This section discusses the resolution of banking sector problems in the context of partially dollarized economies.1 This issue emerged forcefully during the recent Latin American crises but had also been a concern in earlier episodes of distress. Specific constraints imposed by dollarization include the absence of a lender of last resort to respond to bank runs, which, in turn, contributes to exacerbating banking problems by increasing the speed and depth of deposit withdrawals. In many cases, the implementation of administrative measures was necessary to quell runs. Although allowing for some “breathing space,” these measures may complicate longer-term reintermediation efforts. A lesson from these experiences is that the prudential framework in partially dollarized countries should be adapted to address the additional risks. Also, policymakers in such countries should be mindful of the higher fragility of financial systems in the presence of partial dollarization when designing and implementing macroeconomic policies.
Before the onset of the recent Latin American crises, the banking sectors of the countries affected were generally seen as stable and efficient, in part because of the high share of U.S. dollar-denominated assets and liabilities in their financial systems. As the crisis unfolded, the general loss of confidence and the fallout from macroeconomic stop-gap measures led to rapid liquidity losses in several banking systems. Eventually both Argentina and Uruguay faced severe banking crises, and other countries in the region also suffered, although on a more contained scale. Once banking sectors had entered into difficulties, dollarization—which earlier had been seen as a sign of strength—turned into an additional complication. Many traditional means of addressing a banking crisis—including the provision of liquidity support and credible guarantees—were limited or ruled out. In addition, no clear “best-practice” rules on addressing a banking crisis of this nature were available.
This section suggests an approach for dealing with banking crises in dollarized economies that is based on country experience, gained both during the most recent crises in Latin America and in earlier episodes in different countries. It distinguishes between two related issues: addressing bank runs, and the longer-term resolution of banking sector problems, including achieving reintermediation after the immediate causes of the crisis have been addressed. The analysis concludes that dollarization—notwithstanding its well-known advantages in terms of fostering financial deepening and intermediation—can in financially turbulent times constitute both a source of vulnerability for banking systems and a major impediment to banking sector crisis resolution. It is therefore argued that the prudential framework in all dollarized economies, but particularly in those that are only partially dollarized, should be adapted to take into account the more risky environment. In addition, macroeconomic policymakers in dollarized economies should be mindful of the higher fragility of the financial system when designing and implementing macroeconomic policies.
Addressing Bank Runs in a Dollarized Economy
Stopping bank runs—the immediate manifestation of severe banking problems—in a nondollarized system is the first topic examined. The modifications to policy responses that might be called for if the banking system is dollarized are treated afterward.
Background and General Policies to Stop Bank Runs
Bank runs occur when depositors fear that money left with a financial institution or the banking system as a whole is subject to illiquidity or confiscatory risk because deposits may lose either their immediate availability or their value in real terms. In a local bank run, only one bank or a relatively small group of banks is affected, whereas deposits in other banks continue to be perceived as safe. A systemic bank run is a situation where all or most financial institutions of a country suffer large deposit withdrawals. In contrast to local runs, systemic runs are only partially linked to the solvency of the institutions concerned and reflect, to a large extent, macroeconomic events and changes in local and international investor sentiment.
Distinguishing between local and systemic runs is not easy. The treatment of bank runs, however, differs in important ways depending on their nature. In stable periods, and when a run is local, depositor protection can be provided in accordance with the deposit insurance policies already in place; emergency liquidity assistance to the bank in question can be made subject to strict conditions; and under-capitalized or insolvent banks can immediately be subjected to intervention and resolution.2 Nevertheless, even in these circumstances, the authorities need to be mindful of the potential for contagion of runs to other banks.
Addressing systemic bank runs requires a different approach, since policies generally followed in stable periods may aggravate uncertainties in a systemic crisis, worsen private sector confidence, and delay recovery.3 With many banks affected, it is more difficult to identify the true condition of banks. In these circumstances, measures taken to address bank runs should be consistent with achieving the following immediate and medium-term goals: (1) limit the loss of depositor confidence; (2) protect the payments system and a minimum of financial service provision; (3) restore solvency to the banking system; and (4) prevent worsening macroeconomic instability as a result of the bank runs.
Keeping in mind the above principles, and on the basis of past experience, a strategy to address systemic bank runs would be based on the following steps:
-
Accommodate bank liquidity needs in the short run. Emergency liquidity should be provided to all banks in need, except clearly insolvent institutions, under a liquidity policy that applies uniformly to all banks.4 The liquidity support facility should be designed in a way to provide banks with incentives to first tap all other sources of liquidity, including liquidity injections from shareholders. To the extent possible, the central bank should try to sterilize the excess liquidity created.
-
Provide broad liability guarantees to strengthen depositor confidence. Difficulties in distinguishing between solvent and insolvent banks and uncertainties about the future course of the crisis can fuel generalized depositor and creditor runs. A blanket guarantee has proven successful in easing such fears. Requirements for successful application, however, include a credible medium-term fiscal program to back the guarantee. Public relations aspects of announcing the guarantee are also important for credibility. If a blanket guarantee is not feasible, the degree of depositor protection should be clarified (see below for dollarized economies).
-
Implement restrictions on the availability of deposits (administrative measures) if necessary to avoid uncontrollable runs. If provision of liquidity alone does not arrest the outflows, other measures limiting the availability of deposits may need to be considered. For example, a bank holiday can be effective as an emergency stop-gap measure to take stock of the problem and allow for the formulation of a comprehensive plan to address it. Other restrictions could be partial or full deposit freezes. Capital controls may be needed to avoid additional pressure on the currency, which could reinforce withdrawal pressures. Although such restrictions on deposit withdrawals are often unavoidable; they must be managed carefully. Outright rationing or deposit freezes greatly damage confidence in the banking system, put in question the respect for financial contracts, and may seriously imperil the operation of the payments system. They can, therefore, only be a measure of last resort.
