The current account of the balance of payments is an important variable for policymakers. Deficits in the external current account are the result of the accumulation of net claims of foreigners on residents. The empirical literature considers persistent deficits above 5 percent of GDP to be unsustainable in the long run. Although the recent experience of Mexico and Thailand supports this view, some countries—for example, Australia, Canada, and Japan—have in the past persistently run large current account deficits without facing external problems. Because current account balances are a function of the interaction between private agents’ intertemporal decisions and government policies, current account deficits are in themselves neither a curse nor a blessing.
This chapter provides guidelines for assessing the sustainability of current account deficits. It begins by defining the current account balance and its determinants, and it explores the relationship of the current account with key underlying variables, including investment, saving, and capital flows. It then examines the nascent econometric literature on how the sustainability of current account deficits can be determined—whether according to intertemporal solvency or according to such crucial exogenous factors as macroeconomic policy (including policy reversals and credibility) and the willingness of international investors to lend to countries with large deficits. This discussion flows into an examination of the literature that has developed a set of leading indicators (structural, macroeconomic, and overborrowing factors) to predict external crises and thus help to detect whether current account deficits may become unsustainable over the long term. The chapter then details the types of policy responses that can address these constraints so as to reduce the risk of external imbalances. The chapter concludes with a discussion of five countries (Argentina, Canada, Chile, Mexico, and Thailand) that experienced large current account deficits over certain periods and examines whether the deficits were sustainable given the structural characteristics of these economies and their macroeconomic policy stance.
The Current Account: Basic Considerations
The current account balance can be defined as the monetary value of longitudinal international flows associated with transactions in merchandise, services, factor income, and unilateral current transfers. Merchandise consists of tangible commodities. Services include transportation, travel and tourism, insurance, royalties, and license fees. Factor income captures receipts and payments arising from labor services, interest and dividend payments, profit remittances, and retained earnings of direct investment. Formally:
where CAt denotes the current account balance in period t; Xt and Mt denote, respectively, exports and imports of merchandise and nonfactor services; TRt denotes net unilateral current transfers; and NFIt, denotes net factor income from abroad at time t, defined as receipts of income minus payments.
The Current Account and the Saving-Investment Gap
National account statistics can be used to show that the current account deficit is identical to external saving. Arriving at this identity involves several steps.
From the standpoint of the national accounts, gross domestic output, which measures the total market value of all goods and services produced domestically in a given year, can also be defined as
where Y is gross domestic product, C denotes private and public consumption, and I denotes private and public investment. Gross national disposable income (GDI), the most comprehensive measure of income for a country, is defined as GDP plus unilateral current transfers plus net factor income; that is
National saving (S) is the part of gross national disposable income that domestic residents do not consume:
Using the definitions for national saving in equation (5.4), gross national disposable income in equation (5.3), and the current account balance in equation (5.1), and rearranging, one obtains the following expression:
That is, the current account balance is determined ex post by the gap between realized saving and investment. Alternatively, equation (5.5) can be expressed as
Equation (5.6) shows that investment is financed with either national saving or external saving (defined as the current account deficit). Classifying the domestic economy into private (p) and government (g) sectors, equation (5.5) can be expressed as
This expression stresses the relationships among sectoral imbalances. For instance, a fiscal deficit, (S – I)g < 0, can be financed either with private domestic saving in excess of private investment, (S – I)p > 0, or with external saving, CA < 0. Because equation (5.7) is an identity, more information is required to establish causal links among these macroeconomic variables.
The Current Account and Capital Flows
The current account balance (a flow) modifies the stock of net foreign claims on the home country; for example, financing a current account deficit requires reducing the stock of the country’s net foreign assets.
Consider the similarity between a country’s budget constraint and that of an individual who spends his or her total income on different types of goods, including financial assets. For a country as a whole, total disposable income (GDI) can be spent on consumption goods (C), capital goods (I), new monetary assets (Rt – Rt-1), and new net financial assets (Bt – Bt-1).1 Formally:
The term Bt – Bt-1 also measures the capital account in domestic currency. If Bt – Bt-1 > 0, residents are acquiring new net financial assets, and thus an outflow of capital occurs in this period. Such financial assets include not only interest-bearing financial assets, but also the ownership of capital goods located abroad (including foreign direct investment). The term Rt – Rt-1 measures the change in the stock of international currency (expressed in domestic currency) held by the country’s monetary authorities as reserves, such as gold and U.S. currency. Rearranging equations (5.4), (5.5), and (5.8) yields the country’s balance of payments equation:
Because both net financial assets and monetary assets are defined broadly as net foreign assets (NFA), equation (5.9) can be expressed as
The country’s balance of payments, as expressed in equation (5.10), tells us that a country’s current account deficit must be financed with a reduction in foreign assets—or, more specifically, as in equation (5.9), either with the depletion of international reserves or financial assets or with an increase in foreign liabilities.
Determinants of Current Account Balances
As already mentioned, current account balances are a function of the interaction between private agents’ intertemporal decisions and government policies. Equation (5.7) can be seen as an equilibrium condition, establishing behavioral relationships for desired saving and investment in a given institutional framework. On the one hand, when international borrowing and lending are allowed and the home-country economy is small, international interest rates will prevail. On the other hand, private saving depends positively on both the interest rate and the home country’s income level, whereas private investment depends negatively on the interest rate. Because public saving and investment are taken as policy variables, the current account balance is obtained as a residual.2 This framework is represented in Figure 5.1. At point A the current account is balanced; therefore, if the international interest rate is below (above) the “autarky” interest rate, r0, current account deficits (surpluses) emerge.3
Internatonal Interest Rates and the Current Account
Although recent experience shows that current account deficits lead to an accumulation of foreign liabilities that may eventually become unsustainable—that is, deficits that cannot continue to be financed on a voluntary basis in the medium term—we want to show through this simple framework that this experience is not always the case. Current account deficits can be caused by either a reduction in saving or an increase in investment.
A Decline in National Saving
A temporary adverse shock to output causes output to fall temporarily. But because people continue to consume at about the same level, private saving declines, which in turn leads to a current account deficit. Such a deficit is considered sustainable because it is temporary. As output returns to normal, saving increases, improving the current account balance. This relationship is shown in Figure 5.2. Because of the temporary output shock, the S curve shifts to the left. At r*, the current account deficit is shown by I0 – S1. Because the shock is temporary, the S curve shifts back to the right in the next period, although not to its original location, given that interest payments are now higher. Usually, the home economy repays the debt when a temporary positive shock to output is realized.
Impact of a Temporary Decline in Saving on the Current Account
A permanent decline in national saving—for example, due to large and persistent low public saving that is not matched entirely by higher private saving—creates permanent current account deficits, which might imply an unsustainable accumulation of external debt. Within the saving-investment framework, the S curve shifts to the left, and as interest payments increase, the S curve continues its leftward shift, causing the current account deficits to deteriorate further (Figure 5.3). This situation would be unsustainable in the long run unless the economy experienced growth rates above world interest rates.4 In this case, the critical issue is not the decline in saving, but that the decline is perceived to be permanent.
Impact of a Permanent Decline in Saving on the Current Account
An Increase in Investment
Suppose that the country experiences an investment boom (for example, owing to the discovery of natural resources), and that domestic saving is insufficient to finance domestic investment. In this circumstance, large current account deficits emerge (represented by a right-ward shift of the I curve in Figure 5.4). The higher level of investment implies that, under the assumption that the investment projects are profitable, future growth will be more rapid, with the buildup of larger productive capacity. Consumption levels adjust upward given the country’s greater permanent wealth; however, this increase is less than proportional to the increase in output, because external debt-service requirements will be higher. In our framework, the S curve moves to the right so as to generate sufficient current account surpluses to repay the external debt.
Impact of an Increase in Investment on the Current Account
If external savings are invested in unprofitable projects, the impact on future output will most likely be null. If this situation were to continue, and since interest payments on the new debt reduce gross domestic income, and hence the level of saving, the widening current account deficits would become unsustainable. In Figure 5.4, the S curve, rather than moving to the right, would move to the left. In this case, the source of the investment boom is not important; what matters is whether investment expenditures are profitable.
Approaches to Current Account Sustainability
Discussion of whether a large current account deficit is sustainable in the long run is quite recent, and theoretical developments in this area are based on previous research in public finance. Although this approach has proved very useful, some distinctions between the fiscal sector and the external sector must be taken into consideration. In particular, fiscal imbalances depend largely on government decisions about taxation and public spending, whereas the external current account is, as we have seen, a function of the interaction between private decisions (both domestic and foreign) and macroeconomic policies. Thus, defining sustainability is more complex for external current account imbalances.
