Christian Keller, Christoph Rosenberg, Nouriel Roubini, and Brad Setser
The Depth of the capital account crises of the 1990s and the devastation they left in their wake shattered any complacency economic and financial experts may have felt about their ability to accurately assess the financial health of a country relying solely on the traditional analysis of flow variables, such as annual GDP, the current account, and fiscal balances. Many observers realized that signs of impending trouble might have been spotted had they looked more closely at countries’ balance sheets and, specifically, paid more attention to mismatches between the stock of a country’s assets and the stock of its liabilities—that is, stock imbalances. Analysis of flows is, of course, still vital. But those seeking to prevent crises or to react to them more effectively need to ask questions about the health of a country’s balance sheet. For example, does short-term foreign-currency-denominated debt exceed foreign currency reserves? Would a large outstanding stock of debt denominated in foreign currency make it more difficult to correct a flow imbalance (such as a current account deficit) without creating a deep crisis?
The balance sheet approach to crisis prevention and resolution begins with a look at a country’s consolidated external balance sheet—the external debts that the country’s government, its banks, and its firms have relative to their external assets (notably liquid external reserves). But close attention must also be paid to the balance sheets of individual sectors, because mismatches at the sectoral level might not show up on the consolidated balance sheet, yet could trigger a financial crisis just the same. The key sectors include the government sector (for the sake of simplicity, we include the central bank in this sector), the private financial sector (mainly banks), and the nonfinancial sector (corporations and households). The sectoral balance sheets are often linked—that is, one sector’s debt may be another’s asset, in which case, if the first sector has trouble servicing its debt, the second sector’s assets deteriorate and it may, in turn, have difficulty repaying its creditors.
Analysis of balance sheet vulnerabilities is most useful if it is done in time to allow policymakers to identify and correct weaknesses before they contribute to financial difficulties. This puts a premium on timely information on stock variables, which is often hard to come by (see Box 1).
Box 1Not enough information
Countries have not routinely gathered balance sheet information, primarily because they lack the resources to do so—one reason the IMF is placing a high priority on helping countries develop these data. In the meantime, partial information is available from several sources:
• Data collected by national authorities (available on the websites of many finance ministries and central banks); for subscribers to the IMF’s Special Data Dissemination Standard (SDDS), this includes detailed data on reserves and foreign currency liquidity.
• Various databases and publications, including the IMF’s monthly International Financial Statistics (IFS) and the Balance of Payments Yearbook (with International Investment Position data).
• Quarterly statistics, from the perspective of creditors, on the assets and liabilities of banks in the 28 countries and territories that report to the Bank for International Settlements (BIS), available as part of the Joint BIS-IMF-OECD-World Bank debt statistics.
• Annual debt tables published by the World Bank (Global Development Finance).
Together, these data provide reasonable coverage of the public sector’s liabilities and liquid assets and of the country’s aggregate external liabilities. Nonetheless, detailed information on short-term external liabilities, especially those of the nonofficial sectors, and on the domestic claims of the official sector, can be difficult to find. Data on overall assets and liabilities, particularly breakdowns by currency of stocks and flows, in the nonfinancial private sector tend to be scarce.
Anatomy of balance sheet risks
Four different kinds of balance sheet risks could impair an emerging market country’s ability to service its debt.
Maturity mismatches between short-term liabilities and longer-term liquid assets expose borrowers to rollover risk (that they will be unable to refinance maturing debts) and interest rate risk (that interest rates on outstanding debt will rise, particularly if longer-maturity liabilities carry floating interest rates). Financial entities that borrow short term to invest in long-term debt with fixed interest rates are exposed to the risk of rate increases that may reduce the market value of their long-term assets even as they have to pay more on their short-term liabilities. Maturity mismatches in foreign currency can arise even when total foreign currency liabilities match total foreign currency assets if borrowers do not have enough liquid foreign currency assets to cover short-term foreign currency debt.
Currency mismatches arise when borrowers’ liabilities are denominated in a foreign currency but their assets are in domestic currency. In the event of a devaluation, these borrowers will have trouble paying their creditors. Many of the crisis countries had some kind of exchange rate peg, which often encouraged borrowers and lenders alike to ignore the very real currency risk.
Capital structure mismatches may occur when a firm or a country relies on debt rather than equity to finance investment. Equity provides a buffer during hard times, because dividends drop along with earnings, whereas debt payments remain constant. Before the Asian crisis of 1997–98, the Korean government severely restricted foreign direct investment, so most capital inflows were in the form of debt, while Thailand’s tax regime favored corporate debt over equity. As a result of financing current account deficits with debt, particularly short-term debt, rather than with direct investment, these countries accumulated a large external debt stock that increased their vulnerability to crises.
