Chapter

12 An Operational Framework for Addressing the Public Costs of Systemic Bank Restructuring

Author(s):
Charles Enoch, Dewitt Marston, and Michael Taylor
Published Date:
September 2002
Share
  • ShareShare
Show Summary Details
Author(s)
Priya Basu

Systemic bank restructuring entails high costs, which are borne largely by governments and funded through the issuance of vast quantities of public debt. Cross-country estimates show that the stock of public debt issued in the context of recent systemic bank restructuring programs has ranged from single digits to 50 percent or more of GDP. While government support to bank restructuring through the issuance of public debt can help address the immediate crisis, it can result in escalating costs to the government, raising significant, and often unanticipated, medium-term risks related to fiscal and debt sustainability and financial stability. This chapter presents a simple operational framework for quantifying, analyzing, and reducing such costs, on an ex post basis.

In recent decades, a large number of countries—both developed and developing—have experienced systemic banking crises requiring a major—and expensive—overhaul of their banking systems, with the costs of bank restructuring shouldered largely by governments (and thus, ultimately, by taxpayers). These public costs of bank restructuring have arisen from the need for solvency support to ailing banks through the recapitalization of these banks, the purchase of nonperforming loans from banks, the paying out of the guaranteed liabilities of banks that are closed, and the provision of liquidity support. The bulk of these costs have been borne through the issuance of domestic public debt in the form of government bonds, issued directly to ailing banks for recapitalization or in exchange for nonperforming loans. Governments have also issued bonds to central banks and other government agencies to compensate them for past liquidity support to ailing banks. In recent episodes of financial crises, for example, the governments of Ecuador, Indonesia, Jamaica, Korea, Mexico, Romania, and Thailand have shouldered the costs of bank restructuring almost entirely by issuing government bonds, with the stock of bonds issued estimated to range from single digits to 50 percent or more of GDP (Table 12.1).1

While government support for bank restructuring can help address the immediate crisis, the public costs of bank restructuring—manifested in the vast quantities of public debt issues and the servicing of this debt—can raise significant, and often unanticipated, risks to macroeconomic and financial stability, contributing to increased country vulnerability and to macroeconomic and financial shocks.2 Servicing this debt may contribute to chronic fiscal imbalances, necessitating distortionary taxation.3 A high public debt burden may also limit the monetary authorities’ ability to pursue price stability by putting pressure on the central bank to monetize some debt to alleviate the burden.4 If debt dynamics become unsustainable, some policy change will eventually be needed, and this can take the form of fiscal consolidation, inflation (and attendant currency depreciation), or—under extreme circumstances—default.5 Addressing the public debt arising from systemic bank restructuring can thus become an important item on the agenda of policymakers in the aftermath of a banking crisis.6

Table 12.1.Public Costs of Systemic Bank Restructuring: Selected Country Estimatesas percentage of GDP in 2000)
Cross Public

Costs Incurred as

of Mid-2000
Government

Bonds Issued as

of Mid-2000
Estimated Additional

Costs for

FY 00/01
Ecuador (1998-mid-2000)12.312.07.7
Indonesia (1997-mid-2000)n.a.55.03.0
Jamaica (1996-mid-2000)45.040.63.0
Korea (1997-mid-2000)20.412.29.5
Mexico (1994-mid-2000)19.019.00.3
Romania (1997-mid-2000)4.74.7n.a.
Thailand (1997-mid-2000)n.a.10.00.02

The specific nature and purpose of the debt issued in the context of bank restructuring, and the lack of transparency surrounding the issuance and accounting of such debt, highlight the need for an explicit operational framework for quantifying, monitoring, and reducing the debt and its associated debt servicing burden. This chapter attempts to present such a framework. The framework recognizes that any strategy for addressing the public debt arising from a banking crisis must balance the specific objectives of bank restructuring with the government’s broader macroeconomic objectives. It emphasizes that the results of any strategy to address such debt must be integrated with the overall government debt management effort7 and linked to a clear macroeconomic framework under which governments seek to ensure that the level and rate of growth in public debt, and the servicing of that debt, are sustainable.