-
Simultaneously with the above measures, identify the causes of the crisis and announce immediate steps to address these causes. Given time constraints, a quick analysis of the main determinants of the run is needed to formulate a response to the crisis. The strategy should be worked out as the emergency measures described above are put in place, be announced, and be initiated quickly to avoid worsening confidence in the banking system. At the same time, determined actions are needed to tighten macroeconomic policies and implement structural reforms, increase confidence in the overall policy framework, and help to contain the magnitude of the crisis.
-
Implement comprehensive banking resolution measures, supported by a confidence-enhancing communication strategy. Regaining confidence will critically depend on the announcement and implementation of a comprehensive bank restructuring strategy. Failure to address the restructuring needs of the banking sector adequately can trigger a reemergence of bank runs. A bank restructuring strategy would include the following elements: (1) a policy to address weak but viable banks (these banks would submit rehabilitation plans, including cash flow projections and plans for gradual recapitalization); (2) on the basis of triggers set on bank capital levels, insolvent banks should be immediately subjected to intervention and resolution; and (3) finally, if solvency is difficult to determine in the short term, and liquidity assistance has been extensive, intervention could be prompted by triggers set on access to liquidity support as a share of the bank’s capital.
Bank Runs in Highly Dollarized Economies
A high degree of dollarization can severely constrain the policy options available to address bank runs, since it imposes limits on the authorities’ ability to recognize and address emerging banking sector problems. The nature and extent of the constraints imposed differs to a certain degree if dollarization is complete and official—meaning the country officially uses a foreign money as its currency—or partial—when a national currency exists, but economic agents’ currency preference for at least some purposes lies with the dollar.
Policy Environment with Dollarization
Dollarization complicates the task of addressing deposit runs, with partial dollarization creating most challenges (Box 5.1). Uncertainty about the extent of the crisis may be greater in a partially dollarized system. With currency mismatches on the asset and liabilities side of bank balance sheets, rapid changes in the exchange rate may cause sudden changes in banks’ net worth that are difficult to detect promptly. In addition, dollarization may have an impact on credit risk if corporates that borrowed in U.S. dollars are unable to generate U.S. dollar earnings. The assessment of the degree of additional risk is difficult. As a consequence, supervisors may not be able to assess clearly whether a bank is illiquid or insolvent, complicating decisions about liquidity support and bank resolution.
Currency Reform to End Partial Dollarization
An economy is said to be “dollarized” when a foreign money is adopted as the official currency (official or full dollarization) or when the foreign currency is used side-by-side with the local money (partial dollarization). Examples of official dollarization in Latin America include Panama, Ecuador, and El Salvador; among partially dollarized economies Argentina (before the crisis), Bolivia, Peru, Paraguay, and Uruguay all have or had well over half of bank deposits denominated in U.S. dollars.
At the outset of a crisis in a partially dollarized economy, pressures may exist to address the particular problems of partial dollarization by switching all assets and liabilities into one currency. All financial contracts could be rewritten either in foreign currency (dollarization) or in domestic currency (de-dollarization). While both options have been implemented, problems are associated with either course, suggesting caution against reliance on currency reform as a first-choice instrument in crisis management.
-
Official dollarization (e.g., Ecuador in 1999), if implemented in a credible manner and accompanied by other measures addressing banking sector problems, may contribute to restoring depositor confidence. The feasibility of implementing a credible dollarization plan depends, however, on the general macroeconomic conditions in the country, in particular the availability of sufficient international reserves to avoid a destabilizing depreciation in the process of exchanging local currency cash and deposits for dollars. It also critically depends on the prospects for managing fiscal policy under the constraints imposed by dollarization. Once dollarization has been introduced, and even if the initial banking sector problems have been resolved, full dollarization presents many of the same challenges and constraints in addressing future bank runs as those faced by partially dollarized economies.
-
Implementation of a currency board (e.g., Argentina in 1991, Bulgaria in 1997), while closely related to official dollarization and subject to similar preconditions, is not equivalent to official dollarization. The main differences are that a currency board can operate with partial foreign exchange coverage, that a national currency is retained, and that—however unlikely—exiting the arrangement remains an option. In many currency boards, credibility of the national currency is not as high as that of the anchor currency, and partial dollarization (in the form of foreign currency deposits) continues to play an important role.
-
De-dollarization (e.g., Bolivia in 1982, Mexico in 1982, Argentina in 2002) removes the constraints on the central bank’s lender-of-last-resort functions. However, forced currency conversion will make any reintermediation in the financial system even more difficult to achieve, since depositors are prevented from currency diversification. Experience with past de-dollarizations show that, where the accumulation of new U.S. dollar deposits was allowed, dollarization reemerged. Where U.S. dollar deposits were banned in the long run, the policy fostered large offshore U.S. dollar holdings by residents. For example, in Bolivia the financial system eventually redollarized at an even higher rate, while in Mexico a substantial amount of savings left the country.
