Financial Crisis Management

The key features of the Asian financial crisis and the ongoing international turmoil have been extensively documented.¹⁵ Suffice it to note that both episodes have introduced major challenges for policymakers in the region, even though the current turmoil has thus far had much smaller effects on the region's financial systems than the crisis of the late 1990s.¹⁶ As the current crisis is not over, however, some of the spillovers may still be felt and the situation may become more difficult for the region.¹⁷ Most notably, if growth remains subdued for an extended period, there is a risk of a gradual buildup in impaired assets, with associated strains on bank balance sheets; in addition, the recent rapid credit growth in a number of countries in the region may lead to a weakening of financial systems (Asian Development Bank [ADB] 2009a; International Monetary Fund [IMF] 2009a,b).

The epicenters of the 1997–1998 financial crisis were the financial systems of a number of Asian economies that had extended excessive credit without due regard to risk.¹⁸ Not only were the crises in these countries extremely severe, they were also multifaceted in that they frequently involved balance of payments problems, domestic and external debt crises, financial system distress, and, in some cases, political upheaval. In addition, the crises involved complex interactions and spillover effects within and across countries in the region. The costs of the crisis were extremely large, whether measured by (estimates of) the output losses incurred or the fiscal costs of financial sector bailouts, and by some measures not all the economies have fully recovered.¹⁹

Consistent with the pattern of many other emerging market crises, the Asian financial crisis waspreceded by huge inflows of private capital and was triggered by both a sharp pullback of capital (sudden stops and reversals) and market doubts about the sustainability of relatively tightly managed exchange rate regimes (currency speculation). Even more than 10 years after the crises, key features of the crisis management strategies adopted—most notably the use of so-called high interest rate defenses, the handling of some bank closures, and the resort to capital and other direct controls—remain highly contentious.²⁰ Also, in characterizing the policy responses in the worst affected economies (Indonesia, Republic of Korea (hereafter Korea), Malaysia, and Thailand), it is arguable, except in the case of Malaysia, which did not turn to the IMF for financial assistance, whether ownership of the crisis resolution strategies belonged to the countries involved or to the IMF.

The center of the ongoing turmoil is the United States (US) financial system (and those of some European countries) and the crisis was the result of excessive extensions of credit and leverage to finance booms in both real estate and consumption spending. Reflecting the role of the originate and transfer model of finance, extensive short-term funding by the shadow banking system, and the opaque nature of a number of the complex securitized products created, the spillovers from difficulties in one small sector of the US financial system (the subprime market) to the rest of the system and to other countries have been relatively large.²¹ In the early stages of the crisis, counterparty risk rose sharply in response to uncertainty about which entities were holding impaired assets that were hard to value, and many short-term funding markets effectively dried up (Bank for International Settlements [BIS] 2009; Brunnermeier 2009). Such market seizures led to an expanded role for the state as a market maker during the current turmoil in addition to its more traditional roles as a guarantor and lender of last resort (Buiter 2009).

As in the case of many emerging market crises, the US crisis was preceded by large capital inflows, albeit with official rather than private capital inflows playing a very significant role. Unlike many emerging market crises, the US crisis was not triggered by a pullback of these inflows.²² Rather, the triggers were spillovers from the problems that began to emerge in the real estate market in 2006 as the US housing bubble began to burst. Official capital inflows to the US remained strong or increased through 2009 as a result of the international reserve currency role of the dollar and its safe-haven status.²³ The bill is still rising on the costs of the crisis, but they are likely to be very high in terms of output foregone and government budgets.²⁴ Moreover, had the global financial system imploded late in 2008, the costs would have been considerably higher.

Whereas Asia was at the epicenter of the late 1990s crises, it has largely been at the periphery of the current turmoil. This is a reflection of the fact that the current crisis erupted in the US and the financial spillovers to the region thus far have generally been relatively small. This has been the result, inter alia, of limited direct exposure to the “toxic” products and institutions at the centre of the crisis, a general strengthening of conditions and risk management in regional banking systems, and much stronger external macroeconomic fundamentals (ADB 2008; BIS 2009; Adams 2008). Spillovers, however, have recently been increasing.²⁵ For the most part, the spillovers have been the result of real trade linkages and the high export orientation of many Asian economies. Only for a short period in late 2008, when international credit markets began to seize up, did foreign currency liquidity pressure become a significant issue for some countries in the region.²⁶ In line with global markets, however, regional equity and bond markets have seen sharp corrections. Looking forward, the slowdowns in growth in the region resulting from the turmoil will have effects on regional financial systems with systemic risk implications that will depend importantly on the “shock absorbing” capacity of the capital and other cushions in the region.²⁷