The literature has adopted two main approaches. Considering that current account deficits imply an increase in the outstanding external liabilities of an economy, one approach focuses on a country’s solvency—that is, whether at any point in time the net present value of future surpluses on the external balance of goods and services (called primary surpluses) is greater than or equal to outstanding external in debtedness. The other approach considers two factors explicitly: the willingness of international investors to lend to the country, and the country’s current policy stance.
Intertemporal Solvency
A country is said to be solvent at any time t if the present discounted value of the primary balance is sufficient to repay the external debt accumulated up to period t. From this perspective, a country’s current account deficit is sustainable if its intertemporal solvency is not violated. To formalize this concept and to simplify matters, consider the following version of equation (5.9):
where Dt and Dt-1 are the outstanding external debt at the end of periods t and t – 1, respectively. Note that, in this case, Rt – Rt-1 = 0 and Bt – Bt-1 = –(Dt – Dt-1); that is, current account deficits are financed solely by borrowing from abroad. Current account balances can be broken into two components: the primary balance (Tt), or the balance in goods, nonfactor services, and transfers, and the net income factor, which in this case consists solely of interest payments (-iDt-1). Formally:
Combining equations (5.11) and (5.12) obtains:
According to equation (5.13), reducing the external debt requires that the primary balance be greater than zero (that is, a surplus) and greater than interest payments. Rearranging equation (5.13), one can express the stock of debt at time t in terms of the present values of the debt at t + 1 and the primary balance at t + 1. Formally:
Assuming that the number of future periods is finite (N), the forward-looking solution of equation (5.14) is given by
A strict and narrow solvency condition implies that the external debt must be repaid fully at the end of period N (Dt+N = 0). Thus, a country is said to be solvent if the discounted sum of expected future primary surpluses is enough to repay the current debt; otherwise, the country is considered insolvent. The latter was the case with many Latin American countries during the 1982 debt crisis, which were denied access to world capital markets because they were perceived to be bad risks.5
The solvency condition, as defined by equation (5.15), is affected by different factors, including the initial debt stock, its maturity, and the interest rate. For a larger initial debt, the country needs larger primary surpluses to satisfy the solvency condition. Larger primary surpluses are also needed if the interest rate is higher or the length of period N shorter.
A broader definition of solvency can be obtained when weaker conditions are imposed on equation (5.15), such as a constant debt-to-GDP ratio. In this case, dividing equation (5.14) by GDP, under the assumption that the real exchange rate is held constant, and after some algebraic transformations, one obtains
where d denotes the debt-to-GDP ratio, tb is the primary balance as a percentage of GDP, r is the real interest rate, and g denotes the growth rate of real GDP. For a finite horizon N, the forward-looking solution of equation (5.16) is given by
If one imposes the strict solvency condition that dt+N = 0, then the higher the growth rate is in relation to the real interest rate, the smaller are the primary surpluses necessary to repay the debt. However, imposing a less strict solvency condition, such as a constant debt-to-GDP ratio, and provided that g > r, a country can satisfy this solvency condition even when the discounted sum of expected future primary balances is negative.
The following example clarifies how the concept of current account sustainability is made operational. For a country with a debt-to-GDP ratio of 0.5, with the maturity of the stock of debt at 20 years, and with a constant real growth rate and a real interest rate of 2 percent and 4 percent, respectively, the country would meet the strict solvency condition—that is, dt+N = 0—if it ran primary surpluses of 3.1 percent of GDP during the next 20 years. However, if a weaker condition of solvency were imposed—that is, dt = dt+N—it would have to run primary surpluses of only 1 percent of GDP.
Two criticisms have been leveled at the use of intertemporal solvency as a sustainability concept. First, although equation (5.17) tells us about the country’s ability to pay, it does not tell us about its willingness to pay. This distinction, first emphasized by Eaton and Gersovitz (1981), stems from the possibility that sovereign governments can renegotiate debt contracts because foreign lenders are unable to enforce debt repayments. The existence of this imperfection, called sovereign risk, implies that even solvent countries can be denied credit.
Second, equation (5.17) does not capture the macroeconomic policy stance explicitly. Changes in macroeconomic policy, affecting the real exchange rate or the level of domestic absorption (or both), have an impact on the country’s primary balance and, by extension, on the intertemporal solvency condition. Thus, although ex post solvency can always be achieved with changes in the policy regime, the crucial point is determining whether the country’s solvency is satisfied if the current policy stance remains unchanged (ex ante solvency).
Sustainability of the Current Policy Stance
Milesi-Ferretti and Razin (1996) developed another, more complete, approach to current account sustainability. In their view, a crucial restriction is imposed on the intertemporal solvency condition: Can the government continue indefinitely to pursue its current set of policies? Clarifying this approach requires addressing the reasons behind policy reversals and the role of uncertainty.
A policy reversal is needed when continuation of the current policy stance violates the intertemporal solvency condition. Consider, for example, an expansionary fiscal policy accommodated by monetary policy under a fixed exchange rate regime. The resulting expansion in domestic absorption will tend to produce domestic inflation (basically in the nontradables sector) and current account deficits. Because continuing this policy mix increases the accumulation of external liabilities, the policy stance must shift to reverse the current account position as soon as either access to external borrowing is denied or international reserves are depleted or near depletion.
It is important to recognize that an adverse domestic or foreign shock can trigger a macroeconomic policy reversal. For example, a permanent deterioration in the terms of trade will require depreciating the currency in real terms or reducing domestic absorption to maintain macroeconomic stability. Sooner or later, the government will be forced to change the preshock policy stance. Thus, the sustainability of the policy stance depends on the macroeconomic environment.
Because sustainability depends on the macroeconomic environment, the reaction of private agents who believe that a change in the policy stance will take place could actually force the authorities to change it, even if the policy mix were consistent ex ante with the macroeconomic environment (the prediction of a policy change is then a self-fulfilling prophecy). For instance, deficits in the current account can be sustained if the economy regularly receives net capital inflows. If investors suddenly come to believe that the local currency will be devalued, they might change their portfolio composition, substituting foreign financial assets for domestic ones to avoid capital losses. This drastic reversal imposes reserve losses, compelling the government to devalue. Two factors are worth noting. First, a policy shift is not needed for such a speculative attack to occur. Second, the change in investors’ confidence can be driven by exogenous shocks, as has happened to many countries (for example, Argentina experienced net capital outflows following the Mexican crisis).6 In assessing current account sustainability, this approach relies on a set of macroeconomic variables, used as leading indicators, to predict external crises. The next section explains how it does so.
Factors Affecting Current Account Sustainability
As discussed in the previous section, current account deficits are considered unsustainable when the continuation of the current policy stance will eventually trigger a drastic policy shift or when it leads to a crisis (for example, an exchange rate collapse that prevents the country from servicing its external obligations).7 Policy reversals can take the form of a sudden currency devaluation, or the tightening of monetary or fiscal policies or a combination of these, leading to a drastic contraction of domestic absorption and a sudden shift in the current account position. Within this framework, a large body of empirical literature has been devoted to finding variables highly correlated with external crises, identifying not only policy variables but also such structural factors as economic growth, the investment-saving gap, the openness of the economy, the composition of external liabilities, and the financial structure.8
Structural Factors
Economic Growth
As mentioned, a country experiencing growth rates well above world interest rates might run persistent current account deficits without worsening its debt-to-GDP ratio, depending in part on the sectoral composition of its economic growth. For example, an expanding export sector enhances the country’s ability to repay its external debt. However, if GDP growth is due to growth in output from the nontradables sector, the outcome will be mixed. On the one hand, the ability of the country to service its external debt will improve if, because of substitution effects, trade surpluses emerge. On the other hand, if growth in output of the nontradables sector is highly correlated with trade deficits, the country’s external vulnerability will increase, since such a correlation is often evidence of real exchange rate misalignment.
The Investment-Saving Gap
By definition, the current account deficit is the difference between national saving and domestic investment. If external deficits are a product of an increase in investment rates, a more rapid expansion in output growth is to be expected in the future, enhancing the country’s solvency.9 Conversely, if the deficits are the product of a permanent decline in the saving rate, the accumulation of external liabilities will not generate higher output, implying that the country’s external solvency will deteriorate.
Openness and Trade Patterns
As discussed earlier, the size of the export sector in relation to GDP, as a measure of openness, is a determinant of sustainability. In addition, a country is better equipped to deal with external shocks when it has an outward-looking orientation. With a severe change in external conditions (with, for example, unexpected capital outflows), productive factors must shift from nontradables to tradables. This reallocation of resources among sectors entails a cost to the country, which is negatively correlated with the degree of openness. Moreover, having a well-diversified export sector reduces vulnerability to external shocks to the terms of trade.10
Size and Composition of External Liabilities
The size of net external liabilities in relation to GDP is a natural indicator of external sustainability. A large stock of external liabilities implies a large net transfer of resources to the rest of the world in the form of financial services, affecting the country’s intertemporal solvency condition. However, even if the solvency condition is met, current account imbalances can become unsustainable if the country is under liquidity pressures. The risks of suffering such liquidity crises are strongly related to the maturity structure and the composition of external liabilities. For example, in a confidence crisis, a country experiences not only a drastic reduction in capital inflows, but also a surge in capital outflows. The dominance of short-term liabilities over other liabilities with longer maturities poses a high risk of capital outflows (reversing short-term flows is less expensive than reversing long-term flows).11 Furthermore, foreign direct investment flows have positive externalities, such as technology and management expertise, that make this type of capita] flow preferable to others.