More important, dollarization severely limits the availability of policy tools. Dollarization imposes limits on the possible extent of liquidity support— there is no unlimited lender-of-last-resort facility in foreign currency. The knowledge of limits on possible liquidity support for banks experiencing runs in highly dollarized economies may make such runs more likely to occur and, once under way, harder to stop. Similarly, if the banking system is sizable, the government may well be unable to provide a credible blanket guarantee for foreign currency deposits. Whereas in partially dollarized economies lender-of-last-resort facilities and a blanket guarantee may be provided in domestic currency, the relatively smaller domestic monetary base (compared with nondollarized economies) might lead to very fast and high domestic liquidity expansion and an even faster collapse of the exchange rate. In addition, the nature of the guarantee is inconsistent with depositors’ currency preferences, and it is likely to be less effective than a credible guarantee in foreign currency would have been.
Partially or fully dollarized economies also face important constraints on supporting macroeconomic policies to resolve a banking crisis. Dollarization makes monetary policy less effective, with interest rate changes less able to affect money demand. Exchange rate policy is also constrained. In a partially dollarized economy the exchange rate between the local currency and the U.S. dollar may be allowed to move flexibly, but substantial moves in this exchange rate can quickly create severe balance sheet problems for borrowers with liabilities in U.S. dollars and income in the local currency. As a result, banks will suffer, even if their portfolio has a notionally matched currency position. Finally, dollarization also complicates fiscal support for the bank restructuring process, since funds needed for any public recapitalization program may have to be committed in foreign currency to avoid asset/liability currency mismatches, which may not be consistent with a sustainable path of foreign-currency-denominated debt.
Alternatives to lender-of-last-resort and other safeguards—for example, the presence of foreign banks, prearranged credit lines, or sizable earmarked reserves—are likely to be useful mainly in the case of smaller banking sector problems.5 In the event of systemic runs in dollarized economies, these safeguards may not be enough: most important, the presence of a significant share of foreign institutions does not necessarily prevent systemic bank runs (Uruguay in 2002, Argentina in 2002). While the perceived support of the parent typically isolates foreign subsidiaries from a systemic run at the outset of a crisis, such support may not always materialize, particularly when the crisis proves deep and the authorities’ response slow, uncertain, or confiscatory. Under these circumstances, deposit runs are likely to extend to foreign banks. Prearranged credit lines may become unavailable in the time of most urgent needs, owing to strong built-in safeguards. Sizable earmarked reserves (such as used in Bulgaria) can stabilize expectations, but unless they extend to cover a significant portion of banking sector liabilities, there will be residual uncertainty and the perception of risk.
Modified Policies to Stop Bank Runs in Dollarized Economies
The additional constraints and limited role of safeguards in dollarized economies call for more attention to financial sector soundness. Where, in spite of such added vigilance, bank runs do occur, theory and country experiences suggest that a range of additional issues needs to be taken into account when formulating an appropriate strategy (Table 5.1 and Box 5.2).
Country Experiences with Bank Runs
Country Experiences with Bank Runs
| Country case | Deposit Dollarization | Trigger for Run |
Treatment of Dollar Deposits |
Treatment of Local Currency Deposits |
Outcome |
|---|---|---|---|---|---|
| Turkey (1994) | 46 percent | Sovereign downgrade and foreign exchange (f/x) depreciation | Full guarantee of all household deposits, and liquidity of up to 200 percent of capital available to banks | Same as for dollar deposits | Outflows from sound banks were halted by guarantee, allowing insolvent banks to be closed. But lack of supervisory follow-up left banking system vulnerable. |
| Argentina (1995) | 50 percent | Mexican crisis | Heavy, but ad hoc, liquidity assistance provided in dollars. | Liquidity assistance in domestic currency | Outflows diminished in due course, with ten smaller banks closed and ad hoc liquidity to others. There was deposit flight to quality. |
| Bulgaria (1996) | 40 percent | Deposit rationing by two banks | Closure of some banks, choice of full guarantee of f/x deposits in domestic currency or gradual payout in dollars over two years. | Full guarantee and liquidity in local currency | Outflow halted once a second wave of banks closures eliminated all bankrupt banks from the system and full guarantee and earmarked funds were in place for remaining banks. Banking system remained healthy without liquidity needs once currency board in place. |
| Russia (1998) | 18 percent (plus large off-balance sheet liabilities) | Unsustainable exchange rate and fiscal expansion leading to government debt market collapse | Household deposits at six major banks transferred to a state bank, but repaid in rubles at a fixed exchange rate. Liquidity provided only in rubles to the state bank and to other banks. | Same as for dollar deposits | The transfer of deposits to the state banks was an effective guarantee of almost all household deposits, but data do not show whether it stopped the run, and it created a state monopoly. |
| Ecuador (1999) | 76 percent | Failures of major domestic banks | The first attempt at resolution was a bank holiday followed by a deposit freeze and temporary securitization. When the run re-emerged as the freeze was lifted, formal dollarization took place. | Same as for dollar deposits | The deposit freeze halted the outflows initially, but social pressures caused the freeze to be gradually relaxed ahead of schedule. Further runs occurred as the easing took place, leading to renewed currency crisis and government default. Formal dollarization eventually was successful in easing deposit outflows. |
