanding is not high and its maturities are not concentrated at a few dates. Of course, such a debt profile is a product of the past and current stance of macroeconomic policies; in fact, it may be impossible for a government to achieve a desired distribution of maturities if other policies are not in place.
Unsustainable fiscal policies may lead to high levels of debt, and although governments may clearly understand the need to have a debt structure weighted toward longer-term maturities, the government, owing to the stance of policies, may not be able to borrow long term. Empirical evidence shows that at low and moderate levels of government debt there appears to be little relation between the level of debt and its maturity; however, at high levels of debt a strong inverse relationship emerges (Missale and Blanchard (1994)).
The size and maturity structure of public debt (as well as past macroeconomic policies) will affect the ability of a government to issue debt in local currency. If debtholders are required to frequently roll over a large or rising stock of public debt, they could start to fear debt monetization in the case of local currency debt or outright default in the case of foreign currency debt. To the extent that the government is not willing or is unable to raise revenue or cut expenditure, it might eventually be forced to monetize its debt owing to the self-fulfilling expectations of the private sector (Calvo (1988)).
Alternatively, the government could diminish the fear of debt monetization and maintain access to financing on reasonable terms by indexing the debt. However, the decision to substitute indexed for nominal debt carries an additional cost. Although nominal government liabilities create the incentive for inflation, nominal debt can help the government to hedge against shocks that affect its budget. Shocks to the economy generally require unexpected changes in the path of taxes; if debt is nominal, these shocks also change prices and the real value of government debt. The price effect may thus reduce the required changes in the path of taxes. Unless taxes are lump sum, these differences in taxation have important welfare implications (Bohn (1988)). Consequently, resorting to debt indexation (or an equivalent change in the currency denomination of debt) eliminates the incentive to monetize, but it may require significant changes in the path of taxes—thereby creating incentives to default, which may over time be reflected in the premium paid on indexed debt.
Foreign Currency Debt
Foreign-currency-denominated debt typically enables the borrowing government to issue securities with longer maturities than those sold in domestic markets, thereby avoiding a concentration of redemptions. Foreign markets may be more liquid, also potentially allowing for a smoother rollover of outstanding debt obligations. Moreover, issuing debt in foreign markets may provide the international diversification suggested by portfolio theory to reduce both the borrower’s costs and the volatility of the total debt (de Fontenay, Milesi-Ferretti, and Pill (1995)). Under certain circumstances, the credibility of a country’s anti-inflationary policy may be improved by issuing foreign-currency-denominated debt, which could lower the interest rate premium associated with the risk of monetization.
A case also can be made for foreign currency debt based on a hedging motive in an open economy subject to shocks to output (Bohn (1990)). If unexpected foreign inflation and domestic output are negatively correlated, foreign currency debt can serve to hedge the domestic economy against shocks to output. In this context, foreign-currency-denominated debt will protect the economy against worldwide business cycles originating on the supply side, by lowering the real value of debt when the domestic economy is weakening.50
By issuing foreign-currency-denominated debt, however, developing countries may accentuate an already large exposure to foreign exchange risks inherent in their export and import patterns. This additional exposure could occur because debt-service payments would depend on the real value of the currency in which the debt is denominated at the time that the debt matures. The high historical volatility of currencies and export and import prices has had serious implications for the government budgets of developing countries. Countries can be particularly vulnerable (1) if they have large international borrowing requirements and the resulting external debt is denominated in different currencies; (2) if most of the external debt is in obligations with variable interest rates; and (3) if trade in primary commodities is significant (Claessens (1992)).
The currency composition of external debt can be used to minimize exposure to external price risk and, thus, to diminish the exposure of developing countries and their budgets to external price shocks (including exchange rates and interest rates). An optimal currency composition of the debt can be determined by doing an empirical investigation of relationships between cross-country exchange rates and indicators of the balance of payments (such as the terms of trade, exports, and services receipts). For a country that acts as a price taker in world markets (goods and financial) to hedge itself against commodity price, interest rate, and exchange rate movements, the optimal currency composition of its debt would depend only on the co-variances between the effective costs of borrowings and indicators of the country’s ability to generate foreign exchange receipts. However, these covariances may vary over time. Moreover, the currency composition of debt is an imperfect hedging tool against external price uncertainty, particularly because transaction costs can reduce the benefits of adjusting portfolios frequently. Other types of instruments, such as contingent contracts linked to commodity prices, may be better suited as hedging devices against volatile external prices (Claessens (1992)).