Financial Structure
The recent experience in Asia has highlighted the importance of financial sector weaknesses as a determinant of external crises, or as a factor that exacerbates their impact on the economy. If prudential regulations are not in place and supervision is ineffective, banks can take excessive risks due, for example, to inadequate provisions. Against this background, a relatively minor adverse output shock could easily create a serious financial crisis, which in turn could cause an external crisis. In particular, under a fixed exchange rate regime, the decision of the central bank to bail out affected financial institutions would expand domestic credit and thus worsen inflation. This would weaken the country’s external position as the currency appreciates in real terms and current account deficits emerge.
Conversely, an external crisis can also cause financial distress. For example, consider a financial system in which a substantial percentage of liabilities are denominated in foreign currency. In this context, financial vulnerability comes from two sources: exchange rate risks associated with the mismatch between liabilities and assets denominated in foreign currency, and credit risks stemming from the effect of the devaluation on the borrower’s repayment performance. In these circumstances, a large and unexpected devaluation, triggered by an unsustainable current account deficit, could seriously harm the financial system.
The importance of the financial structure as a determinant of external crises has been extensively documented in the literature. For example, Goldstein (1996) found that such variables as the weakness of the banking sector, short-term borrowing, and mismatches between assets and liabilities are crucial to understanding the degree of vulnerability to external shocks. Kaminsky, Lizondo, and Reinhart (1997) also found evidence of the importance of the financial sector, although they used different explanatory variables, such as the change in stock prices, financial liberalization, and banking crises.
Other Factors
Political instability, policy uncertainty, weak credibility, and market expectations are also important for determining the sustainability of current account deficits. For example, the last three external crises in Mexico (in 1976, 1982, and 1994) occurred in election years. Political instability, the absence of credibility, and highly uncertain policy could cause not only capital outflows (creating short-term problems for financing the current account deficit) but also low investment and low growth rates (impairing the country’s solvency).
Macroeconomic Policy
Monetary and Exchange Rate Policy
The choice of nominal anchor is critical to determining the ability of an economy to sustain current account deficits. For example, when a fixed exchange rate regime is not accompanied by consistent monetary and fiscal policies, the resulting real appreciation widens the current account deficit. If this policy mix continues over time, at some point the current account deficits become unsustainable, either because external borrowing is no longer available or because international reserves have been depleted. The anticipation of a policy reversal could trigger a speculative attack (Krugman, 1979). In particular, Cumby and van Wijnbergen (1989) found that domestic credit growth was key to explaining the external crisis in Argentina in the late 1970s, and Blanco and Garber (1986) found evidence that domestic credit growth was an important determinant of the 1982 Mexican crisis.
A continued real appreciation, primarily owing to an inconsistent policy mix, can also be a good indicator of an external crisis. However, a large real appreciation can also be caused by “fundamental” factors— such as high productivity growth in the tradables sector, favorable terms of trade, or the discovery of natural resources. In such cases, the real appreciation does not imply a deterioration of the external account. Currency appreciation can also stem from capital inflows, in which case the sustainability of policies must be assessed according to other indicators (including the variable triggering the inflows). For example, if capital inflows are driven by policy changes that improve the country’s long-term fundamentals (such as structural reform and trade liberalization), the resulting exchange rate appreciation corresponds to the long-run equilibrium in the economy. However, if capital inflows are of a short-term, speculative nature, the ensuing appreciation will cause a misalignment rather than reflect long-run equilibrium.
Fiscal Policy
Under the debt neutrality hypothesis (Barro, 1997), the current account is independent of both the time profile of taxation and, for a given public expenditure path, the public sector deficit. The rationale for this result is that since decisions about consumption, saving, and investment are based on the present value of net disposable income, every dollar of taxes postponed today must be paid with interest in the future. Thus, an increase in the budget deficit leads to an equal, in stantaneous increase in private saving without affecting aggregate wealth. This implies that a link between the current account and the fiscal deficit does not exist.12
The assertion that government budget imbalances are irrelevant to resource allocation is known as the Ricardian equivalence principle. However, very strong assumptions are required for this principle to hold. For example, market imperfections that distort borrowing and lending decisions will invalidate the principle, affecting the relationship between fiscal imbalances and the current account deficit, since private investment and saving decisions also react to changes in fiscal policy.13
In a different context, coordination between fiscal and monetary and exchange rate policies becomes even more important when an economy is experiencing large capital inflows. If the potential volatility of capital flows is a source of concern because of its implications for nominal exchange rate fluctuations, a fixed exchange rate policy can be pursued. However, in this case, a strong fiscal policy is advised in order to cope with the potential misalignment of the real exchange rate. A tight monetary policy is inappropriate both for fighting inflation and for avoiding real appreciation because, by keeping domestic interest rates high, it creates incentives for speculative short-term capital inflows. Thus, fiscal policy must be strong enough to partially offset the adverse effect of capital inflows on private expenditure (Wong and Carranza, 1999).
Capital Controls
Countries using fixed or quasi-fixed exchange rate regimes and lacking fiscal flexibility must depend on tight monetary policies to fight inflation. But since capital mobility implies that monetary policy becomes ineffective, the authorities must rely on capital controls to achieve otherwise inconsistent macroeconomic objectives (such as a high interest rate that is inconsistent with an undervalued local currency). Capital controls can be effective at blocking capital inflows that may cause large current account deficits, at least in the short run, although they tend to become ineffective in the medium term (Dooley, 1996). As the recent Chilean experience shows, however, capital controls may contribute to lengthening the maturity of external liabilities (IMF, 1998).
During the 1970s and 1980s, capital controls were used to prevent capital flight, rather than to block capital inflows. In those years, a combination of expansionary monetary policies, fiscal deficits, active management of the exchange rate, and repressed financial systems created real appreciation, high inflation, and negative real interest rates. Under such circumstances, asset holders sought protection by holding dollar-denominated assets within and outside the country.14 To avoid capital flight and a fall in reserves, governments imposed several types of controls on outflows, which often yielded perverse results. Given the ineffectiveness of such controls and, above all, the negative signals to the markets, these controls, rather than stopping capital flight, helped stop potential foreign investment.15
The Overborrowing Syndrome
Uncertainty, distortions, and imperfect information can lead to over-borrowing and can jeopardize current account sustainability. The phenomenon of overborrowing was analyzed first in the context of trade reforms and the absence of full credibility (Mussa, 1986; Calvo, 1987). If economic agents believe that the elimination of trade restrictions is temporary, they will exploit this window of opportunity to increase consumption, particularly of durable goods. The increase in consumption leads to excessive borrowing, including external borrowing if the capital account is open. In this situation, policy should try to boost credibility by making reforms feasible overall and by implementing policy commitments that are not subject to the time inconsistency problem.
A recent strand of the literature analyzes overborrowing in the context of fully credible macroeconomic stabilization and reform of the real economy (McKinnon and Pill, 1996, 1997). The authors explain the overborrowing syndrome by focusing on uncertainty, moral hazard, and imperfect information. This syndrome entails “excessive” capital inflows that produce excess domestic demand, inflation in nontradables, loss of monetary control, and, finally, likely reversal of the reform process. The authors assert that many liberalizing economies have suffered overborrowing during boom times, with the domestic banking system the principal intermediary between foreign lenders (depositors) and domestic borrowers.
The McKinnon and Pill model assumes that credible reforms raise the expected productivity of new investment above the level attainable before the reform; thus, some agents undertake discrete new investments, which are large in relation to their initial endowments and whose payoff is uncertain. In this scenario, overborrowing might emerge because banks lend too exuberantly (because of unrestricted capital inflows, coupled with the moral hazard effect of deposit insurance schemes), thus signaling higher payoffs for investments than the reform warrants. The gist of their analysis is that the financial system induces market participants to take relatively higher risks, financed by overborrowing.
According to McKinnon and Pill, the main distinguishing features of an overborrowing episode include (1) rapid growth in domestic credit, financed largely with capital inflows intermediated through the domestic banking system, leading to higher levels of domestic consumption; (2) a larger current account deficit; (3) weaker domestic monetary control and higher domestic inflation; and (4) a real appreciation of the currency.