| Argentina (2001) | 70 percent | Exchange rate and government debt concerns | Partial deposit freeze, followed by forced de-dollarization and maturityextension. Voluntary securitization of 12 percent of deposits. | Similar partial deposit freeze and heavy liquidity assistance after de-dollarization | Social unrest and sharpened recession. Leakages occurred under the deposit freeze, partly from legal challenges. |
| Uruguay (2002) | 85 percent | Argentine crisis | Domestic banks: dollar liquidity to sight deposits, maturity extension of time deposits. Foreign banks: headquarters liquidity in case of need. | Domestic banks: liquidity to all deposits; foreign banks: headquarters liquidity in case of need | Outflow halted once 100 percent of unrestricted deposits covered by earmarked funds. |
Country Experiences with Bank Runs
| Country case | Deposit Dollarization | Trigger for Run |
Treatment of Dollar Deposits |
Treatment of Local Currency Deposits |
Outcome |
|---|---|---|---|---|---|
| Turkey (1994) | 46 percent | Sovereign downgrade and foreign exchange (f/x) depreciation | Full guarantee of all household deposits, and liquidity of up to 200 percent of capital available to banks | Same as for dollar deposits | Outflows from sound banks were halted by guarantee, allowing insolvent banks to be closed. But lack of supervisory follow-up left banking system vulnerable. |
| Argentina (1995) | 50 percent | Mexican crisis | Heavy, but ad hoc, liquidity assistance provided in dollars. | Liquidity assistance in domestic currency | Outflows diminished in due course, with ten smaller banks closed and ad hoc liquidity to others. There was deposit flight to quality. |
| Bulgaria (1996) | 40 percent | Deposit rationing by two banks | Closure of some banks, choice of full guarantee of f/x deposits in domestic currency or gradual payout in dollars over two years. | Full guarantee and liquidity in local currency | Outflow halted once a second wave of banks closures eliminated all bankrupt banks from the system and full guarantee and earmarked funds were in place for remaining banks. Banking system remained healthy without liquidity needs once currency board in place. |
| Russia (1998) | 18 percent (plus large off-balance sheet liabilities) | Unsustainable exchange rate and fiscal expansion leading to government debt market collapse | Household deposits at six major banks transferred to a state bank, but repaid in rubles at a fixed exchange rate. Liquidity provided only in rubles to the state bank and to other banks. | Same as for dollar deposits | The transfer of deposits to the state banks was an effective guarantee of almost all household deposits, but data do not show whether it stopped the run, and it created a state monopoly. |
| Ecuador (1999) | 76 percent | Failures of major domestic banks | The first attempt at resolution was a bank holiday followed by a deposit freeze and temporary securitization. When the run re-emerged as the freeze was lifted, formal dollarization took place. | Same as for dollar deposits | The deposit freeze halted the outflows initially, but social pressures caused the freeze to be gradually relaxed ahead of schedule. Further runs occurred as the easing took place, leading to renewed currency crisis and government default. Formal dollarization eventually was successful in easing deposit outflows. |
| Argentina (2001) | 70 percent | Exchange rate and government debt concerns | Partial deposit freeze, followed by forced de-dollarization and maturityextension. Voluntary securitization of 12 percent of deposits. | Similar partial deposit freeze and heavy liquidity assistance after de-dollarization | Social unrest and sharpened recession. Leakages occurred under the deposit freeze, partly from legal challenges. |
| Uruguay (2002) | 85 percent | Argentine crisis | Domestic banks: dollar liquidity to sight deposits, maturity extension of time deposits. Foreign banks: headquarters liquidity in case of need. | Domestic banks: liquidity to all deposits; foreign banks: headquarters liquidity in case of need | Outflow halted once 100 percent of unrestricted deposits covered by earmarked funds. |
Bank Runs in Dollarized Economies: Lessons from Experience
A review of experience with bank runs in several fully or partially dollarized economies suggests some important lessons for bank runs in dollarized systems.
-
Measures directly aimed at assuring the liquidity of banks (or restricting the outflows of deposits) stem runs. In contrast, attempts to halt deposit outflows through macroeconomic measures without assurance of adequate liquidity merely saw outflows accelerate (Argentina in 1995 and 2001, Uruguay in 2002).
-
Once emergency liquidity is being provided, a credible blanket guarantee, backed by sufficient reserves, can play a crucial role in arresting runs. Deposit outflows were successfully halted by deposit guarantees (Russia in 1998), and by both liquidity and deposit guarantees (Turkey in 1994, Bulgaria in 1996, Uruguay in 2002).
-
Where a blanket guarantee cannot be given for all bank liabilities, restrictions on the withdrawals of the deposits may be necessary as a last resort. Turkey (1994, 2002) and Bulgaria (1996) provided blanket guarantees on all deposits. Uruguay (2002) and Russia (1998) upheld existing deposit insurance arrangements, which guaranteed only some liabilities, and resorted to restrictions on many of the other liabilities. Ecuador and Argentina (2001) resorted to deposit restrictions.
-
To limit the loss of confidence associated with deposit restrictions and to minimize disruption to the payments system, transaction accounts should remain liquid. In Ecuador and Argentina (2001) comprehensive deposit restrictions rapidly came under social pressure, which forced a disorderly relaxation. In Russia (1998) and Uruguay (at least to date), more selective restrictions have proved to be more enforceable, in part because they were less disruptive to the payments system.
-
Effective supervision—especially a good off-site system—can help the authorities to focus their assistance on viable core banks within a comprehensive bank restructuring strategy. Turkey, Bulgaria, Russia, and Uruguay all provide examples where a comprehensive approach to intervening against non-viable banks and providing liquidity to viable ones proved effective, after piecemeal approaches had failed to restore confidence. In Turkey and Uruguay, foreign-owned banks were able to provide their own liquidity, so the authorities were able to focus their assistance on domestic banks.
