A Balance Sheet Approach to Financial Crisis
December 1, 2002
Data Provision to the Fund for Surveillance Purposes
April 26, 2002
External Debt Statistics Guide for Compilers and Users
June 25, 2003
Original Sin-Balance Sheet Crises, and the Roles of International Lending
December 1, 2002
The Balance Sheet Approach and its Applications at the Fund
Prepared by the Policy Development and Review Department
June 30, 2003
This note summarizes the main conclusions from the working paper "A Balance Sheet Approach to Financial Crisis" (WP/02/210). It describes the conceptual framework of the balance sheet approach and how it has been reflected in the Fund's work. Based on this, the note explores how the use can be further integrated into the existing analysis, while also pointing out the challenges this involves from an analytical and operational perspective.
1. The Executive Board has on several occasions asked staff to examine the causes and implications of capital account crises. Recent recommendations by the Independent Evaluation Office also emphasize the need for Fund staff to analyze the consequences of potential balance sheet shocks in the context of surveillance as well as program design.1 Indeed, the traditional financial programming approach may be insufficient in explaining some of the dynamics underlying modern-day capital account crisis. Its flow-based analysis focuses on the gradual build-up of unsustainable fiscal and current account positions. The balance sheet approach focuses instead on shocks to stocks of assets and liabilities, which can trigger large adjustments in (capital) flows. Such an approach can therefore be a useful complement to the traditional flow-analysis. Indeed, academics and policy makers have been paying increasing attention to its further development as a result of the capital account crises of the 1990s (see Box 1). The Fund has been applying the insights from the balance sheet approach for some time and many of its elements have entered the work on fiscal and external sustainability, liquidity and debt management, FSAPs etc.
|2. The focus on balance sheets is of particular relevance for emerging market economies. While these countries have gained access to capital markets and benefit from their ability to mobilize foreign savings, the related capital flows are often subject to great volatility. Importantly, emerging market countries' foreign borrowing typically is foreign currency denominated and at shorter maturities. Their local markets are typically thin and the institutional capacity to manage the associated risks remains limited. Although significant differences exist among emerging markets, in general their vulnerability to sudden shocks tends to be higher than that in mature markets.
3. Capital account crises typically occur as creditors suddenly lose confidence in the health of the balance sheets of one of the country's main sectors--the banking system, the corporate sector, or the government. This confidence loss can prompt sudden and large-scale portfolio adjustments such as massive withdrawals of bank deposits, panic sales of securities or abrupt halts of debt roll-overs. As the exchange rate, interest rates, and other asset prices adjust, the balance sheets of an entire sector--which may be largely solvent in absence of these adverse events--can sharply deteriorate. In an integrated financial system and with an open capital account, concerns about asset quality on domestic balance sheets can provoke creditors to shift towards (safer) foreign assets. This will often result in capital outflows, exerting further pressure on the exchange rate or official reserves and ultimately resulting in a balance of payment crisis.
4. The initial shock to a balance sheet may take various forms, its impact depending on the existing mismatches on the balance sheet. Several patterns can be detected in capital account crises of the last decade and have been subject to recent research (Box 1). A currency mismatch--a predominance of assets denominated in domestic currency over liabilities denominated in foreign currency--leaves a balance sheet vulnerable to a depreciation of the domestic currency (exchange rate shock). A maturity mismatch--long-term, illiquid assets against short-term liabilities exposes a balance sheet to risks related both to rollover and to interest rates: If liquid assets do not cover maturing debts, a balance sheet is vulnerable to a rollover risk, because emerging economies can find themselves shut out of capital markets altogether. Furthermore, a sharp increase in interest rates (interest rate shock) can dramatically increase the cost of rolling over short-term liabilities, leading to a rapid increase in debt service. Other potential shocks include any sharp drop in the price of assets such as government bonds, real estate, or equities (market risk), to which the balance sheets of a certain sector may be particularly exposed. Any of the above shocks can bring about a deterioration in the value of a sector's assets compared to its liabilities and hence to a reduction of its net worth; in the extreme that net worth may turn negative and the sector becomes insolvent. The greater a balance sheet's capital structure mismatch--too much debt relative to equity--the smaller its buffer against such an event.