In assessing the cost of borrowing in foreign currencies, careful analysis of the associated foreign exchange risk is critical. This risk needs to be adequately hedged—either implicitly by keeping the composition of foreign currency borrowing in line with the composition of the country’s foreign exchange earnings, or explicitly by the use of financial market instruments such as foreign exchange swaps. An unhedged foreign currency position can expose a country to considerable risks, and it may encounter substantial costs as a result. The recent experiences of countries that have sizable obligations denominated in Japanese yen illustrate this point. As Table 15 shows, several developing countries increased their yen exposure significantly in the period 1989-94, Some of these countries sustained significant increases in their debt-servicing burdens owing to the sharp appreciation of the yen in early 1995.
|U.S. dollar||Yen||Deutsche mark||U.S. dollar||Yen||Deutsche mark||U.S. dollar||Yen||Deutsche mark|
Debt Management and Macroeconomic Policies
In the end, the ability of a country to manage its public debt effectively will depend on the stance of macroeconomic policies. The stronger is the country’s macroeconomic position, the easier it is to pursue appropriate debt-management practices. In sharp contrast, an unsustainable policy stance (accompanied by a reluctance on the part of the authorities to make necessary policy adjustments) may lead a country to engage in debt-management practices that may sow the seeds for the emergence of a financial crisis. At the same time, the government’s options in managing its debt are likely to become increasingly constrained by its policy stance.
In an effort to limit the impact of borrowing costs on the budget, the country may significantly shorten the maturity profile of its debt. As maturities are shortened, interest rates (and borrowing costs) will tend to rise as concerns about the possible monetization of the debt become more prominent. In response, the government may choose to index its debt to alleviate such concerns or may resort to foreign currency borrowing, thereby at least temporarily containing costs and possibly limiting pressures on domestic interest rates. In borrowing foreign currencies, the quest to keep down borrowing costs may lead the country to tap a wide variety of currency markets, potentially increasing the riskiness of the country’s foreign currency exposure. Eventually, pressures on domestic interest rates would reappear in the absence of policy actions because concerns about the government’s debt-servicing ability would bid up interest rates on indexed debts. Likewise, concerns about the ability of the country to meet its foreign currency obligations and expectations regarding the exchange rate would raise the cost of foreign borrowing. At some point, a crisis in confidence is likely to emerge, as the government faces the prospect of being unable to roll over its maturing debt.
The recent crisis in Mexico illustrates how developments can affect the structure of government debt and, in the process, lay the foundation for a financial crisis. During 1994, as investors became increasingly reluctant to hold peso-denominated government debt, the Mexican Government issued increasing amounts of U.S. dollar-indexed short-term debt (tesobonos) (Table 16). As a consequence, the maturity profile of the Government’s debt shifted toward the short term, resulting in a significant bunching of future maturities, and the proportion of foreign-currency-linked obligations in total debt rose. This shift initially produced lower borrowing costs for the Government and reduced pressure on domestic interest rates, but toward the latter part of 1994, upward pressure on rates (including tesobono rates) emerged. With a substantial loss in international reserves prompting a devaluation and subsequent floating of the Mexican peso in December 1994, the maturity structure and currency composition of the public debt added to the severity of the ensuing crisis.
Implications of Private External Borrowing for Macroeconomic Policy Management
Implications of Banking Sector Intermediation of Capital Flows
One of the main functions of financial intermediation by banks is to transform assets and, thereby, facilitate the allocation of financial savings to investment. In carrying out this function, banks provide deposit holders the opportunity to indirectly have a more diversified portfolio, and thus lower risk, while at the same time providing the service of asset evaluation, since depositors delegate to banks the task of monitoring the loans they make (Diamond (1984)).
In carrying out this intermediation, banks face three main types of risks: interest rate risk, exchange rate risk, and asset quality risk. Interest rate risk arises from a potential mismatching of the maturities of banks’ assets and liabilities (Jaffee (1986)). Exchange rate risk stems directly from the banks’ net asset or liability positions maintained in foreign exchange, but it may indirectly result from the banks’ domestic operations that are denominated in foreign currency. Asset quality risk is related to the success of the activities financed by the banks’ loans and other investments. The major effects of capital inflows on the banking systems of developing countries have been through the influence of capital inflows on exchange rate and asset quality risk.51
With access to international capital markets, profitable lending opportunities at home could induce banks to take open foreign exchange positions, which would increase their exposure to foreign exchange risks. Even if regulations place limits on such open positions, banks still could be taking an implicit foreign exchange risk associated with a shift in asset quality arising as a result of banks borrowing abroad to lend domestically in foreign exchange to nonfinancial entities that do not have foreign exchange receipts.52
The importance of the latter point is illustrated by the recent experiences of developing countries. Most episodes of significant capital inflows to developing countries in the 1990s, particularly in Latin America, came after the implementation of stabilization policies and followed the reduction in interest rates in industrial countries (Calvo, Leiderman, and Reinhart (1993)). In the face of strong credit demand, especially from the nontradables sector, banks had incentives to lend domestically both U.S. dollar deposits from domestic residents and foreign exchange raised abroad. In the case of Mexico, lending in dollars to the construction sector showed a significant increase between 1992 and 1994 (Table 17). Dollar credits to the services sector in Argentina also rose significantly during the same period (Table 18). In Argentina’s case, these dollar credits were mostly extended by public banks, cooperative banks, and financieras. Private banks, particularly those that are foreign owned, appear to have taken a more cautious approach regarding dollar loans to the nontradables sector (Table 19).