Policy Responses to External Imbalances
Designing a policy package that reduces the risks of external crises requires identifying the source of imbalances.16 If current account deficits are caused by fiscal deficits, as was the case in Argentina, the optimal policy response would consist of a combination of Keynesian-type fiscal and monetary policies and exchange rate policy.
However, when large capital inflows prompt the private sector to increase expenditure (consumption, investment, or both), which in turn leads to current account deficits, the appropriate policy response depends on the composition and use of the capital inflows. In that case, the first step is to ascertain the effect on the real exchange rate. If capital inflows predominantly take the form of portfolio investment or other short-term inflows, the equilibrium real exchange rate will probably depreciate if these inflows are used to finance consumption or unproductive activities, and it will probably appreciate if they are channeled into productive capital formation.17 If the equilibrium real exchange rate is impossible to estimate reliably, it is better to make the nominal exchange rate sufficiently flexible, so that market forces can establish its level. This isolates the monetary base from changes in net foreign assets, avoids speculation, and further reduces capital inflows by letting foreign investors bear more exchange rate risk. At the same time, flexible exchange rates would enable the authorities to focus on other necessary measures, such as strengthening the financial system.
The second step is to ascertain the viability of a fixed exchange rate regime. Because such a regime is more susceptible to arbitrage and speculative attacks in a world of high capital mobility, it should be implemented only when (1) a country’s monetary and fiscal policies are flexible enough to keep the real exchange rate in line with the fundamentals; (2) ample international reserves are available to defend the exchange rate and maintain the confidence of investors; and (3) the country has both a strong and independent central bank, prepared to let interest rates increase to curb speculative attacks, and a financial system strong enough to withstand large swings in foreign assets and liabilities.
The third step in the policy response is to ascertain the timing of the policy action. Fiscal restraint is usually more effective at the beginning of the external adjustment, when current account deficits are relatively small. Decisive action on the exchange rate should be taken once it becomes evident that a stable real exchange rate cannot be maintained. In this context, it is preferable that real exchange rate realignment be made through changes in the nominal exchange rate, rather than through changes in domestic prices. Domestic price changes are particularly relevant when the fiscal adjustment required to achieve a desirable real depreciation is too expensive.
The final step is to monitor the expansion of private credit. A rapid expansion of credit in the financial system could be the catalyst for a financial crisis, since it might encourage excessive risk taking by financial institutions. Although, in principle, banks are risk-neutral agents, they can take excessive risk in the face of rapid credit expansion. Implicit free deposit insurance schemes will induce banks to operate with huge mismatches between assets and liabilities and with inadequate credit screening, leading to credit misallocation. Strengthened reforms and surveillance of the financial system are necessary measures for a successful adjustment process.
Country Experience
This section briefly discusses five countries with large current account deficits, focusing on the sustainability of these deficits, the structural characteristics of the economy, and the macroeconomic policy stance. Of the five cases, only Canada from 1976 to 1996 had sustainable current account deficits. In Argentina in the 1970s, fiscal imbalances were the primary determinant of current account deficits, whereas in Chile in 1982, Mexico in 1994, and Thailand in 1997 private decisions drove current account deficits. (Figures 5.5-5.9 provide general background information and specific reference points for the discussion.)
Canada: Real Effective Exchange Rate and the Current Account
(Index, 1990 = 100)
Source: IMF, IMF Institute database.Argentina: Real Effective Exchange Rate and the Current Account
(Index, 1990 = 100)
Source: IMF, IMF Institute database.Chile: Real Exchange Rate and the Current Account
(Index, 1988 = 100)
Source: IMF, IMF Institute database.Mexico: Real Effective Exchange Rate and the Current Account
(Index, 1990 = 100)
Source: IMF, IMF Institute database.Mexico: Composition of Capital Inflows
(Billions of U.S. dollars)
Source: World Bank, Global Development Finance database.Canada, 1976–96: A Case of Intertemporal Solvency
Despite having run large current account deficits from 1976 to 1996, Canada is an example of intertemporal solvency. Three main features of Canada’s economy explain the country’s ability to continue its access to international credit markets: consistent government adherence to the policy mix, sustainability of fiscal deficits, and strong real sector performance.
Canada has maintained one of the world’s more flexible exchange rate regimes since the collapse of the Bretton Woods system. Because the nominal exchange rate has been sufficiently flexible to reflect conditions in the tradables and nontradables sectors of the economy, excessive real exchange misalignments have been avoided (top panel of Figure 5.5). This exchange rate policy, combined with prudent fiscal and monetary policies, especially in the 1990s, is responsible for the strong merchandise trade surpluses recorded for most of the period under consideration (bottom panel of Figure 5.5). Furthermore, because the current account deficits were driven primarily by the public sector, the government’s intertemporal solvency is key to understanding Canada’s external sustainability. In fact, a good indicator that the government’s solvency was not compromised is the low level of total public debt.
Canada’s creditworthiness can also be gauged by its sustained economic growth in relation to growth in net debt service. Furthermore, the country’s real external debt burden, defined as the payments that the country needs to make to keep its external debt-to-GDP ratio constant, has remained fairly stable at 1.6 percent of GDP. In contrast, this ratio has fluctuated drastically in other countries: the ratio in Brazil went from zero in 1970 to 1.3 percent in 1983, and that in Argentina went from about 0.5 percent in 1970 to almost 3.0 percent in 1983. The higher is the external debt burden, the greater is the probability that current account deficits will become unsustainable. Aside from economic growth, several other real sector indicators illustrate sustainability: for example, the level of domestic saving, the level of openness, and the composition and maturity structure of external liabilities.
In cases where the solvency condition holds, it means only that the ability-to-pay criterion has been met. It does not tell us anything about the willingness-to-pay criterion. In the context of Canada, other structural factors, such as stable political institutions and reputation, are obviously important factors that explain why Canada was not cut off from international capital markets after the crisis in 1982.
Argentina, 1976–81: Unsustainable Fiscal Deficits
In Argentina, the continuation of expansionary fiscal policies and a quasi-fixed exchange rate regime caused large external imbalances, which ultimately became unsustainable in the long run.
The Argentine experience is interesting because the economy suffered two episodes of external crisis within a period of only a few years, both of which were caused by excessive fiscal deficits. The first crisis was in March 1976, when annual inflation reached more than 300 percent, and the overall fiscal deficit was about 17 percent of GDP. These high rates contributed to current account deficits that were clearly unsustainable in the long run (see Figure 5.6). Consequently, a drastic policy reversal was implemented when the government of Isabel Perón was replaced by a military government in 1976.
The policy changes introduced by the military government included the gradual removal of capital controls, the elimination of interest rate controls, and the progressive removal of trade barriers. The changes also contemplated monetary tightening and substantial fiscal adjustment. Although the government reduced the current account deficit, the absence of progress on inflation prompted a new stabilization plan, which was launched in December 1978 in the throes of an inflation rate of about 100 percent. Under the new plan, the path of the nominal exchange rate (which was the anchor) was governed by a preannounced daily rate of crawl (the tablita). The purpose of this scheme was threefold: to avoid a real appreciation, to reduce inflation gradually, and to enhance credibility by promulgating the nominal depreciation in advance.18
However, because the government was unable to significantly reduce the fiscal deficit—which was financed with a combination of external borrowing and domestic credit expansion—inflation accelerated, and current account deficits widened, undermining confidence in the exchange rate regime. Since these deficits were considered unsustainable and the exchange rate regime was expected to collapse, private capital took flight, accelerating losses in official reserves. This setback prompted the authorities to abandon the exchange rate regime and to devalue the peso by 10 and 31 percent in February 1981 and April 1981, respectively.
Chile, 1979–82: Currency Crisis and Financial Structure
In Chile, the nonfinancial public sector balance was in surplus throughout the period under consideration. The main determinants of the Chilean crisis, which was triggered by external shocks, were a currency that was overvalued in real terms and a weak financial system. As the crisis unfolded, the financial system finally collapsed, deepening and prolonging the crisis.
A new policy regime implemented by the military government in 1973 consisted of orthodox stabilization policies combined with far-reaching structural reforms, including privatization and trade and financial liberalization. Despite the high cost of the policy turnaround for output and employment, the disinflation process was slow. In 1978 the annual inflation rate was still close to 40 percent, and the unemployment rate hovered around 14 percent. The persistence of inflation prompted policymakers to use the nominal exchange rate as the major stabilizing device in the economy. A preannounced crawling peg regime (the tablita) was established in February 1978, and in June 1979 it was replaced by a fixed exchange rate regime. The reasoning behind this plan was that, because the openness of the economy allowed domestic prices for tradables to be determined by their international prices, and because expectations of currency depreciation were assumed to converge to zero, domestic interest rates should converge to international levels. The assumption was that the positive impact of these results on the inflation of nontradables would reduce the overall inflation rate. However, the envisaged reduction in nontradables inflation was not achieved because of price indexation. The ex post formal indexation applied not only to wages but also to the prices of many nontradables, such as housing rentals, mortgages, and tuition fees. The resulting strong real appreciation of the peso caused a widening of the current account deficit, which was unsustainable in the long run (Figure 5.7).