5. Strong balance sheets protect against real as well as financial shocks. Many shocks originate in the real economy. For example, a collapse in the demand for a country's main commodity or other major export product will lead to a deterioration in corporate earnings or government revenues. This will prompt a re-assessment of these sectors' sustainability and thus a re-evaluation of the market value of their debt and other assets. Such real shocks are particularly dangerous when combined with financial vulnerabilities, as a real shock is often correlated with reduced market access. The impact of the commodity price shock of 1998, for example, was magnified in countries such as Russia, where maturity mismatches left balance sheets vulnerable to rollover risk and interest rate shocks.
6. Maturity and currency mismatches are sometimes hidden in indexed or floating rate debt instruments, making them less evident. In some emerging market economies (e.g., Brazil) liabilities may be formally denominated in local currency, but indexed to the exchange rate. Similarly, the nominal maturity of an asset may be long but the interest rate it bears may be floating. Such indexation creates the same mismatches as if the debt were denominated in foreign currency or as if the maturity were as short as the frequency of the interest rate adjustments.
7. Off-balance sheet activities can substantially alter the overall risk exposure. Financial transactions such as forwards, futures, swaps and other derivatives are not recorded on a balance sheet, but imply predetermined or contingent future flows that will eventually affect it. Such transactions can be used to effectively reduce the risk created by balance sheet mismatches: for example, corporations with a foreign currency mismatch enter into foreign currency forward contracts to reduce their exposure to exchange rate risk. By the same token, off-balance sheet activities increase the risk exposure, when they are not used to hedge (taking a position that is negatively correlated to an existing balance sheet risk) but to speculate or, in the particular case of monetary authorities, to support the domestic currency against market pressures.
8. Balance sheet problems in one sector can spill over into other sectors, often snowballing in the process. Balance sheet crises can originate in the corporate sector (as in some Asian countries in 1997-98) or the fiscal sector (as in Russia 1998, Turkey 2001 and recently in some Latin American countries), with the banking sector playing a key transmission role in all these episodes. If a shock causes the corporate sector or the government to be unable to meet its liabilities, another sector, typically the banking sector, loses its claims. By the same token, if banks tighten their lending to prevent their asset portfolio from deteriorating, this further complicates the situation of a corporate sector or a government in dire need for fresh financing or debt roll-overs. Due to these repercussions, balance sheet problems tend to snowball as they spill from one sector into another.
9. A loss of confidence in the banking system often not only triggers a run on deposits but also a flight from the currency. The authorities may expand liquidity or lower interest rates to support the ailing banking system, while the depositors may seek to protect their savings by switching into foreign currency assets. Both create pressure on the exchange rate. A depreciating exchange rate, however, further weakens the asset side of a banking sector that has a currency mismatch on its balance sheet. Thus, banking and currency crisis may reinforce each other, creating the "twin crises" frequently observed in past cases.
10. Although a crisis may not originate in the government's balance sheet, it is likely to spread to it, partly as a result of contingent liabilities. For instance, the banking system's integrity is often explicitly or implicitly guaranteed by the government. In the event of a crisis, such contingent (off-balance sheet) commitments become definite (balance sheet) liabilities, further adding to the deterioration of the government's balance sheet and the fiscal pressures that are created by the crisis' macroeconomic disruptions. Contingent commitments may even exist to bail out corporations, especially when governments are involved in their investment and borrowing decisions. Furthermore, monetary authorities may be engaged in forward contracts and other off-balance sheet transactions, which can entail large contingent drains on their foreign currency assets.
11. The interaction between financial balance sheets also magnifies the negative impact of a shock on real output levels. Autonomous investment cuts by corporations to restore the financial health of their balance sheets are usually compounded by a forced reduction in credit from distressed banks and lower consumption by households that experience a negative wealth effect. All this may accumulate to a sharp decline in aggregate demand.