The Specific Nature of Public Debt Issued for Bank Restructuring

Two main features distinguish the public debt issued for bank restructuring from other forms of public debt. First, unlike most government debt, which is floated with the objective of raising cash to help finance the fiscal deficit, the primary purpose of most of the debt issued for systemic bank restructuring is to provide solvency support to ailing banks, and may be viewed largely as an exercise in creative accounting. Second, and related to the first point, government bonds—which are the main instruments through which public debt is issued for the purposes of bank restructuring—are not issued and sold through the standard market-based (auction) system, but rather are placed directly on the books of the recipient institution (appearing on the “assets” side of the recipient institution’s balance sheet). The interest paid by the government on these bonds is often the main source of liquidity for the bank after it has been recapitalized. Therefore, in restructuring these bonds to reduce the government’s debt burden and servicing costs, the authorities have to balance macroeconomic objectives with the objectives of bank restructuring, namely, recapitalizing banks, covering losses, and bringing a troubled banking system back to soundness and profitability. For example, restructuring all bonds into zero coupon bonds can help reduce the medium-term fiscal burden associated with bank restructuring—because the government would not have to provide immediate cash in the form of interest payments. But this can also slow down the process of restoring ailing banks to health, because a bank that has a large part of its assets in the form of such bonds would likely continue to suffer severe cash-flow constraints. Converting bonds into fixed coupon bonds with coupon rates lower than prevailing market interest rates may ease short-term fiscal constraints, but it may present a problem over the longer term, to the extent that if banks want to sell down these bonds, they may be forced to do so at a discount and bear the capital loss.

Bank restructuring bonds may be issued either directly by the central government (as in Thailand) or by government agencies such as bank restructuring agencies (as in Indonesia, Jamaica, and Malaysia), asset management companies (as in Korea), and deposit insurance agencies (as in Mexico). Bonds that are issued by a government agency, rather than directly by the central government, are typically guaranteed by the central government. The holders of the bonds may include intervened financial institutions; the central bank—in cases where the government issues bonds to the central bank to compensate it for past liquidity support to banks; other government entities; and the central government—in cases where the bank restructuring agency issues debt to the central government in exchange for initial cash support provided to assist in covering the operational expenses of the restructuring agency (Figure 12.1).

Figure 12.1.Issuers and Holders of Bonds

The bonds issued include a mix of fixed coupon, variable coupon, zero coupon (i.e., bonds issued at a discount, earning no interest but redeemable at par value), and index-linked bonds, which may be tradable or nontradable, denominated in domestic or foreign currency. In exchange for the bonds issued, governments may receive equity or subordinated debt in the intervened institution, or nonperforming loans (Table 12.2).

Table 12.2.Types of Bonds
Debt Instrument IssuedReceived
Indonesia
  • Variable and fixed coupon bonds
  • Variable coupon bonds
  • Index-linked bonds; tradable
  • Equity
  • Nonperforming loans
  • Equity
Jamaica
  • Fixed and variable coupon bonds; tradable
  • Fixed and variable coupon bonds; nontradable (allowing for interest capitalization)
  • Equity
  • Equity, subordinated debt, nonperforming loans
Korea
  • Variable coupon bonds; tradable
  • Fixed and variable coupon bonds
  • Equity or preference shares
  • Nonperforming loans
Malaysia
  • Zero coupon bonds; not easily traded
  • Zero coupon bonds
  • Convertible preference shares or subordinated debt
  • Nonperforming loans
Mexico
  • Ten-year promissory notes, nontradable (allowing for interest capitalization)
  • Nonperforming loans
Thailand
  • Fixed coupon bonds; tradable
  • Fixed coupon bonds; nontradable
  • Equity
  • Subordinated debt

Key Steps in Addressing the Public Costs of Bank Restructuring

Addressing the public costs of bank restructuring involves a series of steps, including cost recognition and quantification of the total costs, the specification of alternative scenarios to reduce the costs, and determination of the preferred strategy. Each step requires the concerned authorities—in most cases the ministry of finance, in coordination with the bank restructuring agency—to make difficult decisions, as illustrated in Figure 12.2.

Cost Recognition and Quantification of the Debt Problem

The first step is to assess the magnitude of public costs, to serve as a basis for projecting the evolution of costs over time and assessing their impact on medium-term debt and fiscal sustainability. This requires a full and transparent accounting of the total public costs of bank re-structuring and the incorpotation of these costs into the budget.

Figure 12.2.Decision Sequence

Because the current guidelines for classifying government costs in bank restructuring are founded on the use of the cash-based balance of the general government, as described in the 1986 Manual on Government Finance Statistics, and do not allow for the explicit inclusion of the stock of debt issued for bank restructuring, it may be misleading to follow these guidelines in quantifying the total magnitude of the public costs.8 An “augmented fiscal balance” approach, involving the explicit recognition of all the major quantifiable budgetary and quasi-fiscal costs of bank restructuring, is more appropriate for gaining an accurate picture of the total public costs of bank restructuring.9 Under this approach, the public costs would be defined as the entire stock of debt issued in support of bank restructuring, plus the costs of servicing that debt. The approach requires identifying each and every bond issued and gathering information on the maturity, currency, interest rates, schedule of coupon payments, and amortization schedule of each bond.