Moreover, because capital controls had been lifted in the wake of financial liberalization, and prudential regulations and banking supervision remained inadequate, private borrowing from abroad became excessive, with two dire consequences. First, inertial inflation was sustained as monetary aggregates, fueled by capital inflows, grew rapidly. Second, the financial system became highly vulnerable, as banks engaged in unsound practices without covering exchange rate and credit risks.
By the end of 1981, the economy had also suffered severe external shocks—a sharp decline in the terms of trade, a large increase in international interest rates, and an abrupt contraction of capital inflows— along with an unfolding financial crisis. The bailout of private financial institutions by the central bank compromised its ability to defend the currency, as the public sector, including the central bank, accumulated large stocks of debt. This development led to a self-fulfilling crisis, as the market began to expect that the debt would be monetized in the future (Edwards, 1987). In June 1982, with the country in the midst of a confidence crisis, the authorities devalued the peso by 18 percent and then allowed it to float. The large exchange rate adjustment led to bankruptcies in the private sector, a deep recession, and a worsening financial crisis, which in turn led in 1985 to quasi-fiscal losses by the central bank of 18 percent of GDP, declining slowly thereafter.
Mexico, 1987–95: Current Account Deficits and Private Decisions
The Mexican crisis of 1994-95 was caused by an inconsistent policy response to a combination of adverse exogenous shocks and misalignment of the exchange rate. In the late 1980s, the government sought to lower inflation and foster growth through economic reform—debt restructuring and trade and financial liberalization—and sound monetary and fiscal policies. The economy responded: inflation fell from 160 percent in 1987 to single-digit levels at the end of 1993, and annual growth averaged 3.5 percent over the same period. However, the external position remained weak. The exchange rate-based stabilization program led to a decline in real interest rates and an expansion in aggregate demand, which in turn led to a large real appreciation of the peso—and thus to protracted current account deficits that were unsustainable in the long run (Figure 5.8). The real appreciation, however, was deemed consistent with the fundamentals, in light of the ongoing comprehensive structural reforms. Because the private sector anticipated higher future income, consumption and investment increased, expanding imports rapidly and thus widening the current account deficits. Moreover, solid developments in the export sector reinforced this perception that the peso’s appreciation was in line with the fundamentals.19
Although many analysts considered the current account deficits sustainable because they were caused by private decisions, Mexico suffered a severe external crisis at the end of 1994, which continued during the first few months of 1995. The crisis had its roots in exogenous shocks in early 1994 (political difficulties, as well as increases in U.S. interest rates). In a few months, from January to April, the exchange rate approached the most depreciated margin of its band, and continuous capital outflows imposed losses of almost half of official reserves. As investors continued to have doubts about the stability of the exchange rate, they moved from financial assets denominated in domestic currency to financial assets denominated in foreign currency. This shift was reflected in a dramatic recomposition of public domestic debt (Figure 5.9). The stock of tesobonos (dollar-indexed instruments), which was practically zero at the beginning of the year, jumped to almost 50 percent of total public domestic debt by October 1994.
Although the combination of negative external shocks and real appreciation called for a much tighter monetary policy, the authorities, believing that the confidence crisis was temporary, chose instead to sterilize the reserve losses so as to prevent a further increase in interest rates. Furthermore, rather than using fiscal restraint to reduce the exchange rate misalignment and current account deficits, the government also experienced fiscal policy slippages throughout 1994. The confidence crisis worsened in late 1994, and on December 20, when reserves were almost depleted, the peso was devalued by 15 percent. The pressure on the foreign exchange markets continued unabated, however, and on December 22 the authorities were forced to let the exchange rate float. Investors were reluctant to roll over the debt because they recognized the high risk of default on government bonds in denominated foreign currency given that the reserve holdings of the Bank of Mexico were so low. To avoid a self-fulfilling debt crisis, the international community had to come to the rescue.20
Thailand, 1991–97: A Recent Experience of Currency Crisis
From 1991 to 1997, Thailand’s current account deficits, which were caused by private decisions, became unsustainable because (1) highly distorted incentive mechanisms led to overinvestment in the economy; (2) the financial system was poorly regulated and supervised; (3) the composition of external liabilities was untenable; and (4) external conditions had deteriorated.21
On July 2, 1997, the Thai authorities let the baht depreciate after fruitless attempts to defend it. Until the eve of the attack on May 7, 1997, the Thai economy had been considered an example of sustained growth, with strong macroeconomic performance characterized by a conservative fiscal stance, high saving rates, and low inflation. However, some concerns had been expressed about the country’s large current account deficits (Figure 5.10).
Thailand: Real Effective Exchange Rate
(Index, 1990= 100)
Source: IMF, IMF Institute database.Several factors were behind this crisis. First, structural and financial weaknesses had given rise to low-quality and excessive investment. The rapid increase in the rate of investment, from an average of 28 percent of GDP to about 40 percent in the 1990s, had gone to real estate, other speculative activities, and industries with significant overcapacity. The increase in the incremental capital-output ratio was a clear signal that the productivity of investment was declining sharply.
Thailand: Composition of Capital Inflows by Type and Maturity
(Billions of U.S. dollars
Source: World Bank, Global Development Finance database.Second, this low-quality investment was financed by a credit boom stoked by large net private capital inflows. And since investment leaned heavily toward real estate, the financial system became more vulnerable to shifts in asset prices. Furthermore, poor financial regulation and supervision, reflected in inadequate lending practices, an absence of corporate governance, and deficient risk management, contributed to the vulnerability of the financial system. In this connection, the financial system was operating with huge liquidity and currency mismatches, which in turn allowed the possibility of speculative attacks.
Third, the composition of capital inflows was flawed (Figure 5.11). From 1994 to 1996, short-term inflows represented 7 to 10 percent of GDP each year, while foreign direct investment flows were about 1 percent of GDP. Because the dominance of short-term capital increased the risk of unexpected capital outflows, the economy became more vulnerable to shifts in market sentiment. Moreover, the external environment changed. In 1996 Thailand’s export sector experienced a slowdown because of temporary factors such as a decline in world trade. This trend was heightened by strong competition from China. Furthermore, as the U.S. dollar recovered against the yen, Thailand’s export sector suffered further losses in international competitiveness.
When it became clear that the excessive investment was unprofitable, that the financial system was weak, and that the economy was suffering from adverse external shocks, the authorities realized that the baht was overvalued and that the current account deficits were unsustainable. Despite concern about the weakening external position, they were forced to defend the baht as they recognized the large currency exposure of the financial system.
Conclusions
This chapter has provided a set of general guidelines for assessing the sustainability of current account deficits. Analysis of whether a certain level of current account deficit is sustainable in the long run is quite recent, with two dominant approaches in the literature.
The first approach focuses on a country’s solvency: whether at any point in time the net present value of future surpluses on the external balance of goods and services (the primary surplus) is enough to extinguish or reduce the outstanding external indebtedness. The second approach improves on the concept of solvency by imposing more structure on the analysis. In particular, it takes into account the degree of difficulty in obtaining foreign borrowing, and the current policy stance on which the concept of solvency is made conditional. In assessing current account sustainability, the second approach relies on a set of macroeconomic variables as leading indicators to predict external crisis, identifying not only policy variables but also such structural factors as openness, the composition of external liabilities, the financial structure, and economic growth.
Appendix 5.1: Indicators of External Crises: A Survey
This appendix surveys the literature on indicators of currency crises (from Kaminsky, Lizondo, and Reinhart, 1997).