II. Policy implications for crisis prevention and crisis response
12. As a first line of defense, the private sector is responsible for protecting its balance sheets against shocks. As explained above, equity on a balance sheet constitutes a general buffer against any type of shock. Similarly, the availability of contingent credit lines or liquid assets can provide a cushion against the liquidity problems that may arise as a result of a shock. In addition, hedging strategies can help to limit the specific risks stemming from currency and maturity mismatches.
13. Public policies can promote buffering and hedging on private balance sheets. While the optimal degree of balance sheet protection depends on its costs and benefits, such a cost-benefit analysis is often distorted by economic policies that encourage inadequate buffering or hedging: for example, the implicit promise of government bail-outs, a tax system that favors debt over equity or an exchange rate peg that creates the illusion of currency stability. Furthermore, the policies that actively promote balance sheet protection may be insufficient: for example, financial system supervisors may not enforce capital requirements or properly assess the aggregate risks of maturity and currency mismatches in the banking system. Governments can lessen both types of shortcomings, by reducing involvement in private sector investments, moving towards more flexible exchange rate regimes, eliminating tax distortions, and, at the same time, strengthening supervision, prudential regulation and data dissemination.
14. Governments can also help to create the conditions that allow private entities to finance themselves in ways that avoid large balance sheet mismatches and that encourage the development of markets for hedging instruments. Private agents often have to assume exchange rate and maturity risk because in many emerging economies there is no market to raise long-term investment capital in domestic currency. Through a combination of sound macroeconomic policy, appropriate financial regulations, and strategically planned public debt issues, governments can promote the development of domestic markets for equity and local currency long-term debt, thus helping to reduce reliance on short-term and foreign currency debt financing. In addition, they can try to promote the development of markets for instruments that allow private agents to hedge their existing balance sheet risks more effectively. If a government does this by supplying the needed financial instruments itself, for example, by issuing exchange rate-linked securities, it must take care that the exchange rate risk it thereby assumes does not overwhelm its own balance sheet.
15. In general, governments need to address their own balance sheet risks by keeping debt at prudent levels, building more insurance against shocks into their debt structure, and maintaining adequate reserve levels. Given that a government's main asset is its power to generate a future stream of tax revenue from residents, its liabilities would ideally be denominated mainly in domestic currency and issued at long maturities. The shift from foreign currency debt--often incurred for reasons explained above--towards long-term domestic currency borrowing, however, will have to be a gradual process. In the meantime, governments need to limit the risk created by their short-term foreign currency liabilities by accumulating a buffer of liquid foreign currency assets. The amount of buffering should also take into account off-balance sheet positions and other contingent liabilities, including those that may eventually emerge from currency and maturity mismatches in the private sector. This said, governments should limit their contingent commitments where possible, for example, by avoiding implicit guarantees for private sector investments.
16. Given that the private sectors' hedging of exchange rate risk will typically be incomplete in emerging market economies with high liability dollarization, official reserves may play a role in providing emergency liquidity. In an emerging market economy, attempts by one sector to hedge its currency mismatch may just transfer the risk to another sector. For example, banks lending in foreign currency to the local corporations to reduce their net open foreign exchange liabilities position are simply shifting the vulnerability to an exchange rate shock to their clients--and in turn, raising their exposure to credit risk (non-performing loans of the corporate sector). If the corporate sector hedges its foreign currency mismatch through off-balance sheet transactions, for example, currency forward contracts with banks, the exchange rate risk is transferred back to the banks. While this may help to direct risks to those particular banks or corporations that can manage them somewhat better than others, in an economy where liability dollarization is high, the overall exposure to exchange rate risk will have to remain largely unhedged. In such circumstances a government may want to hold ample foreign currency reserves, which allows it to become a potential provider of financing to a sector with a foreign currency liquidity problem. Although it has inherent risks, the provision of emergency liquidity by the government at an early stage of a crisis may help to restore confidence and avoid a sectoral problem snowballing into a broader crisis.