Once the stock of debt has been calculated, projections need to be made on the future evolution of this stock, over the medium term (generally defined as the next three to four years). Preparing accurate stock projections requires that the following key issues be considered:

  • The stock projections should properly reflect whether any (or all) of the bonds allow for the capitalization of interest accrued. If so, all bonds to be issued over the medium term in lieu of interest payments must be fully reflected in the stock projections, based on thorough information about the coupon payment schedule of each bond. For variable coupon bonds on which interest is being capitalized, assumptions need to be made about the projected interest rate path so that the amount of future bonds to be issued in lieu of interest payments can be measured accurately. For foreign exchange-denominated bonds on which interest is being capitalized, medium-term exchange rate projections need to be made so that the amount of future bonds to be issued in lieu of interest payments can be estimated accurately.
  • A clear understanding needs to be reached on the government’s existing commitments to issue new bonds to support financial restructuring; the level of commitments, types of bonds to be issued, and details of the proposed coupon payments need to be defined.
  • Estimates also need to be made of the likelihood of any additional bonds that may be required to provide solvency or liquidity support to financial institutions over the medium term. This requires an assessment of the current and future soundness and stability of the financial system, including capital adequacy, asset quality, management, liquidity, and profitability. In cases where the assessment suggests that financial institutions are likely to require solvency or liquidity support in the future, estimates need to be made of the extent of such support needed, and of whether this would need to be funded through the issuance of new government bonds (as opposed to funds raised by the private sector). These estimates must be factored into the stock projections.

Based on the stock projections, an assessment has to be made of whether the government has the capacity to sustain the projected debt burden and the attendant fiscal costs of servicing the debt. If the debt burden and its carrying costs are deemed to be unsustainable, then a strategy for addressing the debt problem would clearly be warranted.

Scenario Specification

Any strategy for debt reduction must begin with the specification of alternative scenarios under which the problem may be addressed. A starting point in the scenario specification exercise is to define the objectives of the strategy. While the specific objectives would vary depending on the nature and the extent of the problem at hand, and the particular circumstances of the country in question, it is reasonable to assume that the strategy would aim, at the very minimum, to reduce the stock of debt and associated debt servicing costs to levels that are sustainable over the medium term, while ensuring that the objectives of the bank restructuring program are not jeopardized.

Once the objectives of the resolution strategy have been defined, the authorities have to make a number of decisions (Figure 12.2). They must first decide on which components of the debt will be addressed through the strategy (e.g., whether debt issued both to the public and private sectors should be reduced/eliminated). Next, the amount by which the various components of the debt will be reduced needs to be determined (e.g., in the case of debt issued to public sector entities, will all of the debt be eliminated, or will debt owed to certain entities be eliminated selectively; similarly, by how much will the debt issued to various private institutions be reduced?). The period over which the reduction will be implemented also needs to be determined. These decisions tend to be guided mainly by the specific policy goals of the authorities. For example, if the goal is to privatize the recapitalized banks as soon as possible, and if a certain recapitalized bank is considered to be potentially more “salable” than others, with the main obstacle to its sale being the large portfolio of nonmarketable recapitalization bonds on the asset side of its balance sheet, the authorities may decide to buy back the bonds of that bank (in exchange for cash) on a larger scale than bonds held by other banks in the system.

The next steps involve selecting the appropriate financial engineering techniques. The available techniques typically include (1) write-offs, (2) debt buy-back operations, (3) swaps, (4) debt restructuring through a conversion of the debt instruments (e.g., the conversion of nonmarketable debt into marketable instruments, which make regular coupon payments, based on market-determined interest rates), and (5) simply “inflating out” of the debt problem, so that real interest rates on the debt can be significantly reduced without actually implementing a systematic debt reduction program (Table 12.3).

Not all of these techniques are equally preferable. For example, writing off debt issued by the government carries the important risk of moral hazard, creating disincentives for the government to honor its debt obligations in the future. Moreover, if governments use write-offs for dealing with intragovernment debt (e.g., debt issued by the treasury to the central bank, or by a government-owned bank restructuring agency to the treasury), this could send adverse signals to the private sector, which may be led to believe that such write-offs are also an option for dealing with government bonds held by private institutions. If governments use write-offs for addressing debt issued to selected private institutions, this could also have systemic implications on the private sector’s perception of the government. In general, writing off debt issued by the government may contribute to a loss of credibility for the government, which risks generating panic on the part of private holders of government debt. Similarly, the option of inflating out of the debt problem carries serious adverse consequences, not least of which is the risk of macroeconomic instability. On the other hand, debt buy-back operations, swaps, and improvements in the marketability of bonds are associated with fewer obvious adverse consequences.