Provision of Emergency Liquidity
Liquidity support is a key element in addressing systemic runs. In a dollarized economy, U.S. dollar liquidity support will be limited by the availability of resources, and it may be necessary to have a more clearly defined strategy for providing liquidity. Dollar assistance can be provided so long as sufficient U.S. dollar resources are available. Sources of liquidity can include high international reserves prior to the run (e.g., Argentina in 1995); contingent facilities (swaps, repos); international financial support (e.g., Uruguay in 2002); or liquidity from owners and shareholders (e.g., Uruguay in 2002). Drawing down some of those resources can, however, have costs in terms of credibility and limitations to macroeconomic policymaking.
With limited U.S. dollar assistance possible, a decision needs to be taken whether to shift assistance in part or fully to local currency (e.g., Russia in 1998, Bulgaria in 1996, Ecuador in 1999). This option avoids the need for other administrative measures (see below) and circumvents constraints on the lender of last resort since the central bank can print local currency. However, the resulting expansion in the money supply and the currency mismatch are likely to result in a combination of loss of international reserves, exchange rate depreciation, and inflation.
Depositor Protection
Depositor protection, in particular through a blanket guarantee, is generally regarded to be of key importance in addressing systemic bank runs. In a dollarized economy, funding constraints emerge, similar to those discussed above in the case of liquidity support.
-
A guarantee of deposits in foreign currency can be effective so long as sufficient U.S. dollar backing is available. In some countries, most notably the transition economies, banking systems tended to be small enough that a full blanket guarantee of foreign currency deposits might have been an option. In most countries, limited international reserves and fiscal constraints rendered a blanket guarantee issued in foreign currency not credible.
-
A blanket guarantee in local currency is unlikely to instill depositor confidence, because depositors will demand U.S. dollars rather than the local currency counterpart. Moreover, the announcement of a local currency guarantee may aggravate the crisis if depositors fear that the banking system does not have sufficient U.S. dollar resources and, therefore, withdraw deposits and seek to buy U.S. dollars in the exchange market. Guaranteeing deposits in local currency at a fixed exchange rate (Russia in 1998) implies a “haircut” to depositors in U.S. dollar terms when the exchange rate depreciates.
-
One option for depositor protection would be to transfer deposits in failed banks to a bank (public or private) that is perceived to be sound and has significant foreign exchange holdings, backed up by a corresponding amount of central bank securities to ensure liquidity. In the context of a systemic run, however, such banks will exist only in rare cases.
Administrative Measures
Given the added constraints of dollarization for liquidity support and depositor protection, administrative measures are more likely to be needed. Measures mentioned earlier—securitization of deposits, extension of deposit maturities, or the imposition of bank holidays or other restrictions on deposit with-drawals—are also available in dollarized settings. Similarly, the general principles to minimize the fallout from such policies stressed for general bank runs (e.g., uniformity of policies applied to banks) continue to apply.
Considerations in designing and implementing administrative measures include the following:
-
Securitization of deposits (e.g., Argentina in 2002, Ecuador in 1999) may be the less damaging option for the depositors compared with deposit freezes or outright defaults, since a secondary market in such instruments could provide some liquidity for those instruments. Depositors may incur losses if the bonds trade at a discount and are sold for cash before maturity. There are a variety of modalities for swapping bank deposits into bonds, and in designing the strategy, but possible trade-offs need to be considered. For example, one choice involves whether bonds should be issued in the form of a government bond (as was done in Argentina) or in the form of a transferable claim on the bank. While in principle a claim issued by the bank might be preferable because it maintains a degree of “normal banking relations,” other concerns—including the banks’ conditions and the marketability of the security—could favor government bonds and, in some cases, might be overriding.
-
Securitization should be accompanied by steps to mitigate its impact on financial intermediation, on the payments system, and on economic activity. Specific steps could include the following: securitization can be limited to certain accounts; securities can be made transferable; the authorities can seek to stimulate a secondary market in bonds; bonds can be issued in sufficiently small denominations; and, if the bonds are government issued, they can be made redeemable at par for certain transactions (payment of tax arrears, purchase of government assets). Although early redemption will strengthen demand for such instruments and foster the secondary market, the extent to which early redemption can be used is limited by the issuer’s cash needs. To avoid circumvention and loss of credibility, there should be few exceptions from the general policy concerning securitization of deposits.
-
Extension of deposit maturities is an alternative to securitization. This option was implemented, for instance, in both Argentina and Uruguay in 2002. While deposits with extended maturities are frozen in the banking system and not converted into negotiable instruments, measures should be considered to improve the functioning of the payments system under such a deposit freeze. For example, depositors could be allowed to write cheques on frozen deposits to give them full mobility within the banking system, and limited weekly/monthly withdrawals could also be allowed. In this way individuals and corporations with frozen balances could still carry out transactions. The authorities should be aware, however, that mobility within the system may lead to a flight to quality as depositors move frozen deposits to the strongest banks in the system.
-
Undefined limitations on deposit withdrawals without preestablished “rules of the game” (e.g., freezes without specified minimum weekly or monthly withdrawal amounts, or outright bank holidays) are an unstable solution to deposit runs and should be avoided. If established, they should only be in place for limited time periods, to buy the authorities time to work out a permanent solution.
Crisis Resolution and Reintermediation Under Dollarization
This subsection discusses lessons from country experiences for the longer-term resolution of a banking crisis in a dollarized economy. As in the previous subsections, it emphasizes how standard recommendations may need to be modified to take account of the specific challenges posed by dollarization.