17. The balance sheet approach also highlights the fact that macroeconomic policies adopted in response to shocks may be constrained by domestic balance sheet mismatches. A now familiar policy dilemma in the Asian crisis was the following: tight monetary policy aimed at preventing an excessive real depreciation may protect balance sheets with large currency mismatches (e.g., the corporate sector); but at the same time higher interest rates create further pressures on balance sheets with significant maturity mismatches (e.g., the banking sector). Knowing whether open currency positions or the maturity mismatch is the dominant concern will assist in making a better-informed policy choice between two painful alternatives--letting the exchange rate depreciate or raising short-term interest rates. It also helps inform a discussion of whether a restriction of capital outflows may be useful as a temporary emergency measure that makes lower interest rates compatible with a more appreciated exchange rate--thereby providing some breathing space for policy makers to implement the necessary policy adjustments.
III. The balance sheet approach in the Fund's current work
18. The Fund has been using insights based on balance sheet concepts in its surveillance as well as its program work for some time. There is, for example, increased emphasis on the adequacy of official reserves in relation to short-term debt, money aggregates, and other stock variables; the focus on dollarization risks; and the efforts at promoting better public liability management. Other important work that uses balance-sheet-related concepts are the new framework for debt sustainability analysis (DSA), the quarterly external vulnerability exercise, and the FSAP:
19. The efforts to incorporate the balance sheet approach into the Fund's work have been supported by statistical and transparency initiatives. These have improved the availability of some key balance sheet stock data and the accuracy of these data. In particular, the SDDS requirements have improved the dissemination of data and metadata on public and external debt, international reserves and foreign currency liquidity, international investment positions, and analytical accounts of the banking sector.3 The Funds' Coordinated Portfolio Investment Survey has improved the availability and comparability of statistics of countries' portfolio investment positions. An inter-agency task force chaired by the Fund has developed a new guide on the measuring and monitoring of external debt.4 The new Government Finance Statistics Manual 2001 supports the balance sheet approach through a new statistical framework that systematically links flows and stocks and introduces the concept of a government balance sheet.
IV. Further use of the balance sheet approach for surveillance purposes
20. Identifying the mismatches and linkages of sectoral balance sheets, including their off-balance sheet activities, can further sharpen the Fund's understanding of members' vulnerabilities. Such an analysis requires taking into account the asset and liability positions, including their currency denomination and maturities, not only for an economy as a whole (vis-à-vis non-residents) but also for each of an economy's sectors (vis-à-vis each other). This reveals the liability positions among residents--information that can be crucial in detecting a capital account crisis, but is often netted out in the published data (for example, in the IIP). In some cases, not only the linkages between sectors but also the interactions within a sector may be important (for example, trade credit among corporations or interbank lending). Importantly, such an analysis has to account for off-balance sheet positions to correctly gauge risk exposures.
21. Indeed, in a number of current surveillance and program cases, the balance sheet perspective can help to systematize the policy discussion. Wide-spread liability dollarization is causing concern in Latin America, as is its counterpart in several Eastern European countries (euroization).5 For the latter, the prospects of EU accession have become a driver for capital inflows and exchange rate movements that make sectoral balance sheets vulnerable to a potential reversal of investors' sentiment. To evaluate the adequacy of monetary and exchange rate policies, and to assess the soundness of the banking system (e.g. its foreign currency lending practices to an unhedged corporate sector), a closer analysis of sectoral balance sheets seems indispensable.6
22. Balance sheet analysis can also help to recalibrate the Fund's policy advice, for example, regarding the size of liquidity buffers.7 In its current approach to reserve adequacy, the Fund focuses on reserves to match the external debt falling due in the short term. In countries where governments issue significant amounts of domestic debt in foreign currency, however, the question may arise whether and to what extent the reserve coverage should take account of such foreign currency liabilities to residents. In countries where banks have large short-term foreign currency liabilities to residents (e.g., in the form of foreign currency deposits), it may be sensible for official reserves to cover the resulting foreign currency liquidity gap in the banking system. Better knowledge of the currency risks in sectoral balance sheets, especially the banking sector, helps to address these issues and to gauge whether policy advice concerning higher reserves or more limited issuance of foreign currency debt might have to be more ambitious. Carrying out this analysis for more countries would also facilitate the development of benchmarks that allow vulnerabilities to be put into perspective, for example by comparing risk levels among emerging economies.