Indicators of External Crises
Indicators of External Crises
| Study | Sample and Frequency | Country Coverage | Indicator |
|---|---|---|---|
| Bilson (1979) | 1955–77, annually | 32 countries, with emphasis on Ecuador, Mexico, and Peru | (1) International reserves/base money (2) “Shadow” exchange rate |
| Blanco and Garber (1986) | 1973–81, quarterly | Mexico | (1) Domestic credit growth |
| Calvo and Mendoza (1996) | 1983–94, Monthly and quarterly | Mexico | (1) M2 (in dollars)/reserve (2) The money demand-supply gap |
| Collins (1995) | 1979–91, annually | 18 countries with pegged exchange rates at the beginning of 1979 | (1) International reserves/GDP (2) Real GDP growth (3) Change in the real exchange rate (4) Multiple exchange rate dummy (5) Inflation (6) Current account/GDP (7) Foreign aid |
| Cumby and van Wijenbergen (1989) | 1979–80, monthly | Argentina | (1) Domestic credit growth |
| Dornbusch, Goldfajn, and Valdes (1995) | 1975–95, annual and quarterly | Argentina, Brazil, Chile, Finland, and Mexico | (1) Real exchange rate (2) Real interest rates (3) GDP growth (4) Inflation (5) Fiscal deficit/GDP (6) Credit growth (7) Trade balance/GDP (8) Current account/GDP (9) International reserves (10) Debt/GDP |
| Edin and Vredin (1993) | 1978–89, monthly | Denmark, Finland, Sweden, and Norway | (1) Money (2) Output (3) Foreign interest rate (4) Foreign price level (5) Real exchange rate (6) International reserves/imports (7) Trade balance |
| Edwards (1989) | 1962–82, pooled quarterly data and annually | 24 developing countries | (1) Central bank foreign assets/base money (2) Net foreign assets/M1 (3) Domestic credit to public sector/total credit (4) Bilateral real exchange rate (5) Parallel market premium (6) Growth of credit (7) Growth of credit to the public sector (8) Public-sector credit growth/GDP (9) Fiscal deficit/GDP (10) Current account/GDP (11) Terms of trade (12) Errors and omissions plus short-term capital (13) Exchange controls |
| Eichengreen, Rose, and Wyplosz (1995) | 1959–93, quarterly | 20 industrial countries; 78 crises | The authors mention that many of (1) –(16) are defined with respect to the same variable in Germany, but do not specify which of them. (1) Change in international reserves (2) Real effective exchange rate (3) Credit growth (4) M1 growth (5) Bond yield (6) Interest rates (7) Stock prices (8) Inflation (9) Wage growth (10) GDP growth (11) Unemployment rate (12) Employment growth (13) Fiscal deficit/GDP (14) Current account/GDP (15) Change in exports (16) Change in imports (17) Government victory (18) Government loss (19) Elections (20) Change in government (21) Capital controls (22) Left-wing government (23) New finance minister (24) Past exchange market crisis (25) Past exchange market event |
| Flood and Marion (1995) | 1957–91, monthly | 17 Latin American countries; 80 peg periods of a duration of at least 3 months | (1) Drift of the real exchange rate (2) Variance of the real exchange rate |
| Frankel and Rose (1996) | 1971–92, annually | 105 developing countries | (1) Credit growth (2) Fiscal deficit/GDP (3) Per-capita GDP growth (4) External debt/GDP (5) Reserves/imports (6) Current account/GDP (7) Deviations from PPP in the bilateral real exchange rate (8) OECD GDP growth (9) Foreign interest rate The following variables as a share of total debt: (10) Commercial bank loans (11) Concessional loans (12) Variable rate debt (13) Short-term debt (14) Public-sector debt (15) Multilateral development bank loans (16) Flow of FDI |
| Goldstein (1996) | Annually and monthly | Argentina, Brazil, Chile, and Mexico (other crises are also discussed). | (1) International interest rates (2) Mismatch between the government’s and the banking sector’s short-term assets and liabilities (such as M3/reserves) (3) Current account/GDP, particularly one driven by a fall in savings (4) Boom in bank lending followed by a decline in asset prices (5) Real exchange rate (6) Short-term borrowing (7) Weak banking sector |
| Humberto, Julio, and Herrera (1991) | Monthly | Colombia | (1) Credit growth (2) Parallel market premium |
| Kamin (1988) | 1953–83, annually | 107 cases with devaluations of at least 15 percent with respect to the U.S. dollar | (1) Trade balance/GDP (2) Import growth (3) Export growth (4) Capital flows/GDP (5) Changes in reserves (6) Inflation (7) The real exchange rate (8) Real GDP growth (9) Change in export prices |
| Kaminsky and Leiderman (1995) | 1970–95, monthly | 20 countries (5 industrial and 15 developing) | (1) Export growth (2) Import growth (3) Bilateral real exchange rate-deviation from trend (4) Terms of trade changes (5) Changes in reserves (6) Money demand/supply gap (7) Changes in bank deposits (8) Real interest rates (9) Lending-deposit spread (10) Domestic-foreign real interest rate differential (11) M2 money multiplier (12) M2/international reserves (13) Growth in domestic credit/GDP (14) Changes in stock prices (15) Output growth (16) Financial liberalization (17) Banking crises |
| Klein and Marion (1994) | 1957–91, monthly | 87 peg episodes, as in Flood and Marion (1995) | (1) Bilateral real exchange rates (2) Real exchange rate squared (3) Net foreign assets of the monetary sector/M1 (4) Net foreign assets of the monetary sector/M1 squared (5) Openness (6) Trade concentration (7) Regular executive transfers (8) Irregular executive transfers (9) Months spent in the peg |
| Krugman (1996) | 1988–95, annually, quarterly, and some daily | France, Italy, Spain, Sweden, and the United Kingdom during the 1992–1993 ERM crises | (1) Unemployment rate (2) Output gap (3) Inflation (4) Public debt/GDP |
| Milesi-Ferretti and Razin (1996) | 1970–94, annual | Chile and Mexico have the 4 crises cases; Ireland, Israel, and South Korea do not have crises cases due to policy reversal; and Australia does not have a crisis case or a policy change | (1) Debt service/GDP for GDP growth and changes in the real exchange rate (2) Exports/GDP (3) Real exchange rate versus historical norm (4) Savings/GDP (5) Fiscal stance (6) Fragility of the banking sector (7) Political instability (8) Composition of capital flows |
| Moreno (1995) | 1980–94, monthly and quarterly | Indonesia, Japan, Malaysia, Philippines, Singapore, Korea, and Thailand | (1) Change in bilateral exchange rate (2) Changes in net foreign assets (central bank) (3) Domestic-foreign interest rate differential (4) Exports/imports (5) Output gap All the following are in relation to the United States: (6) Growth of domestic credit/reserve money (7) Growth in M1 (8) Growth in broad money (9) Fiscal deficit/government spending (10) Inflation |
| Otker and Pazarbasioglu (1994) | 1979–93, monthly | Denmark, Ireland, Norway, Spain, and Sweden (the sample covers 15 devaluations and 10 realignments of all central rates) | (1) Domestic credit (2) Real effective exchange rate (3) Trade balance (4) Unemployment rate (5) German price level (6) Output (7) Reserves (8) Central parity (9) Foreign-domestic interest rate differential (10) Position within band |
| Otker and Pazarbasioglu (1995) | 1982–94, monthly | Mexico (during the sample, there are 4 devaluations; 3 increases in the rate of crawl and 2 reductions; and 2 shifts to a more flexible exchange system) | (1) Real exchange rate (2) International reserves (3) Inflation differential with the United States (4) Output growth (5) U.S. interest rates (6) Central bank credit to the banking system (7) Financial sector reform dummy (8) Share of short-term foreign currency debt (9) Fiscal deficit (10) Current account balance |
| Sachs, Tornell, and Velasco (1995) | 1985–95, monthly and annually | 20 emerging market countries | (1) The real exchange rate (2) Credit to the private sector/GDP (3) M2/international reserves (4) Savings/GDP (5) Investment/GDP (6) Capital inflows/GDP (7) Short-term capital inflows/GOP (8) Government consumption/GDP (9) Current account/GDP |
Indicators of External Crises
| Study | Sample and Frequency | Country Coverage | Indicator |
|---|---|---|---|
| Bilson (1979) | 1955–77, annually | 32 countries, with emphasis on Ecuador, Mexico, and Peru | (1) International reserves/base money (2) “Shadow” exchange rate |
| Blanco and Garber (1986) | 1973–81, quarterly | Mexico | (1) Domestic credit growth |
| Calvo and Mendoza (1996) | 1983–94, Monthly and quarterly | Mexico | (1) M2 (in dollars)/reserve (2) The money demand-supply gap |
| Collins (1995) | 1979–91, annually | 18 countries with pegged exchange rates at the beginning of 1979 | (1) International reserves/GDP (2) Real GDP growth (3) Change in the real exchange rate (4) Multiple exchange rate dummy (5) Inflation (6) Current account/GDP (7) Foreign aid |
| Cumby and van Wijenbergen (1989) | 1979–80, monthly | Argentina | (1) Domestic credit growth |