Reversing the Causes of the Crisis
Country experience shows that financial reinter-mediation after a banking crisis, in the first place, depends on a reversal of the factors that caused the crisis.
Macroeconomic Stabilization
When macroeconomic imbalances were a major factor leading to a banking crisis, strong stabilization programs have often generated a swift market response. The stabilization program undertaken by Mexico (1988–94)—based on an exchange rate anchor, severe fiscal adjustment, and financial liberalization—led to a massive return of capital and a credit boom. The experience of Peru following the reform program of 1991–93 was similar. The program was sufficiently credible that intermediation levels increased by 8 percentage points between 1991 and 1996, albeit in U.S. dollars. Argentina too, after the introduction of the Convertibility Plan in 1991, experienced a 14 percentage point improvement in intermediation levels between 1990 and 1994. Even after the problems of 1994, following the election in June 1995 almost half of the deposits lost during the Tequila crisis returned within two months, albeit to stronger banks and in the form of U.S. dollar deposits. Poland stands out as an interesting case of rapid reintermediation in local currency. Implemented in 1990, a comprehensive stabilization program, with an important real interest rate differential favoring the zloty, produced almost immediate results. Already in 1990, the share of foreign currency deposits decreased from 63 percent at the beginning of the year to 32 percent at year end.6
At the same time, however, financial reintermediation will not endure if the improved macroeconomic environment is not sustained. In Mexico and in Argentina, for example, the financial reinter-mediations were relatively short-lived because the exchange rate anchor proved ultimately to be unsustainable, while fiscal policy was unable to offset the large emerging current account imbalances and increases in external debt. A number of episodes of rapid banking reintermediation, including Chile (pre-1982), Mexico (post-1988), Argentina (during the tablita years and during the 1990s), Ecuador (post-2000) and Poland (post-stabilization) were associated with large and continued real exchange rate appreciation in the context of an open capital account. By allowing a large positive spread between local interest and foreign interest rates to develop (thereby attracting capital inflows), without raising real interest rates (thereby encouraging lending), such real exchange rate appreciations seem to have been a key underlying factor behind these episodes. The risk, of course, is that the exchange rate appreciation and current account deficits could become unmanageable and, in time, result in twin currency and banking crises.
Restoring Currency Credibility
Financial reintermediation may be slow even after macroeconomic policies have become sound because of a continued distrust in the local currency. Restoring currency credibility can be a long and painful process in countries with histories of high inflation or repeated stabilization failures, regime changes, and episodes of acute financial repression, as in Argentina and Peru during the late 1980s or Mexico after the 1982 and 1995 crises. Thus, during the post-Tequila years, Mexican real interest rates remained quite high and have fallen only recently as the Banco de Mexico has started to make inroads in establishing its credibility.7
Countries have attempted to speed up the process of regaining currency credibility through several approaches, all of which have been associated with problems of their own.
-
Increased use of a foreign currency (usually the U.S. dollar). Although the use of the U.S. dollar can speed up reintermediation, it can also introduce important sources of tension. In Peru, for example, the use of the U.S. dollar as the prime currency for setting financial contracts appears to have had important short-term benefits because it greatly facilitated financial reintermediation. However, partial dollarization entails significant risks—as discussed earlier.
-
Moving to full dollarization. Although full dollarization is an internally more consistent arrangement than partial dollarization, it can also expose the financial system to greater near-term liquidity pressures and to longer-term stress. As the recent Ecuadorian experience demonstrates, full dollarization can provide the immediate credibility boost needed for bank reintermediation to occur swiftly. However, dollar backing of deposits remains low, leaving the financial system exposed to liquidity risk. The financial system is also exposed to longer-term stress in the event of significant deviations of the real exchange rate from equilibrium, due to weak fiscal management, shifting terms of trade, or large exchange rate adjustments by the country’s main trading partners (i.e., when the country’s trade is not mostly within a U.S. dollar zone).
-
Introduction of price-indexed instruments. When possible, the promotion of price-indexed instruments is a prudentially preferable alternative to dollarization in order to accelerate and deepen reintermediation, particularly as regards longer-term instruments such as mortgages. The main prudential advantage lies in the fact that price-indexed instruments increase the share of local currency in banks’ portfolios, and, hence, increase the degree of lender-of-last-resort coverage. In Chile after the 1982 crisis, the use of the UF (a unit of account indexed to the consumer price index) was instrumental in promoting rapid financial reintermediation. However, the development of price-indexed instruments needs to be accompanied by a decisive effort to stabilize (to avoid indexation in the real sector) and must be gradual, not the least because it must take place symmetrically on both sides of banks’ balance sheets.8
Unwinding Emergency and Administrative Measures
Notwithstanding their important potential role during a crisis, emergency and administrative measures may damage the longer-term credibility of the banking system. Their successful removal—in itself a complicated and possibly time-consuming endeavor—will be the key to lasting reintermediation.
-
Blanket guarantees, while playing a key role in stopping a bank run, must be designed in a way that reduces moral hazard. This includes planning up-front for the eventual replacement of the guarantee with a limited deposit insurance scheme, as well as resolving quickly weak and excessively risk-taking banks. Phasing out a blanket guarantee requires confidence in the stability of the system, which often makes it a drawn-out process. In addition, it requires strong leadership, since depositors will have a significant interest in retaining the guarantee in place. The difficulties are illustrated by country experience. In Korea there was considerable resistance to changes in the arrangement, and the upper limit of the partial guarantee was set at about W 50 million (about $50,000) instead of the original, pre-crisis W 20 million (about $20,000).