23. The balance sheet approach may also be useful in defining new directions in banking supervision. In past crises, e.g., in Asia, the relationship between credit risk, foreign currency risk and liquidity risk proved to be insufficiently incorporated in the risk measurement and management of banks. Often supervisory authorities did not address this issue forcefully enough, and they did not always assess the macroeconomic risk arising from the aggregate mismatches in banks' balance sheets and the currency-induced risk coming through corporate sector balance sheets onto the books of the banks. Given the banking sector's key role in the transmission of shocks, more emphasis may need to be put on the detection of such vulnerabilities, emphasizing, for example, that regulators try to slow down excessive credit growth in boom periods and that they ensure that banks' credit analysis takes into account a borrowers' foreign currency exposure.
24. The simple matrix presentation of sectoral asset and liability positions introduced in working paper WP/02/210 can serve as basis for a sectoral balance sheet analysis. The paper has prompted attempts by staff to undertake a more extensive sectoral balance sheet analysis in several countries (Brazil, Peru, Lebanon, the Philippines, and Thailand). The concept allows highlighting the key balance sheet mismatches and sectoral inter-linkages at one point in time. Simple simulations can also be carried out, which, despite their static nature, help to better understand the impact of possible shocks, e.g., a sudden withdrawal of bank deposits or a decline in roll-over rates for external debt, and how such shocks could spill from one sector into another. There is of course no single well-established way of presenting and analyzing sectoral balance sheet data, and obviously more complex ways of modeling are possible, e.g., by explicitly incorporating measures of volatility.
25. Although some of this work can be done with existing data, significant data constraints remain and it may require substantial resources to address these shortcomings. National sources, the Fund's IFS, the Fund's Coordinated Portfolio Investment Survey, and some international data sources (e.g., the BIS banking statistics) can provide the basis for a sectoral analysis. But often balance sheet data--especially with regard to the corporate sector--may not always be readily available or in a useful format, not only in emerging markets but also in industrial countries. The same is true to an even greater extent for information on off-balance sheet positions, which is nevertheless crucial for a complete analysis of risk exposures. To compile the necessary data can be very resource intensive, and while the Fund can provide guidance, much depends on the efforts member countries are able to make in this direction. However, country authorities may see the advantages of collecting such data, in particular on the financial and the corporate sector.8 Indeed, such data is not only useful in identifying vulnerabilities and addressing them, but it can also be used to build confidence by underscoring strengths and ensuring that the right benchmarks are chosen for comparison. A notable example of such efforts is the Bank of Thailand's non-bank private debt survey.
26. Even an analysis limited to the balance sheet of the banking sector and the sovereign provides valuable information about an economy's resilience to potential shocks, but data constraints and other caveats must be carefully taken into account. In cases where the heterogeneity of the corporate sector unduly complicates a comprehensive sectoral balance sheet analysis, a thorough examination only of the banking sector and the government--for which data is more readily accessible--will still provide a useful perspective of the economy's vulnerability to balance sheet shocks. But the usual caveats about drawing premature policy conclusions apply even more strongly to such partial analysis. In particular, it is prone to the danger of providing a misleading picture of the risks to an economy, especially if information on off-balance sheet activities is incomplete.
27. Sectoral balance sheet analysis can be useful to distinguish between cases where Fund lending can help from those where it cannot.9 In addition to the standard debt sustainability analysis that aims at safeguarding Fund resources, balance sheet information would help to identify whether a crisis is mainly generated by short-term liquidity problems rooted in maturity mismatches or due to a debt overhang. While the first may be resolved by a combination of policy adjustments and financing--provided the mismatches are not too large--the latter could require a debt restructuring to restore solvency. Balance sheet analysis can help inform better judgments about the appropriate role of official finance in crises and the policies necessary for recovery. Official financing can help to restore confidence and prevent a balance sheet problem in one sector from snowballing into other sectors. In the recent arrangement with Uruguay, for example, Fund resources were used to rebuild official reserves to support a restructuring of the banking system, a comprehensive restructuring of the sovereign's debt, and a change in the exchange rate regime.