| Dornbusch, Goldfajn, and Valdes (1995) | 1975–95, annual and quarterly | Argentina, Brazil, Chile, Finland, and Mexico | (1) Real exchange rate (2) Real interest rates (3) GDP growth (4) Inflation (5) Fiscal deficit/GDP (6) Credit growth (7) Trade balance/GDP (8) Current account/GDP (9) International reserves (10) Debt/GDP |
| Edin and Vredin (1993) | 1978–89, monthly | Denmark, Finland, Sweden, and Norway | (1) Money (2) Output (3) Foreign interest rate (4) Foreign price level (5) Real exchange rate (6) International reserves/imports (7) Trade balance |
| Edwards (1989) | 1962–82, pooled quarterly data and annually | 24 developing countries | (1) Central bank foreign assets/base money (2) Net foreign assets/M1 (3) Domestic credit to public sector/total credit (4) Bilateral real exchange rate (5) Parallel market premium (6) Growth of credit (7) Growth of credit to the public sector (8) Public-sector credit growth/GDP (9) Fiscal deficit/GDP (10) Current account/GDP (11) Terms of trade (12) Errors and omissions plus short-term capital (13) Exchange controls |
| Eichengreen, Rose, and Wyplosz (1995) | 1959–93, quarterly | 20 industrial countries; 78 crises | The authors mention that many of (1) –(16) are defined with respect to the same variable in Germany, but do not specify which of them. (1) Change in international reserves (2) Real effective exchange rate (3) Credit growth (4) M1 growth (5) Bond yield (6) Interest rates (7) Stock prices (8) Inflation (9) Wage growth (10) GDP growth (11) Unemployment rate (12) Employment growth (13) Fiscal deficit/GDP (14) Current account/GDP (15) Change in exports (16) Change in imports (17) Government victory (18) Government loss (19) Elections (20) Change in government (21) Capital controls (22) Left-wing government (23) New finance minister (24) Past exchange market crisis (25) Past exchange market event |
| Flood and Marion (1995) | 1957–91, monthly | 17 Latin American countries; 80 peg periods of a duration of at least 3 months | (1) Drift of the real exchange rate (2) Variance of the real exchange rate |
| Frankel and Rose (1996) | 1971–92, annually | 105 developing countries | (1) Credit growth (2) Fiscal deficit/GDP (3) Per-capita GDP growth (4) External debt/GDP (5) Reserves/imports (6) Current account/GDP (7) Deviations from PPP in the bilateral real exchange rate (8) OECD GDP growth (9) Foreign interest rate The following variables as a share of total debt: (10) Commercial bank loans (11) Concessional loans (12) Variable rate debt (13) Short-term debt (14) Public-sector debt (15) Multilateral development bank loans (16) Flow of FDI |
| Goldstein (1996) | Annually and monthly | Argentina, Brazil, Chile, and Mexico (other crises are also discussed). | (1) International interest rates (2) Mismatch between the government’s and the banking sector’s short-term assets and liabilities (such as M3/reserves) (3) Current account/GDP, particularly one driven by a fall in savings (4) Boom in bank lending followed by a decline in asset prices (5) Real exchange rate (6) Short-term borrowing (7) Weak banking sector |
| Humberto, Julio, and Herrera (1991) | Monthly | Colombia | (1) Credit growth (2) Parallel market premium |
| Kamin (1988) | 1953–83, annually | 107 cases with devaluations of at least 15 percent with respect to the U.S. dollar | (1) Trade balance/GDP (2) Import growth (3) Export growth (4) Capital flows/GDP (5) Changes in reserves (6) Inflation (7) The real exchange rate (8) Real GDP growth (9) Change in export prices |
| Kaminsky and Leiderman (1995) | 1970–95, monthly | 20 countries (5 industrial and 15 developing) | (1) Export growth (2) Import growth (3) Bilateral real exchange rate-deviation from trend (4) Terms of trade changes (5) Changes in reserves (6) Money demand/supply gap (7) Changes in bank deposits (8) Real interest rates (9) Lending-deposit spread (10) Domestic-foreign real interest rate differential (11) M2 money multiplier (12) M2/international reserves (13) Growth in domestic credit/GDP (14) Changes in stock prices (15) Output growth (16) Financial liberalization (17) Banking crises |
| Klein and Marion (1994) | 1957–91, monthly | 87 peg episodes, as in Flood and Marion (1995) | (1) Bilateral real exchange rates (2) Real exchange rate squared (3) Net foreign assets of the monetary sector/M1 (4) Net foreign assets of the monetary sector/M1 squared (5) Openness (6) Trade concentration (7) Regular executive transfers (8) Irregular executive transfers (9) Months spent in the peg |
| Krugman (1996) | 1988–95, annually, quarterly, and some daily | France, Italy, Spain, Sweden, and the United Kingdom during the 1992–1993 ERM crises | (1) Unemployment rate (2) Output gap (3) Inflation (4) Public debt/GDP |
| Milesi-Ferretti and Razin (1996) | 1970–94, annual | Chile and Mexico have the 4 crises cases; Ireland, Israel, and South Korea do not have crises cases due to policy reversal; and Australia does not have a crisis case or a policy change | (1) Debt service/GDP for GDP growth and changes in the real exchange rate (2) Exports/GDP (3) Real exchange rate versus historical norm (4) Savings/GDP (5) Fiscal stance (6) Fragility of the banking sector (7) Political instability (8) Composition of capital flows |
| Moreno (1995) | 1980–94, monthly and quarterly | Indonesia, Japan, Malaysia, Philippines, Singapore, Korea, and Thailand | (1) Change in bilateral exchange rate (2) Changes in net foreign assets (central bank) (3) Domestic-foreign interest rate differential (4) Exports/imports (5) Output gap All the following are in relation to the United States: (6) Growth of domestic credit/reserve money (7) Growth in M1 (8) Growth in broad money (9) Fiscal deficit/government spending (10) Inflation |
| Otker and Pazarbasioglu (1994) | 1979–93, monthly | Denmark, Ireland, Norway, Spain, and Sweden (the sample covers 15 devaluations and 10 realignments of all central rates) | (1) Domestic credit (2) Real effective exchange rate (3) Trade balance (4) Unemployment rate (5) German price level (6) Output (7) Reserves (8) Central parity (9) Foreign-domestic interest rate differential (10) Position within band |
| Otker and Pazarbasioglu (1995) | 1982–94, monthly | Mexico (during the sample, there are 4 devaluations; 3 increases in the rate of crawl and 2 reductions; and 2 shifts to a more flexible exchange system) | (1) Real exchange rate (2) International reserves (3) Inflation differential with the United States (4) Output growth (5) U.S. interest rates (6) Central bank credit to the banking system (7) Financial sector reform dummy (8) Share of short-term foreign currency debt (9) Fiscal deficit (10) Current account balance |
| Sachs, Tornell, and Velasco (1995) | 1985–95, monthly and annually | 20 emerging market countries | (1) The real exchange rate (2) Credit to the private sector/GDP (3) M2/international reserves (4) Savings/GDP (5) Investment/GDP (6) Capital inflows/GDP (7) Short-term capital inflows/GOP (8) Government consumption/GDP (9) Current account/GDP |
Bibliography
Barro, Robert J., 1989, “The Ricardian Approach to Budget Deficits,” Journal of Economic Perspectives, Vol. 3 (Spring), pp. 37–54.
Barro, Robert J., 1997, Macroeconomics (Cambridge, Massachusetts: MIT Press).
Bernanke, Ben S., 1983, “Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression,” American Economic Review, Vol. 73 (June), pp. 257–76.
Bernheim, Douglas, 1987, “Ricardian Equivalence: An Evaluation of Theory and Evidence,” NBER Macroeconomics Annual, Vol. 2, pp. 263–304.
Blanco, Herminio, and Peter Garber, 1986, “Recurrent Devaluations and Speculative Attacks on the Mexican Peso,” Journal of Political Economy, Vol. 94 (February), pp. 148–66.
Calvo, Guillermo A., 1987, “On the Costs of Temporary Policy,” Journal of Development Economics, Vol. 27, pp. 147–69.
Calvo, Guillermo A., and Enrique Mendoza, 1996, “Mexico’s Balance-of-Payments Crisis; A Chronicle of a Death Foretold,” journal of International Economics, Vol. 41 (November), pp. 235–64.
Calvo, Sara, and Carmen Reinhart, 1996, “Capital Flows to Latin America: Is There Evidence of Contagion Effects?” in Private Capital flows to Emerging Markets, ed. by Morris Goldstein (Washington: Institute for International Economics).
Claessens, Stijn, Michael Dooley, and A. Warner, 1995, “Portfolio Capital Rows: Hot or Cold?” World Bank Economic Review, Vol. 9, pp. 153–74.
Cole, Harold, and Timothy Kehoe, 1996, “A Self-Fulfilling Model of Mexico’s 1994–95 Debt Crises,” Journal of International Economics, Vol. 41 (November), pp. 309–30.
Corden, W. Max, 1977, Inflation, Exchange Rates, and the International System (Oxford: Oxford University Press).
Corsetti, Giancarlo, and Nouriel Roubini, 1991, “Fiscal Deficits, Public Debt, and Government Solvency: Evidence from OECD Countries,” Journal of the Japanese and International Economies, Vol. 5 (December), pp. 354–80.