Solvency risk arises when an entity’s liabilities exceed its assets. These assets include future net income—for example, the gap between the government’s future revenue and its future expenditures (excluding interest), as well as financial assets like reserves. Maturity, currency, and capital structure mismatches can all increase the risk that a negative shock to a country—for example, a sudden deterioration in its terms of trade, bad political or economic news, or a crisis in neighboring countries—will drive large parts of one or more sectors into insolvency. Insolvency can also occur if an entity borrows too much and invests in poor-quality assets. To assess a government’s solvency, sovereign debt is often measured against GDP or revenues, and a country’s overall solvency is measured in relation to the ratio of total debt to GDP or to exports. Such measures are most helpful when used in conjunction with other measures of risk exposure. Two countries with identical debt-to-GDP ratios may not be equally vulnerable to a solvency crisis: a country that borrows exclusively in foreign currency is likely to be more at risk than a country that borrows exclusively in domestic currency.
Spotting these types of balance sheet risks may not enable us to predict the exact timing of a crisis—weaknesses in balance sheets can linger for years—but it can contribute to our understanding of the dynamics of crises that do occur and thus inform measures to prevent other such crises (see Box 2). After all, the balance sheet mismatches at the heart of the 1990s capital account crises did not arise by accident. Persistent deficits eventually became stock problems. Moreover, governments and other borrowers that were having trouble financing flow imbalances often took on more currency and maturity risk to obtain needed credit, further weakening balance sheets.
A government that manages its own balance sheet prudently—by maintaining a healthy cushion of reserves, not borrowing too much, and avoiding currency and maturity risk—contributes to the health of the country’s aggregate balance sheet. Claims on the government are usually the largest single financial asset in an economy. Long-term domestic-currency-denominated debt offers the best match for the government’s key asset (its capacity to run future primary surpluses) which is illiquid, long term, and, typically, a revenue stream in domestic currency. Countries that cannot borrow long term in domestic currency are thus unable to sustain as high a debt-to-GDP ratio as countries that can.
There is a growing consensus on the policies that can create incentives for prudent financial behavior by the private sector:
• Flexible exchange rate regimes can discourage borrowers from accumulating too much exposure to currency risks and ease adjustment to shocks, although currency risk is unlikely to disappear altogether, given foreign and domestic investors’ reluctance to take on local currency exposure.
• Domestic markets for equity, long-term debt in local currency, and financial products that enable firms to hedge their risks can make it easier for the private sector to raise the right kind of financing. Tax or regulatory distortions that favor debt—particularly foreign-currency-denominated debt—over equity should be eliminated.
• Prudential regulation of banks needs to take into account both direct and indirect exposure to foreign currency risk. Measures to discourage foreign-currency-denominated deposits and short-term foreign-currency-denominated interbank borrowing—for example, higher reserve requirements—may be worth considering. Policies that discourage short-term external borrowing by firms as well as banks—such as Chile’s encaje—may also warrant consideration in some circumstances.
When crisis strikes
But even with improved crisis-prevention techniques, crises will still occur. When they do, policymakers need to move quickly to keep crises from deepening. The balance sheet approach can help policymakers evaluate some inherently difficult choices.
• A government can sell foreign exchange reserves to help banks and firms hedge their foreign exchange exposure in an attempt to prevent a private sector crisis. But it thereby risks ending up with insufficient reserves to cover its own short-term foreign currency liabilities, particularly if it seeks to defend an overvalued nominal exchange rate. Even when the government’s balance sheet is strong enough to make a rollover crisis unlikely, it needs to take care to limit the creation of expectations that could encourage private firms to incur more foreign currency debt.
• A government may also have to choose between raising interest rates to halt capital outflows and check currency depreciation and keeping interest rates constant so as not to dampen economic activity. If the country’s stock of short-term debt denominated in domestic currency is large, raising interest rates may do more harm than good to firms and households. If the stock of foreign-currency-denominated debt is large, the country might be better off increasing interest rates to reduce the risk of further depreciation. Policymakers typically need to seek a balance between monetary tightening and exchange rate adjustment.
A case for external financing
When is external financing necessary to help a country head off a pending crisis or keep a crisis that has already occurred from snowballing out of control? And how much external financing is needed to provide the country with a reasonable chance? The balance sheet approach can shed light on these questions, though it does not provide a mechanistic way to determine the right amount of official support.
Obtaining financing in the early stages of a crisis, before difficulties in one sector have spilled over into others, can prevent a generalized loss of confidence, averting a surge in demand for external assets and a need for external financing too large to be satisfied by the official sector. For example, the international community’s intervention in Mexico during the 1994–95 crisis kept a government rollover crisis from turning into an even deeper crisis and allowed a relatively quick rebound in output, with an improved balance in the current account. However, not all financial crises in emerging market economies call for official intervention. Some governments have enough resources of their own to avoid a cascading crisis. In certain crises, additional external loans may undermine sustainability, and debt restructuring may be better for a country than taking on new debt.