28. Knowledge of the type and magnitude of sectoral balance sheet mismatches helps in formulating policy advice that can restore confidence. In a financial crisis, monetary and exchange rate policy choices involve complicated trade-offs, depending on the dominant mismatch (maturity vs. currency) that exists on the balance sheets of the economy's main sectors. Accurate information about sectoral balance sheets and their linkages enables policy makers to take informed decisions regarding these policy trade-offs, and for the Fund to lend with more confidence in support of the authorities' program.
29. The balance sheet approach is an important contribution to our understanding of capital account crisis, and it has been used in the Fund's work for some time. The approach highlights vulnerabilities that are crucial for explaining crises in emerging market economies and therefore usefully compliments the traditional flow-based analysis. Insights from the balance sheet approach have indeed guided important Fund initiatives. Staff's analysis of economies' sectoral balance sheets and their potential interaction often has to remain limited, however, owing to the lack of adequate data and other constraints that complicate the real-world application of the balance sheet approach.
30. A further integration of sectoral balance sheet analysis into the Fund's work can strengthen existing surveillance initiatives. It can help to refine the existing vulnerability exercise, by adding an assessment of sectoral balance sheet risks and their interactions. It can add to the DSA framework, by helping to identify where, in the event of a shock, existing mismatches on sectoral balance sheets could turn a liquidity into a solvency (debt sustainability) issue. It also can generally promote a better integration of the specialized FSAP work into the main country analysis.
31. More balance sheet information can also help the Fund to recalibrate its policy advice. A more systematic detection of an economy's sectoral vulnerabilities allows identifying more clearly the type and magnitude of (liquidity) buffers needed. This has implications for the ambitiousness of current reserve levels and fiscal cushions, for debt management, and for prudential regulation and supervision of the banking system.
32. Finally, the balance sheet perspective can be valuable when handling crises. Better knowledge about sectoral balance sheet mismatches helps to separate the cases where Fund's foreign currency lending can help and where such lending needs to be accompanied by a debt restructuring. It may also help to ensure that scarce financing is used where most needed and help make policy packages more effective in restoring confidence.
33. This said, the analytical framework can be difficult to apply and its results must be interpreted with caution. Incomplete data and the static nature of the approach may create misleading perceptions about overall risks. And even if the existence or size of balance sheet mismatches can be identified, this does not necessarily provide clear policy answers. Balance sheet analysis should thus be seen as a useful compliment to the range of other approaches the Fund employs in its work.
1A similar recommendation was given in the Report of the Working Group on Capital Flows (chaired by Mario Draghi) presented at the meeting of the Financial Stability Forum in March 2000: "Risk monitoring at the national level could be assisted by compiling a balance sheet for the economy as a whole and for key sectors, designed to identify significant exposures to liquidity, exchange rates, and other risks." (p. 2)
2See Assessing Sustainability (SM/02/166 and BUFF/02/86).
3See also "Data Provision to the Fund for Surveillance Purposes" (SM/02/126).
4See "External Debt Statistics-Guide for Compilers and Users" (final draft, November 2001).
5See also "Financial Stability in Dollarized Economies" (SM/03/112) and "Macroeconomic Policies in Dollarized Economies" (SM/03/126).
6Financial markets have also reacted to the risk posed by foreign exchange liabilities among residents. For example, Moody's recently announced adjustments to its sovereign rating methodology for countries with highly dollarized banking systems, introducing elements of balance sheet analysis and arguing that the "distance between exchange rate collapse and deposit run has become shorter".
7These issues will be taken up in more detail in a forthcoming paper on liquidity management.
8See also the recommendations of the FSF Working Group on Capital Flows ("Draghi Report"; p. 17)
9For a discussion of this issue, see Jeanne and Zettelmeyer "Original Sin-Balance Sheet Crises, and the Roles of International Lending" (WP/02/234).