Table 12.3.Objectives and Techniques
ObjectivesFinancial Engineering Techniques
  • To reduce the stock of debt to a level that is sustainable over the medium term
  • To ensure that the fiscal burden (i.e., the debt-servicing costs) is sustainable and in line with the overall macroeconomic program
  • Write-offs
  • Debt buy-back operations
  • Swaps
  • Conversion of debt instruments
  • Inflating out of the debt problem

Governments may not always have the luxury of being able to choose from among the various alternatives presented above. The choice of techniques is limited by a number of factors, including the composition of the debt, time constraints for debt resolution, and perhaps most important, the resources available to the government to implement the selected techniques. The use of debt buy-back operations, for example, is constrained by the budgetary or extra budgetary resources available for such purposes. Governments may wish to use resources from asset recovery to finance debt buy-back operations, which would imply that the scale of buy-back operations would be constrained by the success with which governments are able to recover value on “core” assets (i.e., the equity of various ailing financial institutions that is transferred to the government/government entities in exchange for the bonds issued to those institutions) and “noncore” assets (i.e., non-performing loans and collateral transferred to the government/government entities in exchange for the bonds issued).

The techniques chosen may be combined in different ways, under varying assumptions about interest payment and accrual, resources available to reduce the stock of debt, and changes in the stock of debt held by the public and private sectors, so as to specify a menu of alternative scenarios for debt resolution. In theory, available techniques may be combined in an infinite number of ways—under different assumptions—to develop a menu of scenarios. A set of scenarios, based on rather simplified assumptions, is presented below.

Assume that the stock of debt, issued in the context of the bank restructuring program, is defined as follows: D = Dpb+Dpv, where D is the total stock of public debt issued by the government in the context of a systemic bank restructuring program, when Dpb is the stock of debt held by other public sector entities, that is, intragovernment debt (e.g., debt held by the central bank), and Dpv is the stock of debt held by the private sector.

Assume further that the debt burden and associated fiscal costs are unsustainable over the medium term and that the authorities have determined the need to design a debt resolution strategy, with the objectives of reducing D, so that both D, and the annual costs of servicing it, are sustainable over the medium term.

It is agreed that the strategy should be implemented over a period, say, from t through t+3. It is also decided that the reduction in D will be implemented through efforts to eliminate Dpb, and to reduce Dpv by as much as possible, given available resources.

Having agreed on these definitions, the authorities need to identify appropriate techniques to develop alternative scenarios. Let us assume that the preferred tool to bring about a reduction in Dpb is an intragovernment debt write-off. Changes in Dpb are determined by a parameter δ, which denotes the government’s ability to negotiate write-offs: if δ = 0, Dpb remains unchanged; if δ = 1, Dpb is fully retired.

Let us also assume that the preferred tool to bring about a reduction in Dpv is a debt buy-back operation, and that the only resources available to fund this buy-back operation are proceeds from the sale of “core” or “noncore” assets transferred to the government in the course of bank restructuring. Core assets comprise the government’s equity in the recapitalized financial institutions (often representing 100 percent of the equity of the institution concerned), while noncore assets comprise nonperforming loans, real estate assets, and similar instruments purchased by the government in the restructuring process. It follows that the amount by which Dpv can be reduced through a buy-back operation is determined by the amount that can be recovered through the sale of core and noncore assets, with the assumption that proceeds from asset recovery would be devoted entirely to buying back Dpv.

Recovery on core assets is denoted by a value parameter α: if α=0, this would mean that there is no recovery on core assets for any given year, and therefore that no resources are available from the sale of core assets to reduce Dpv; if α = α*, this would mean that some amount is recovered from the sale of core assets, which the government would use toward buying back an equivalent amount of Dpv.10

Recovery on noncore assets is denoted by the value parameter γ: if γ = 0, this would mean that the recovery value on noncore assets for any given year is zero; if γ =γ*, this would represent (in present value terms) the government’s recovery value from the sale of its noncore assets, which it would then use toward achieving a reduction in by an equivalent amount.11

The cost to the government of servicing the debt (i.e., interest payments) is denoted by parameter i: if i = 0, this would mean that all interest on outstanding debt in any given year is capitalized; if i=1, this would mean that cash interest would be paid out on the debt outstanding as of year t+1, commencing in t+1; if i = 2, this would mean that cash interest would be paid out on the debt outstanding as of year t+2, commencing in t+2.

Based on the above, various alternative scenarios for debt resolution may be developed. Table 12.4 presents some of these scenarios.

The final step is to evaluate the effectiveness of each scenario in meeting the predefined objectives of the strategy to address the public costs of bank restructuring, so that a preferred or “optimal” strategy may be selected. This requires estimating and analyzing the impact of each scenario on the evolution of the stock of debt, on the government’s annual interest payment obligations for the period t through t = 3, on other predefined macroeconomic goals, and on the bank restructuring program. The optimal strategy would be one that would best achieve the predefined objectives of debt reduction and fiscal sustainability while assisting the banking system to return to soundness and profitability. It is also important that the strategy does not jeopardize the government’s credibility vis-a-vis the market. Thus, before the optimal strategy is chosen, it would be necessary to consider, for example, how the private sector would perceive a debt write-off. At the same time, the impact on banks’ balance sheets of buying back the recapitalization bonds issued to troubled banks, swapping those bonds, or converting them into marketable instruments would need to be carefully evaluated—-with a view to ensuring that a debt resolution strategy involving such measures does not in any way jeopardize the solvency, liquidity, or profitability of the banks in question.