Cumby, Robert E., and Sweder van Wijnbergen, 1989, “Financial Policy and Speculative Runs with a Crawling Peg: Argentina 1979–81,” Journal of International Economics, Vol. 17 (August), pp. 111–27.
Deppler, Michael, and John Williamson, 1987, “Capital Flight: Concepts, Measurement, and Issues,” in Staff Studies for the World Economic Outlook (Washington: International Monetary Fund).
Dooley, Michael, 1996, “A Survey of Literature on Controls over International Capital Transactions,” Staff Papers, International Monetary Fund, Vol. 43 (December), pp. 639–87.
Dornbusch, Rudiger, Ilan Goldfajn, and Rodrigo Valdes, 1995, “Currency Crises and Collapses,” Brookings Papers on Economic Activity: 1, pp. 253–315.
Eaton, Jonathan, and Mark Gersovitz, 1981, “Debt with Potential Repudiation: Theory and Estimation,” Review of Economic Studies, Vol. 48 (April), pp. 289–309.
Edwards, Sebastian, 1987, “Exchange Controls, Devaluations and Real Exchange Rates: The Latin American Experience,” NBER Working Paper 2348 (Cambridge, Massachusetts: National Bureau of Economic Research).
Edwards, Sebastian, 1995, “Why Are Saving Rates So Different Across Countries? An International Comparative Analysis,” NBER Working Paper 5097 (Cambridge, Massachusetts: National Bureau of Economic Research).
Frenkel, Jacob A., Assaf Razin, and Chi-Wa Yuen, 1996, Fiscal Policy and Growth inthe World Economy (Cambridge, Massachusetts: MIT Press).
Garber, Peter, and Lars E.O. Svensson, 1995, “The Operation and Collapse of Fixed Exchange Rate Regimes,” Handbook of International Economics, Vol. 3, ed. by Gene Grossman and Kenneth Rogoff (Amsterdam and New York: North-Holland).
Ghosh, Atish, and Jonathan D. Ostry, 1995, “The Current Account in Developing Countries: A Perspective from the Consumption Smoothing Approach,” World Bank Economic Review, Vol. 9 (May), pp. 305–33.
Goldstein, Morris, 1996, “Presumptive Indicators/Early Warning Signals of Vulnerability to Financial Crises in Emerging Market Economies” (unpublished; Washington: Institute for International Economics).
International Monetary Fund, 1995, International Capital Markets: Developments, Prospects, and Policy Issues, World Economic and Financial Surveys (Washington).
International Monetary Fund, 1998, International Capital Markets: Developments, Prospects, and Policy Issues, World Economic and Financial Surveys (Washington).
Kaminsky, G., J. Saul Lizondo, and Carmen Reinhart, 1997, “Leading Indicators of Currency Crises,” IMF Working Paper 97/79 (Washington: International Monetary Fund).
Kiyotaki, Nobuhiro, and John Moore, 1993, “Credit Cycles” (unpublished; University of Minnesota).
Krugman, Paul, 1979, “A Model of Balance of Payments Crises,” Journal of Money, Credit and Banking, Vol. 11 (August), pp. 311–25.
Leiderman, Leonardo, and Assaf Razin, 1991, “Determinants of External Imbalances: The Role of Taxes, Government Spending, and Productivity,” Journal of the Japanese and International Economies, Vol. 5 (December), pp. 421–50.
Lindert, Peter, and Peter Morton, 1989, “How Sovereign Debt Has Worked,” in Developing Country Debt and Economic Performance, Vol. 1, ed. by Jeffrey D., Sachs (Chicago: University of Chicago Press).
McKinnon, Ronald I., and Huw Pill, 1996, “Credible Liberalizations and International Capital Flows: The Overborrowing Syndrome,” in Financial Deregulation and Integration in East Asia, ed. by T. Ito and Anne Krueger (Chicago: University of Chicago Press), pp. 7–42.
McKinnon, Ronald I., and Huw Pill, 1997, “Credible Economic Liberalizations and Overborrowing,” American Economic Review, Papers and Proceedings, pp. 189–93.
Milesi-Ferretti, Gian Maria, and Assaf Razin, 1996, Current Account Sustainability, Princeton Essays in International Finance, No. 81 (October).
Mussa, Michael L., 1986, “The Adjustment Process and Timing of a Trade Liberalization,” in Economic Liberalization in Developing Countries, ed. by A. Choksi and D. Papageorgiou (Oxford: Blackwell).
Obstfeld, Maurice, 1996, “Models of Currency Crises with Self-Fulfilling Features,” European Economic Review, Vol. 40 (April), pp. 1037–47.
Obstfeld, Maurice, and Kenneth Rogoff, 1996, Foundations of International Macroeconomics (Cambridge, Massachusetts: MIT Press).
Ostry, Jonathan D., 1997, “Current Account Imbalances in ASEAN Countries: Are They a Problem?” IMF Working Paper 97/51 (Washington: International Monetary Fund).
Razin, Assaf, 1997, “Current Account Reversals, Exchange Rate Misalignments, and the Government Budget,” notes presented at an IMF Institute seminar (August).
Reisen, Helmut, 1997, “The Limits of Foreign Savings,” in Promoting Savings in Latin America, ed. by Ricardo Hausmann and Helmut Reisen (Washington: Inter-American Development Bank).
Sachs, Jeffrey D., 1982, “The Current Account in the Macroeconomic Adjustment Process,” Scandinavian Journal of Economics, Vol. 84, No. 2, pp. 147–59.
Velasco, Andres, 1991, “Liberalization, Crisis, Intervention: The Chilean Financial System, 1975–85,” in Banking Crisis: Cases and issues, ed. by Tomás Baliño and V., Sundararajan (Washington: International Monetary Fund), pp. 113–74.
Wong, Chomg-Huey, and Luis Carranza, 1999, “Policy Responses to External Imbalances in Emerging Market Economies—Further Empirical Results,” Staff Papers, International Monetary Fund, Vol. 46 (June), pp. 225–37.
The acquisition of net financial assets implies either that gross financial assets are being acquired or that gross claims of foreigners are being extinguished.
Barro (1997) provides an extensive explanation of this hypothesis.
The autarky interest rate is the rate at which CA is equal to zero—that is, national saving is equal to investment.
A complete explanation is provided in the next section.
This crisis is not the only example of national bankruptcy. Similar experiences in which countries defaulted on their external debt can be found since the establishment of the international capital market, Lindert and Morton (1989) provide a more in-depth discussion.
Calvo and Mendoza (1996) and Cole and Kehoe (1996) contain discussions related to the Mexican crisis.
This section is based on Milesi-Ferretti and Razin (1996) and McKinnon and Pill (1996 and 1997).
Kaminsky, Lizondo, and Reinhart (1997) found that the best indicators for anticipating current account sustainability problems include the behavior of the real exchange rate, domestic credit, international reserves, credit to the public sector, and domestic inflation. The appendix, from Kaminsky, Lizondo, and Reinhart (1997), summarizes some of the most relevant empirical research.
This condition does not prevail, however, when the economy is experiencing major price distortions. In that case, investment would probably be unproductive, and external vulnerability would increase.
For example, two countries with the same degree of openness and the same export-to-GDP ratio may have very different vulnerabilities to foreign shocks if one is exporting only primary products and the second only high-end manufactured goods. The first country would be more vulnerable because prices for commodities are more volatile than those for manufactured goods.
This statement could be controversial. For example, Claessens and others (1995) did not find any evidence that long-term flows are less volatile than short-term flows.
Although this proposition has been the predominant one in the literature, it does not apply when the time profile of public expenditure changes. Since the taxation time profile remains unchanged, private saving also remains unchanged. Thus, a fiscal deficit leads to a current account deficit of the same magnitude.
Empirical evidence on the Ricardian equivalence principle is inconclusive (see, for example, Bernheim, 1987 and Barro, 1989). Edwards (1995) estimated an average Ricardian offset coefficient of 0.5 for developing countries.
Dollarization is fairly common in several countries. For example, in Peru dollar-denominated deposits represented about 80 percent of total deposits in the mid-1990s.
Overvaluing imports and undervaluing exports are ready ways to avoid these types of controls. Deppler and Williamson (1987) report that, from 1975 to 1985, cumulative capital flight from Latin America was $106.6 billion. To put this in perspective, total Latin American external debt in 1987 was estimated at $300 billion.
This section is based on Wong and Carranza (1999).
The saving-investment framework presented earlier shows this relationship clearly.
With a gradual reduction in the depreciation rate to zero, inflation rates were supposed to converge to international levels.
The Mexican crisis has been studied extensively. See, for example, IMF (1995); Dornbusch, Goldfajn, and Valdes (1995); and Calvo and Mendoza (1996).
IMF (1995) provides details of the events and actions.
This section is based on IMF (1998).