-
The timing of the removal of a deposit freeze, when the freeze has been introduced in response to the lack of credibility of a blanket deposit guarantee, is also crucial. Whereas a premature removal exposes the banking system to the risk of a new run, an excessively drawn-out process can harm confidence in the banking system and delay rein-termediation. Moreover, deposit restrictions tend to lose effectiveness quickly as market participants learn ways of circumventing them. Country experiences suggest that restrictions used to substitute for necessary policy adjustments to address the fundamental causes of crises cannot provide lasting protection. Thus, as macroeconomic stabilization proceeds, restrictions should be lifted quickly so as to minimize their negative effects on the economy and facilitate reintermediation.
Adapting Bank Restructuring and Prudential Arrangements
Whether the crisis was primarily induced by widespread bank insolvency or poor supervisory oversight or both, return of deposits can take place only in the context of measures to restore the profitability and solvency of individual banks and strengthen the supervisory framework. In addition to the short-term measures to address banking sector problems noted earlier, a longer-term restructuring strategy is needed that includes a number of confidence-enhancing elements. None of these elements by itself can restore a banking system’s health; each must be introduced as a component of a wider strategy.
-
Decisive action to resolve weak banks. To ensure credibility in the banking system, unviable banks need to be removed, while weak but viable banks need to be visibly strengthened. For example, in Korea the initial commitment by the authorities to solving the financial crisis played a key initial role in assuaging the fears of local and foreign investors. In Turkey, the bank audit and subsequent public recapitalization scheme, along with bold steps to intervene with problem banks, assisted in preserving confidence. However, imposing nominal losses on depositors to restore the financial health of the banking system should be considered only as a last resort. Although such measures could return banks to soundness in a technical sense, they are likely to have a traumatic impact on depositors because they fully undermine trust in the banking system as a store of value.
-
Improving prudential oversight. A visible strengthening of bank supervision and the legal and institutional framework for crisis management will support the credibility of the restructured banking system. Measures to specifically address the risk related to the limitations or absence of the lender-of-last-resort function in a dollarized system will be particularly useful to reassure those wishing to hold U.S. dollar deposits. Country examples where prudential measures proved important include Bulgaria, where the redesign of prudential norms, along with technical strengthening of the supervision department, contributed to a more credible supervisory environment. In Korea, the adoption of a resolution framework emphasizing a prompt corrective action approach and its quick implementation helped to assuage investors’ fears.
-
Revising prudential norms. Other specific prudential initiatives have also been used to underpin confidence building and improve risk management in a dollarized setting. To assure markets of the banking systems’ resilience to liquidity shocks in a dollarized environment, Peru has subjected foreign currency deposits to a marginal reserve requirement. Bolivia’s regulations differentiate liquidity and reserve requirements for foreign currency deposits by the maturity of the liability, so that the most “volatile” maturity incurs the highest liquidity requirements. Moreover, in managing foreign exchange risks, both Bolivia and Peru impose asymmetric limits on foreign exchange exposures, allowing banks to hold relatively more long than short foreign exchange positions.9 To contain credit risk, Chile imposed a restriction on U.S. dollar lending such that banks could only lend to firms in the tradables sector. To underpin local currency reintermediation, Korean regulations temporarily differentiated between residents and nonresidents such that, while both were allowed to have deposits in local and foreign currency, for residents accounts in local currency were not convertible into foreign currency.10 Such steps can help to reduce and manage the specific prudential risks from dollarization.
Reviewing the Role of Foreign Banks
The presence of foreign banks can contribute to the process of reintermediation on the liability side of banks’ balance sheets. While in many cases the initial increase in the market shares of foreign banks was purely the result of the lifting of entry barriers and the acquisition of failed banks, perceptions of quality and parent backing, and the introduction of new products, may have contributed to increase presence further. Deposit market shares of foreign banks increased after the 1980 crisis in Argentina, from 8.6 percent in 1980 to 15 percent in 1984, and increased further from 16 percent in 1994 to almost 50 percent in 2000. Although relatively small after the crisis in 1983, the role of foreign banks in Chile has grown to more than 50 percent of the market in the 1990s. In Korea, a deliberate effort has been undertaken since the crisis to enhance the role of foreign ownership and management of local banks. This is driven partly by the need to reduce the “relationship” banking that existed between local banks and large corporations and by the need to import risk-management expertise of large foreign banks.
The evidence on the contribution of foreign banks to reintermediation on the asset side of banks’ balance sheets is less clear. Whereas more financial savings may be available to the financial system for investment, foreign banks may be more risk-averse than domestic banks, which may have a better knowledge of the local market. The experience of Mexico after the Tequila crisis, for instance, shows that foreign banks may concentrate lending on the larger (often foreign) corporations, leaving smaller potential borrowers with limited access to credit.
Asset Reintermediation
The restoration of an adequate level of bank credit to the economy (asset reintermediation) following a banking crisis is a major problem, with dollarized systems facing the most challenges. In this context the implementation of comprehensive and credible programs to strengthen banks and their institutional environment are necessary conditions for lending to domestic borrowers to recover. Credit may, however, continue to be depressed, often because of banks’ reluctance to lend in the post-crisis environment or because of the lack of investment opportunities. Banks’ unwillingness to lend often reflects an adequate response to the poor shape of the corporate sector— that in some cases was an important contributor to the crises.