Table 12.4.A Menu of Alternative Scenarios
Policy VariablesExamples of Alternative Strategies
δ = 0 or 1α = 0 or α*Baseline: δ = 0, α = 0, γ = 0, i = 0 for each year from t through t + 3
I = 0, 1 or 2γ = 0 or γ*Scenario 1: δ = 0, α = 0, γ = 0 for each year from t through t + 3; i = 1
Scenario 2: δ = 0,α= 0, γ = 0 for each year from t through t + 3; i = 2
Scenario 3: δ = 1,α=α*, γ = 0 for each year from t through t + 3; i = 1
Scenario 4: δ = 1, α = α*, γ = γ* for each year from t through t + 3; i = 1

A Numerical Example

This section provides a numerical example to illustrate the steps involved in developing a strategy to address the public costs of bank restructuring.

Assumptions

Consider an economy where the government has financed a systemic bank restructuring program entirely through the issuance of bonds. All bonds have been issued by the government-owned bank restructuring agency and are guaranteed by the government. The total stock of public debt (bonds) issued (D) is $100. Bonds issued to, and held by, the public sector (Dpb) amount to $40 of this (comprising a combination of bonds held by the central bank to compensate it for past liquidity assistance provided to troubled banks, and bonds held by the central government to compensate it for cash support provided in the past to help cover operational expenses). Bonds issued to, and held by, the private sector (Dpv) amount to $60 (comprising a combination of bonds held by troubled banks and other financial institutions, including nonbank financial institutions).

The bonds are denominated in domestic currency, are nontradable, and have a maturity structure in the 3–15 years range, with interest payable on a biannual basis, either at a fixed rate or at a floating rate (typically, at the six-month treasury bill interest rate plus 100 basis points). Interest on bonds is capitalized through the issuance of new bonds.

The majority of the bonds issued carry variable interest rates; therefore, projections on the future evolution of the debt stock require assumptions about the nominal interest rate path. Nominal interest rates are projected to decline over the medium term, and GDP is assumed to increase over the medium term (Figure 12.3).

Figure 12.3.Illustrated Interest Rate and Exchange Rate Projections

(baseline scenario)

Assume also that the government does not need to issue any additional bonds to meet liquidity or solvency needs of the banking system over the medium term, but that all interest is capitalized through the issuance of new bonds.

Quantifying and projecting the stock of debt

The first step for the authorities would be to quantify the existing debt problem. With interest capitalized, and with no reductions in the stock of debt, the total stock of debt is estimated at $100, and it is projected to evolve as shown in the baseline scenario (Figure 12.4) increasing from $100 in base year FY1999/2000 to about $175 by FY2003/2004, or an annual growth rate of about 15 percent. Based on these estimated projections, let us assume that the authorities determine that the debt problem would become unsustainable, and that a strategy for addressing this debt is therefore warranted.

Figure 12.4.Evolution of Public Debt (Baseline Scenario)

(In U.S. dollars

Scenario specification

Now assume, as before, that the only technique available to the authorities for reducing Dpb is a write-off: if δ = 0, there is no write-off; if δ = 1, this means that the entire amount of Dpb ($40) is written off. Also, assume that the only technique available for reducing Dpv is a buy-back operation—funded through proceeds from the sale of core assets (a) or noncore assets (γ), where the maximum recovery from the sale of core assets (α*) is $10 and the maximum recovery value on noncore assets (δ*) is estimated at $5, in present value terms. Against this background, the scenarios defined in the section on the public costs of bank restructuring can be developed, based on alternative combinations of key choices on interest accruals and payments, changes in the stock of debt held by the public sector effected through a write-off of public debt, and changes in the stock of debt held by the private sector effected through debt buy-backs, funded by proceeds from the sale of core assets and noncore assets. Summary results of the impact of each scenario on the stock of debt, and on the government’s debt-servicing costs (shown as a percentage of GDP, on an annual basis for the period FY 1999/2000 through to FY 2002/2003) are presented in Figure 12.5.