Reflecting the key differences between the late 1990s crises and the current turmoil, crisis managers in the region have thus far faced very different challenges during the two episodes. Most importantly, whereas the 1990s crises called for the adoption of wide ranging measures to manage collapsing financial systems in several countries, the key challenges during the current episode for the most part have involved taking preemptive measures to reduce crisis spillovers and efforts to keep credit flowing.²⁸ Exceptions are the cases of Korea and Indonesia, which had US dollar liquidity shortages in late 2008 that needed to be addressed, and Japan, where banking system exposure to falling stock markets has been high (BIS 2009). In addition, towards the end of 2008 there were some short-lived pressures in a number of interbank markets in the region associated with spillovers from the Lehman's bankruptcy.

A range of orthodox and unorthodox measures has been taken to stabilize advanced country financial systems during the current turmoil, and the state has assumed explicitly or implicitly an important market maker role (Buiter 2009). Some of the measures taken in the advanced countries—such as the resort to blanket guarantees and the nationalization of parts of financial systems²⁹—inadvertently may have added to pressures in the region by encouraging withdrawals of foreign currency liquidity. At the same time, especially in the case of countries in the region with a strong foreign bank presence, there were spillovers to domestic financial markets and institutions from distressed foreign financial firms, especially in the wake of the Lehman's failure in late 2008. As a result, officials in the region have been “living on the edge” responding to the potentially adverse consequences of spillovers from foreign banks and institutions, and responding to measures taken by the advanced countries to stabilize their financial systems (BIS 2009; IMF 2009a).

Consider now the approaches taken to the management of systemic crises. Even though we focus mainly on the current crisis and the Asian financial crisis, a number of references are also made to other crises. Most importantly, the lessons drawn about crisis management come from the full range of crisis experiences and hence include what are typically seen as the “successful” experiences in countries such as Sweden and the less successful experience of Japan following the bursting of its asset price bubbles (Boorman et al. 2000).

To facilitate exposition, the stages of crisis management are broken down into three phases. Even though there is necessarily some overlap between the phases, the breakdown facilitates the analysis of the key decisions that need to be made at critical points during crises and the roles of different instruments and state bodies. The three phases are: stabilization and containment, asset write-downs and absorption, and rehabilitation and normalization.³⁰ In making the breakdowns, the intention is not to suggest that crisis management is a clean and structured process that proceeds in clearly defined stages. Nor do most crises necessarily move step by step along a clear straight-line road map. By their nature, financial crises are messy and disorderly with many unpredictable developments (Hoelscher and Quintyn 2003). Every crisis, however, involves a number of critical stages during which certain key decisions need to be made and the breakdown facilitates the consideration of the key decisions.

Stabilization and Containment Phase: This is the first and, arguably, the most important phase for systemic crisis management. Except in those instances where a major crisis erupts completely unexpectedly, crises frequently begin with the emergence of pressures in particular segments of financial systems or in particular financial institutions (the “tremors”). Hence, for example, in the case of Thailand's 1997–1998 financial crisis, difficulties emerged as early as 1996 in a number of local finance companies that had extended excessive credit to the commercial real estate market. Korea began to see pressures in some of its merchant and commercial banks early in 1996 and 1997, well before its major crisis erupted later in the latter year (Boorman et al. 2000). Alternatively, in the case of the current turmoil, the warning signs began to emerge in the sub-prime segment of the US real estate market in 2006 as default rates began to increase (ADB 2009; IMF 2008a, b). Even though the emergence of these problems can be seen ex post as the starting points for the full-blown crises, they are frequently not seen in such terms at the time they emerge. Either because the sectors facing difficulties are seen as too “small” to have large spillovers, the excesses in these sectors are not judged to be symptomatic of wider problems, or as a consequence of political unwillingness to confront reality (the “denial” phase) a relatively common response is to regard the first problems as “isolated”. As I argue later, there is a need to address directly the shortcomings in systemic risk oversight that frequently led to a failure to correctly assess the systemic risk implications of the early “tremors.”