Box 2Unraveling the East Asian crisis
The balance sheet approach clarifies what happened in East Asia in 1997–98. Capital flows to East Asian countries stopped suddenly, making it impossible for countries to maintain their exchange rate regimes and leading to a sharp depreciation of their currencies. In Thailand, a deterioration in the corporate sector’s financial health and export performance made it more difficult to continue to finance a current account deficit and led currency speculators to bet against a pegged exchange rate. In Thailand and other countries, doubts about the ability of banks and firms to roll over short-term external debt helped trigger a capital flow reversal, and the need to buy foreign currency to repay existing debts was a key source of pressure on the currency.
The resulting currency depreciation, in turn, undermined the corporate sector’s financial health. When firms could not service their foreign currency debts, the banks that had borrowed abroad in foreign currency to finance these loans were hurt. Output declines in several Asian crisis countries were much worse than expected. Although some firms enjoyed increased export opportunities after currencies tumbled, these gains were initially more than offset by an increase in the corporate sector’s real debt burden and banking sector weaknesses, both of which battered corporate investment.
The crisis thus fed on itself, as weaknesses in one sector led to weaknesses in other sectors. Plunging demand for a country’s assets resulted in massive capital outflows, depletion of reserves, and a sharp currency depreciation. Eventually, the sharp adjustment in the current account and strengthened policies led to the restoration of confidence, currency stabilization, and the rebuilding of reserves in most countries.
The IMF’s ability to disburse large amounts of foreign exchange quickly—and to discern when such financing is necessary and which measures are necessary to ensure sustainability—can be crucial for effective crisis resolution. The IMF can strengthen the monetary authorities’ balance sheet, helping to restore confidence in the government’s own balance sheet and leaving the authorities in a better position to help rebuild confidence in other sectors. By providing the reserves needed to tide a country over a crisis stemming from a maturity mismatch, the IMF can help it avoid the need to create money, a potentially destabilizing suspension of payments and debt restructuring, and the draconian external adjustment required to generate reserves immediately from current export earnings. Such assistance should be combined with efforts by the country to help itself—such as letting the exchange rate float to reduce current account imbalances or tightening its fiscal belt to reduce financing needs.
“The balance sheet approach can inform judgments about the scale of IMF support needed, but it does not provide a formula for calculating the exact sums.”
It is worth distinguishing between crises that arise because of mismatches on the private sector’s balance sheet and those that arise because of mismatches on the government’s balance sheet. The best approach to most private financial problems is to allow borrowers and creditors to negotiate a workout or to leave matters to the bankruptcy courts. Intervention by the national government is warranted only when there is a large risk of spillovers to the broader economy. The government can finance most such interventions itself, but, in some cases, it may need external official financing to build up its foreign currency reserves so that it can intervene effectively. Governments should support only the banks and firms that are structurally sound, closing unsound institutions immediately while strengthening incentives for prudent financial management in the private sector.
If the government itself has a balance sheet problem, the risk of a broader crisis is more acute. The government’s debt is often one of the largest—if not the largest—financial asset held by domestic banks and other financial institutions, so a government financing crisis could snowball into a banking crisis. There are also limits on the ability of the government to draw on the strength of the private sector’s balance sheet to correct its own financial problems. Often, the official sector—including the IMF—is the only viable source of foreign currency liquidity.
A key finding of balance sheet analysis is that demand for foreign currency reserves during a crisis is proportionate to outstanding claims on short-term foreign currency liquidity. A country without a current account deficit may still have trouble rolling over its short-term debt. Demand for foreign exchange to repay external debt surged during all of the recent crises, far surpassing official foreign currency reserves. In many cases, foreign exchange may also be needed to cover domestic debts, notably foreign-currency-denominated bank deposits held by the country’s own residents. Information about the outstanding stock of claims helps determine only the maximum amount that might be needed to repay existing debts, not the actual amount, which is determined by how creditors and investors behave—for example, whether domestic and foreign banks are willing to roll over their short-term claims.
However, it is unrealistic—and, in most cases, undesirable—to expect official financing to cover all stock imbalances. The balance sheet approach can inform judgments about the scale of IMF support needed, but it does not provide a formula for calculating the exact sums or identifying whether a crisis is best dealt with through debt restructuring or official lending. It is also important to remember that only a limited set of balance sheet problems can be effectively addressed by nonconcessional official financing. Official lending increases the supply of foreign currency available in the short run and can thus help address a maturity mismatch. But it cannot reduce the currency mismatch on a country’s consolidated balance sheet, nor does a loan from a preferred creditor like the IMF improve a country’s capital structure or eliminate solvency risk, though it can buy the country time to make policy adjustments to rebuild its finances and regain market confidence.
Christoph Rosenberg is a Deputy Division Chief, Christian Keller is an Economist, and Brad Setser is a Visiting Scholar in the IMF’s Policy Development and Review Department. Nouriel Roubini, Associate Professor of Economics and International Business at New York University, was a Visiting Scholar in the IMF’s Policy Development and Review Department and Research Department, and contributed to the study on which this article is based.