Figure 12.5.Illustrated Evolution of Debt Stock and Debt Servicing Costs

Assessment12

Figure 12.5 (top panel) shows the impact of the various scenarios on the evolution of the total stock of debt over the medium term. Scenarios 1, 3, and 4 each result in stock reductions over the medium term, in varying degrees, achieved through different combinations of policy tools to reduce the stock of debt held by the public sector and the private sector and assumptions about the capitalization of interest on the debt. In contrast, Scenario 2 results in an increase in the stock of debt as a percentage of GDP, with the ratio being higher for 2002-2003 than for 1999—2000, because it assumes that interest would continue to be capitalized up until FY 2001/2002, resulting in a sizable increase in the stock of debt, and that no reductions would be made in the stock of debt held by the public or private sector.

Figure 12.5 (bottom panel) shows the impact of the various scenarios on the evolution of the government’s interest payment obligations, and hence on the fiscal balance, over the medium term. Scenarios 1 and 2 result in the maximum drain on government cash resources in the outer years, but a minimum drain in the initial years. These scenarios assume that the government would commence cash interest payments on the debt in FY 2001/2002 and 2002/2003, respectively. Furthermore, they assume that the government would not be able to recover any value on core or noncore assets, and therefore would have no resources available to bring about a reduction in the stock of private sector debt (which is predicated on the availability of resources to buy back the debt); and that the government would not be able to reach an agreement on the write-off of debt held by other public sector entities. Scenarios 3 and 4 show the impact on the government’s cash flow of different degrees of realization from core and noncore assets.

In this hypothetical case, Scenario 4 presents an optimal strategy for addressing the public debt problem, with optimal recoveries on the sale of both core and noncore assets, which would help provide significant resources for the buy-back of private sector debt. It also assumes a resolution/write-off of debt held by the central government and public sector entities. The scenario would represent a “first best” strategy, in terms of both its impact on the government’s debt servicing burden and the stock of debt. Interest payments at the nominal rate of interest would yield a cost to the government of about 2.7 percent of GDP for 2001-2002, trending down to 2 percent of GDP for 2002-2003. At this level of effort, the government would be lowering the stock of debt in real terms. The stock of debt would decline from 35 percent of GDP in FY 1999/2000 to 18 percent in FY 2002/2003.

Some Caveats

This chapter has aimed to present an operational framework for quantifying, analyzing, and addressing, on an ex post basis, the public costs arising from systemic bank restructuring. The framework developed in this chapter is based on a set of simplified assumptions on the composition of public costs, on how the debt is structured in terms of its maturity, currency, and interest rate composition, on the specifics of the debt problem, and on the factors guiding both the specification of various alternative resolution scenarios and the choice of an optimal strategy. In practice, developing a strategy for addressing the public costs of bank restructuring is likely to be more complex, requiring governments to make difficult, and often competing, choices at various stages. Because no decision on an optimal strategy can be made independently of its potential impact on medium-term macroeconomic and financial stability, its implications for the bank restructuring program under way, and the perceived impact on the government’s credibility vis-à-vis market participants, each of these factors needs to be carefully considered in selecting the preferred strategy.

Finally, while the chapter focuses on addressing public costs on an ex post basis, this should not be seen to undermine the importance of the ex ante containment of the public costs arising from systemic bank restructuring. Indeed, decisions on the extent of public costs, the types of instruments to be used to finance these costs, and the pricing and marketability of those instruments, have important implications for the medium-term debt and fiscal dynamics and, more generally, for the overall macroeconomic and financial stability of the country in question. They should be addressed up front in the design of a systemic bank restructuring program.

References

    BeckerTorbjörn1999“Public Debt Management and Bailouts,”IMF Working Paper No. 99/103 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    BemheimB.Douglas1987“Ricardian Equivalence: An Evaluation of Theory and Evidence,”NBER Working Paper No. 2330 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation

    BlanchardO.R.Dombusch andW.Buiter1986“Public Debt and Fiscal Responsibility,”in Restoring Europe’s Prosperity: Macroeconomic Papers from the Centre for European Policy Studiesed. by OlivierJ.BlanchardRudigerDombusch andRichardP.G.Layard (Cambridge, Massachusetts: MIT Press).

    • Search Google Scholar
    • Export Citation

    DanielJamesA.JeffreyM.Davis andAndrewM.Wolfe1997“Fiscal Accounting of Bank Restructuring,”IMF Paper on Policy Analysis and Assessment 97/5 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    DooleyMichaelP.1998“Government Debt and Asset Management in Financial Crises; Sellers Beware,”Department of Economics Working Paper No. 409 (Santa Cruz: University of California).

    • Search Google Scholar
    • Export Citation

    EnochCharlesGillianGarcia andSundararajanV.2001“Recapitalizing Banks with Public Funds,”IMF Staff PapersVol. 48No. 1 pp. 58110.

    • Search Google Scholar
    • Export Citation

    FeldsteinMartin1982“Government Deficits and Aggregate Demand,”Journal of Monetary EconomicsVol. 9 (January) pp. 120.