At some point, the difficulties in particular segments of the financial system no longer remain contained. The crisis becomes systemic and an arsenal of policy and other instruments begin to be deployed to stabilize the financial system. Key policy and other policy instruments that can be employed at this stage include central bank liquidity support operations, partial and blanket government guarantees of deposit and other liabilities and, in some instances, the use of direct measures such as deposit freezes, bank holidays, and capital controls.³¹ In many instances, these measures are also complemented by government communication strategies that seek to calm the markets and convince them that the situation will be brought under control. And if balance of payments pressures accompany the financial crisis—as was the case during the Asian financial crises—the government may seek the assistance of the international financial institutions or regional partners (Boorman et al. 2000; Adams, Litan, and Pomerleano 2000).

Numerous issues arise with regard to the roles of different policy instruments to contain the crisis during the first phase, and there will be a need for a relatively high degree of coordination between various state agencies and the central bank in assessing and responding to the problems. Typically, central banks will play a major role as suppliers of domestic currency (and, perhaps, foreign currency) liquidity³² but finance ministries, supervisory and regulatory bodies, and deposit insurance agencies may also play important roles. Generally, liquidity support operations are at the forefront of the initial efforts and will be undertaken by the central bank implying that it absorbs the risk. Key operational issues include whether standing liquidity facilities are sufficient including whether stigma effects limit their use), the terms and conditions for the liquidity provided (including collateral and counterparty requirements), and whether the central bank is able to channel the liquidity to where it is needed in the system (Hoelscher and Quintyn 2003; Boorman et al. 2000). As evidenced during the current turmoil, some of the advanced countries, including the US, found it necessary to expand the counterparties eligible to use lender last of resort facilities as liquidity seized up in both the capital markets and the banking system.³³ And new facilities were created in some instances either to address stigma effects or to allow liquidity to be channeled to where it was needed.

Generally, the challenges associated with getting liquidity to where it is needed can be addressed under the flexibility provided by the “exceptional” circumstance provisions of many central bank charters and, as necessary, new regulation or legislation can be enacted.³⁴ Both during the current crisis and during the Asian financial crisis, central banks were generally able to provide domestic currency liquidity to where it was needed even though in some cases this called for the “flexible” use of existing facilities and, in some instances, the liquidity was provided to institutions not usually covered under lender of last resort. Generally, the technical ability to inject domestic currency liquidity is not a problem except, perhaps, in currency unions or where there are relatively rigid currency board type arrangements, which limit domestic currency issuance.³⁵

In practice, a key challenge in providing liquidity support is to ensure that it does not compromise the overall macroeconomic policy stance. What this means is that the central banks must have the tools to sterilize injections of liquidity to ensure that monetary and exchange rate policy are not compromised (liquidity recycling). The importance of sterilization depends on the scale of the liquidity support provided and the nature of the financial crisis. During the early stages of the current turmoil, major central banks appear to have been able to sterilize their liquidity support operations through a range of relatively standard monetary operations.³⁶

Conversely, during the Asian financial crises, sterilization operations were very important given external sector weakness and the so-called high interest rate defenses judged necessary to support the external sector (Adams 2003). A failure to fully sterilize liquidity support in Indonesia in the context of heightened concerns about the state of the banking system arguably was a major factor in fueling a collapse of the currency and a sharp increase in inflation in 1998.³⁷

Additional important issues in liquidity provision include the conditions under which liquidity is provided and when the support should be backstopped by other measures. Even though there appears to be widespread support for the so-called Bagehot principles for the lender of last resort function—lend freely at penal rates against good collateral—systemic crises frequently lead to departures from these principles. Even though prudence might dictate that liquidity support should only be provided to solvent institutions, it is very difficult in practice to distinguish between solvency and liquidity problems in the heat of a major crisis (Boorman et al. 2000). Very often, major crises tend to call for central banks to loosen their collateral (and hair cut) requirements and it may not be feasible to charge distressed financial institutions penal interest rates. Generally, as evidenced by the current crisis, the scale of central bank liquidity support can become very large and the credit quality of central bank balance sheets can deteriorate sharply, especially in circumstances where the central bank seeks to circumvent breakdowns in financial intermediation (BIS 2009).