|Agriculture, forestry, and fishing||10.2||13.4||23.0||13.8||13.5|
|Services and other||24.2||23.2||24.2||23.5||24.0|
|Water, electricity, sanitation||8.9||3.2||4.0|
|Water, electricity, sanitation||2.3||5.4||6.0|
These credit booms were generally accompanied by an investment boom in the nontradables sector (mostly construction but also other services) and a consumption boom. The stabilizing economies also typically experienced a real exchange rate appreciation, as the prices of nontraded goods responded to increased demand while supply tended to be constrained, and as returning flight capital and other external borrowing and investment flooded these economies with foreign exchange. Consequently, new foreign currency loans made by banks generally had assets valued at inflated prices in foreign currency terms as collateral, and loan recipients often had inflated foreign exchange earnings.53
This situation gives rise to two sources of concern for the authorities: there may be a tendency to over-borrow by domestic firms and households, and the existence of explicit or implicit insurance on bank deposits provides incentives to financial intermediaries to take on significant risks (McKinnon and Pill (1993)). At the same time, that a significant part of bank lending was directed to the nontradables sector, which may be prone to suffer more from real domestic shocks and exchange rate shocks, raises questions about whether the interest rates charged by banks on these loans realistically reflected the risks involved. Although domestic interest rates generally are higher than international rates, owing to “country risk,” they may not adequately reflect sector-specific risks.
To guard against and minimize potential dangers and to maximize benefits from the intermediation of capital flows through the banking system, a country needs an adequate supervisory and regulatory framework. As part of this framework, prudential regulations limiting open foreign exchange positions are useful, but supervisors will have to carefully check the lending and investment practices of the banks to ensure that undue implicit foreign exchange risks do not arise from lending domestically in foreign currencies. If banking regulations are applied properly, bank assets will be appropriately valued, and bank capital should serve as a necessary brake on banks’ intermediation of capital inflows. Adequate private capital at risk in the banking system should serve as a check on excesses. However, supervisors in developing countries will have to adapt international standards to fit their domestic situations. In particular, in establishing risk-weighted levels of capital adequacy, the difficulties in assessing risk in newly deregulated and liberalized economies will need to be given careful consideration. In the case of public banks, the authorities should be alert that they do not engage in behavior that could be costly to taxpayers. In particular, the government, as the main shareholder, should ensure that prudential rules applied to public banks are not discriminatory and that appropriate information on the banks’ performance and risk taking is publicly available.
Implications of Foreign Borrowing by Nonfinancial Corporations
Nonfinancial corporations should, in principle, be able to manage their own risks and should undertake such risks at their own peril. However, in the case of public sector corporations, the government is a significant shareholder, and it needs to ensure that these corporations achieve adequate returns with acceptable degrees of risk. Some of these corporations produce exclusively for the domestic market or have a concentrated import and export pattern. In these cases, the government (as a shareholder) should require that borrowing in foreign currency, for example, does not expose the firm to risks that are not appropriately hedged, or that the currency composition of debt reflects basic principles of diversification consistent with the export and import pattern of the firm.
In the case of private corporations, the government has a role in ensuring that the activities of these corporations can be adequately assessed by establishing and enforcing standards with regard to the public provision of appropriate financial information. In addition, the government should carefully monitor the foreign borrowing of domestic corporations for statistical purposes in order to aid in macroeconomic policy design and to alert the authorities specifically of potential dangers posed by the external exposure of major corporations that, because of the corporation’s size, may pose risks for the whole economy. The government should abstain from singling out “strategic” sectors that are supported. Such support not only could be quite costly (in the form of bailouts) but also could create significant moral hazard problems. As a general rule, private nonfinancial corporations that find themselves overexposed should be allowed to go bankrupt or to be bought by other corporations on market terms.