    GiavazziFrancesco andMarcoPagano1990“Can Severe Fiscal Contractions Be Expansionary? Tales of Two Small European Countries,”NBER Working Paper No. 3372 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation

    GiovannmiAlberto andGustavoPiga1992“Understanding the High Interest Rates on Italian Government Securities,”CEPR Discussion Paper Series No. 720 (London: Centre for Economic Policy Research).

    • Search Google Scholar
    • Export Citation

    HonohanPatrick andDanielaKlingebiel2000“Controlling Fiscal Costs of Banking Crises,”Policy Research Working Paper No. 2441 (Washington: World Bank).

    • Search Google Scholar
    • Export Citation

    International Monetary Fund2001Guidelines for Public Debt Management (Washington: IMF). Available via the Internet at http://www.imf.org/external/np/mae/pdebt/2000/eng.

    • Search Google Scholar
    • Export Citation

    International Monetary Fund2001. Government Finance Statistics Manual (Washington).

    International Monetary Fund1986A Manual on Government Finance Statistics (Washington).

    IMF/World Bank2000Draft Guidelines on Public Debt Management (Washington: International Monetary Fund).

    LaneTimothy1997“High Public Debt: Consequences for Investment and Growth,”in Banking International Capital Flows and Growth in Europe: Financial Markets Savings and Monetary Integration in a World with Uncertain Convergenceed. by PaulJ.J. Welfens andHolgerC. Wolf (New York; Springer Verlag).

    • Search Google Scholar
    • Export Citation

    Lindgren Carl-JohanTomás BaliñoCharles EnochAnne-Marie GuideMarc Quintyn andLeslie Teo1999Financial Sector Crisis and Restructuring—Lessons from Asia IMF Occasional PaperNo. 188 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    PlosserCharlesI.1982“Government Financing Decisions and Asset Returns,”Journal of Monetary EconomicsVol. 9 (May) pp. 32552.

    • Search Google Scholar
    • Export Citation

    SargentThomasJ. andNeilWallace1981“Some Unpleasant Monetarist Arithmetic,”Federal Reserve Bank of Minneapolis Quarterly ReviewVol. 9 (Winter) pp. 1531.