Failure to fully recapitalize banks has also been a major constraint to the resumption of lending. In such an environment banks need to recoup losses and restore their capital through higher spreads, which will limit the demand for loans. Empirically, this has slowed the pace of recovery in private lending in Turkey during the recent crisis, and in Mexico during the post-Tequila years. In both cases the interest spreads between deposit and lending rates increased sharply, as did the spreads between deposits and government securities.11 Thus, the failure or delays in full recognition and in measures to deal with banks’ capital losses upfront require that banks recovered their capital over time by “taxing” their customers.12 In Korea, the high market interest rates (over 20 percent) resulted in raising the cost of capital and in reducing the value of cash flows. Combined, these two effects weakened the balance sheets of borrowers and increased drastically the number of bankruptcies. The high interest rates resulted in a severe credit crunch in the first part of 1998 and in an overall drop in domestic demand.
Lingering problems in the corporate sector, including the failure to properly restructure bank loans and deal with insolvent or illiquid debtors (i.e., a banking debt overhang) has also been a major factor in delaying reintermediation. Hence, decisive action is needed to restructure nonperforming borrowers, both from a financial and an operational viewpoint. In dollarized economies hit by a twin currency and banking crisis, borrowers in U.S. dollars without an explicit or “natural” hedge (i.e., U.S. dollar revenues) are likely to be the worst hit, and their restructuring needs to be major. This could also include prime borrowers with access to external (foreign currency) financing prior to the crisis.
Finally, a depressed economic environment or unstable political conditions can of course continue to slow down the recovery of financial intermediation, even when banks’ financial conditions have healed.
Concluding Remarks
Dealing with banking crises in dollarized economies is both more difficult and subject to greater risks than in other cases. The absence of a lender of last resort in U.S. dollars has a potential to make dollarized systems more prone to bank runs, and bank runs more difficult to stop when they occur. A strategy to address runs cannot rely to the same extent on the provision of liquidity and on deposit guarantees, the two key elements identified for other cases. Consequently, addressing bank runs in dollarized economies needs to resort earlier and more frequently to administrative measures that restrict the availability of deposits, or else seek international support on a larger scale. However, even if these steps are carefully designed, credibility loss in stopping bank runs in dollarized systems is likely to be more serious than in other systems.
Longer-term reintermediation following a banking crisis is, in all cases, an “uphill battle” to reestablish lasting macroeconomic stability and credibility, but it may be even more of a struggle in dollarized systems. A reflow of deposits into the banking system will, in all cases, depend on a visible and credible strategy to overcome the initial weaknesses that led to the crisis. In addition, prudential and other safeguards will be needed to assure the public that the reemergence of the earlier problems is unlikely. In the case of dollarized systems, the reintermediation process is additionally hampered by the loss of confidence from administrative measures that may have been unavoidable to stop the deposit runs.
The added difficulties in dealing with banking crises in partially dollarized economies bear important policy lessons. Most significant, regulators and supervisors need to be mindful of the additional risks of dollarization, and the prudential framework needs to be adapted accordingly. This requires added safeguards in terms of higher liquidity and solvency margins, and proper systems to restrict the possible impact of exchange rate changes on banks’ balance sheets. Equally important, macroeconomic policies need to be mindful of the possibly lower resilience of dollarized financial systems.
Reference
Basel Committee on Banking Supervision, Task Force on Dealing with Weak Banks, 2002, Supervisory Guidance on Dealing with Weak Banks, Basel Committee on Banking Supervision Publications No. 88 (Basel, Switzerland: Bank for International Settlements, March).
The section benefited from inputs from Ralph Chami, Monica DeBolle, Peter Hayward, Gianni de Nicolò, Claudia Dziobek, Olivier Frécaut, Shogo Ishii, Chee Sung Lee, Carlos Medeiros, Zuzana Murgasova, James Roaf, Jorge Roldos, Alejandro Santos, and Rupert Thorne.
Resolution of individual problem banks in normal times is the subject of a recent report by the Task Force on Dealing with Weak Banks. See Basel Committee on Banking Supervision (2002)
The closure of a bank could be taken as a signal by the depositors in other banks that their institutions might also be closed, and the run could accelerate.
Liquidity may be provided in a variety of ways, including temporary reduction in reserve requirements; central bank rediscounts using either local or foreign currencies; central bank purchase of government paper held by banks; or central bank lending collateralized by illiquid assets.
The issue is also relevant in economies with currency boards, which face similar challenges.
The speed of stabilization and reintermediation in Poland may be explained partly by the historical context. High inflation rates in 1988–89 that led to the deposit movements had not been experienced since World War II, thereby making the stabilization program more credible than in countries with repeated crisis/inflation episodes.
As a result, the government has recently been able to issue seven-to ten-year peso-denominated, fixed-rate securities, opening the way for a deepening of banking intermediation, with eight-year fixed-rate peso mortgages starting to be issued by banks.
Thus, although the conversion from floating-rate mortgages to price-indexed mortgages in Colombia following the 1998–99 banking crisis did reduce borrowers’ exposure to interest rate risk, it magnified that of banks because most banks’ liabilities continued to be expressed in pesos at very short maturities or with floating interest rates.
In this case banks are exposed to an appreciation of the local currency but will gain from a depreciation.
Following macroeconomic stabilization and progress in banking sector restructuring, these restrictions were lifted in 2001.
In the immediate aftermath of a crisis, banks’ assets may comprise to a large extent government or central bank paper, which may be the only safe asset at the time.
Although authorities may want to issue paper at interest levels that would enable banks to offer adequate positive real interest rates on deposits (as recently done in Argentina), such strategies would need to be consistent with the overall fiscal envelope.