    • Search Google Scholar
    • Export Citation

List of Authors

  • Richard K. Abrams
  • Advisor, Monetary and Exchange Affairs Department of the IMF.
  • Michael Andrews
  • Financial Sector Advisor, Monetary and Exchange Affairs Department of the IMF.
  • Priya Basu
  • Senior Economist at the World Bank in India. Previously Economist in the IMF’s Monetary and Exchange Affairs Department.
  • Luis Cortavarria
  • Senior Economist in the Systemic Banking Issues Division in the IMF’s Monetary and Exchange Affairs Department.
  • Fernando Delgado
  • Senior Economist in the Systemic Banking Issues Division in the IMF’s Monetary and Exchange Affairs Department.
  • Claudia Dziobek
  • Assistant to the Director of the Statistics Department of the IMF. Previously Senior Economist in the Monetary and Exchange Affairs Department of the IMF.
  • Charles Enoch
  • Senior Advisor in the Statistics Department. Previously Assistant Director of the Systemic Banking Issues Division of the Monetary and Exchange Affairs Department of the IMF.
  • Gillian Garcia
  • Previously Senior Economist in the IMF’s Monetary and Exchange Affairs Department.
  • Peter Hayward
  • Financial Sector Advisor in the Banking Supervision and Regulation Division of the Monetary and Exchange Affairs Department of the IMF.
  • Dong He
  • Senior Economist in the Systemic Banking Issues Division of the Monetary and Exchange Affairs Department, IMF.
  • J. Kim Hobbs
  • Director of Financial Risk Management at Capital One in Virginia. Previously consultant for the IMF.
  • David Hoelscher
  • Division Chief, Systemic Banking Issues Division, Monetary and Exchange Affairs Department of the IMF
  • Akihiro Kanaya
  • Investment Banking Division at Morgan Stanley Japan Ltd. Previously Economist, Monetary and Exchange Affairs Department of the IMF.
  • Daniel Kanda
  • Economist in the IMF’s Asia and Pacific Department. Previously in the Monetary and Exchange Affairs Department of the IMF.
  • David Marston
  • Division Chief, Banking Supervision and Regulation Division, Monetary and Exchange Affairs Department of the IMF.
  • Greta Mitchell
  • Senor Economist, Banking Supervision and Regulation Division, Monetary Affairs and Exchange Department of the IMF.
  • Armando Morales
  • Senior Economist, Monetary and Exchange Policy Review Division, Monetary Exchange Affairs Department of the IMF.
  • Inwon Song
  • Senior Economist, Banking Supervision and Regulation Division, Monetary and Exchange Affairs Department of the IMF.
  • Michael Taylor
  • Financial Sector Issues Representative in Indonesia of the IMF’s Monetary and Exchange Affairs Department. Previously Senior Economist in the IMF’s Monetary and Exchange Affairs Department.
  • V. Sundararajan
  • Deputy Director of the Monetary and Exchange Affairs Department of the IMF.
  • David Woo
  • Vice President of the Economic and Market Analysis Department of Citigroup. Previously Economist in the IMF’s Monetary and Exchange Affairs Department.
The author would like to thank the staff of the IMF’s Monetary and Exchange Affairs Department for their comments and assistance, in particular David Hoelscher, Claudia Dziobek, Oliver Frécaut, Yuri Kawakami, Delisle Worrell, and Mark Zelmer.
1These estimates represent the gross costs to the government. The net costs can be known only after proceeds from (re)privatization of banks and recoveries of assets accruing to the government have been taken into account. A more complete picture of the costs would also involve indirect effects of the crisis and subsequent reforms. For a detailed discussion of the costs of restructuring, see Lindgren, and others (1999) and Honohan and Klingebiel (2000).
2This chapter does not attempt to address the theoretical debate on the relevance or welfare implications of debt management. It may be useful to note, however, that there is a rich theoretical literature on the relevance of public debt management, stemming from the debate over the validity of the “Ricardian Paradigm.” In its purest form, Ricardian equivalence would render irrelevant the level and composition of government debt and its management. Whether or not Ricardian equivalence is a good approximation has received considerable attention in the empirical literature. Papers by Plosser (1982) and Giavazzi and Pagano (1990), for example, have lent support to Ricardian equivalence, while several other papers, for example, by Feldstein (1982) and Bemheim (1987), have argued against Ricardian equivalence. On balance, empirical work does not appear to support the idea that the private sector systematically offsets changes in the government’s net debt position. Moreover, there are several theoretical frameworks in which government deficits do matter. Bianchard, Dombusch, and Buiter (1986), for example, present an intertemporal model in which deficit finance can make an important contribution to stability and growth. Suffice to say that deviations from the assumptions underlying Ricardian equivalence do suggest that debt management is important, particularly in light of its implications for financial stability. For an excellent discussion on the role of sound debt management policies in reducing the susceptibility of countries to contagion and financial risk, see the Guidelines for Public Debt Management prepared by the International Monetary Fund and the World Bank (IMF/World Bank, 2000).
3For a discussion on the possible real effects of public debt, see Lane (1997).
4In an extreme case of “unpleasant monetary arithmetic,” monetary policy is fully determined by the debt dynamics; see Sargent and Wallace (1981).
5In most industrialized countries, inflation and depreciation are the most likely alternatives (see Giovannini and Piga, 1992). In developing countries, the risk of default presents a more realistic possibility (see Dooley, 1998).
6While the focus of this chapter is on addressing public costs on an ex post basis, this should not be seen to undermine the importance of the ex ante containment of the public costs arising from systemic bank restructuring programs.
7For a broader discussion on issues concerning the strengthening of public debt management, see IMF/World Bank (2000).
8Under the existing guidelines, the fiscal balance is affected only by financial assistance that involves cash operations by the general government. Noncash expenditures and quasi-fiscal operations associated with bank restructuring are not accounted for in the overall fiscal balance. Thus, for example, recapitalization via the issuance of government bonds to troubled banks affects the standard fiscal balance only through interest payments, not through the principal, even if the bank immediately sells the debt. Likewise, in circumstances where the government issues bonds to compensate the central bank for past liquidity support, or where the bank restructuring agency or deposit insurance agency issues government-guaranteed bonds for bank restructuring, then only the interest accrued by these bonds (and not the principal) would be included in the budget. Furthermore, quasi-fiscal operations such as the extension of credit to troubled banks by the central bank or other public agencies (e.g., bank restructuring agencies or deposit insurance funds) are either entirely excluded from the government budget or their inclusion comes indirectly. If, on the other hand, governments were to float the debt and transfer proceeds to ailing banks and other financial institutions, that debt would be classified as government budgetary expenditure. A new manual, the 2001 Government Finance Statistics Manual, which is likely to take several years in many countries before it is adopted, takes an accrual-based approach to the recording of government transactions which therefore addresses the issues raised here.
9For a detailed discussion of the augmented fiscal balance approach, see Daniel, Davis, and Wolfe (1997).
10In practice, scenario specification should involve assigning a variety of different values for a, based on different estimates of recovery values on core assets—ranging from pessimistic to optimistic estimates.
11In practice, scenario specification should involve assigning a variety of different values for γ based on different estimates of recovery values on noncore assets—ranging from pessimistic to optimistic estimates.
12The model showing details on the evolution of debt and debt-servicing costs, under each scenario, is available from the author on request.

    Other Resources Citing This Publication