Generally, countries do not rely exclusively on liquidity support, except in the very short run before there is time to introduce other measures. Key complementary or alternative measures are the provision of partial or blanket government guarantees to the financial system or the introduction of direct controls on withdrawals from the banking system. As these measures typically require changes in government policy and may have implications for taxpayers, it is normal for finance and other ministries to play a key role in their introduction. At the same time, critical input and cooperation has also typically been required from a range of supervisory and regulatory agencies and deposit insurance funds in order to help assess the solvency of different financial institutions.³⁸

Notwithstanding the existence of deposit insurance funds, a common response in many crises is the introduction of “blanket” guarantees that expand the scope of deposit insurance coverage. By providing confidence that deposits (and, perhaps, other liabilities) will be covered by the state, policymakers intend to avoid wholesale or retail bank runs and reduce the pressure on the lender of last resort function. All the countries at the center of the Asian financial crisis eventually introduced blanket deposit guarantees. In the case of Indonesia, the blanket guarantee was introduced very late following bank runs in late 1997 (Boorman et al. 2000). During the current turmoil, the United Kingdom was slow to impose a blanket guarantee for the banking system and only did so following the run on Northern Rock Several countries in Asia have used blanket guarantees during the current turmoil so as to limit contagion or in response to similar moves by their neighbors as discussed by the ADB Monitor (ADB 2008b).

Based on past experience, blanket guarantees have never been a complete solution and, even when they work to install confidence, can imply very high fiscal costs.³⁹ Moreover, in circumstances when they call into the question the sustainability of the macroeconomic framework, blanket guarantees may have the unintended consequence of shifting financial pressures into fiscal and external problems.⁴⁰

Key operational issues with blanket guarantees include the scope and breadth of their coverage (and, in particular, whether they should only cover certain bank liabilities), whether foreign as well as domestic banks should be covered, and the degree to which financial sector supervision is strengthened, or levies are imposed, so as to help prevent the covered institutions from misusing the guarantees. No one size fits all approach describes the range of actions taken during the current or earlier crises even though in broad terms—and especially during the current turmoil—there has been a tendency for blanket guarantees to be widely used and to be relatively broad in scope. During the current turmoil, some developed ountries have also guaranteed not only bank deposit liabilities but also some of the liabilities of nonbank institutions such as money market funds.⁴¹

Not only are substantive issues about how and when the guarantees will be scaled back raised, there are important implications for moral hazard. More than 10 years after the Asian crisis, both Thailand and Indonesia still had not fully removed their blanket guarantees even before the current turmoil erupted and depositors in these countries had limited incentives to monitor bank behavior.

As in the case of liquidity support, experiences show it is critically important that blanket guarantees do not compromise the overall macroeconomic framework. Given the potentially huge fiscal costs, blanket guarantees can easily raise questions of fiscal sustainability and, in the cases of heavily dollarized economies or where the guarantees cover external currency funding, may also threaten external sustainability (Collyns and Kincaid 2003). Based on these considerations, best practice in the case of low intensity financial crises has typically been to use blanket guarantees very sparingly, and to employ them only after losses have been imposed on the stakeholders of troubled institutions.⁴² Clearly, however, the situation has been much more difficult in the case of fast moving systemic crises; as noted, the tendency has been for blanket guarantees to be widely used during recent crises. It is possible that their widespread use during the current crisis—especially in the case of countries at the periphery—has been a necessary “tit for tat” to the adoption of blanket guarantees in other economies.

Even though blanket guarantees may be introduced to lessen the possibility of bank runs, their introduction can also play a potentially important stabilizing role by signaling that the state is prepared to absorb at least some of the financial system losses. As a result, guarantees can represent an initial critical stage in the process of allocating and distributing losses across different stakeholders and making clear that particular claimants (such as depositors) will be protected. Needless to say, the earlier blanket guarantees are introduced, the less likely it is that the state will have a full accounting of the scale of potential losses it will be covering and the fiscal implications. By drawing a line in the sand—and indicating that the state will cover some of the losses—blanket guarantees can, however, play an important stabilizing role provided the fiscal and other resources are judged to be there to cover the (as yet, unknown) losses. By signaling that certain deposits will be protected, the state, of course, is also making clear that other claimants, including capital owners, may not be covered.

Finally, should liquidity support and government guarantees not be sufficient to stabilize the financial system, the state may impose direct controls on the withdrawal of funds from financial institutions and, in extremis, impose bank “holidays.” Even in the cases of very severe crises, these have typically been used as measures of last resort not only on account of their potentially damaging effects on confidence and efficiency but also because they can be very hard to administer. Realistically, however, it is important to recognize that some direct controls are frequently used during crises—as was the case during the Asian financial crisis—but typically as a backstop to front line efforts to control the crisis. As discussed by Collyns and Kincaid (2003), such measures have been used on a number of occasions in Latin America, including in the Argentine crisis early this decade.

Drawing together the key elements, the stabilization and containment phase of crisis management includes potentially important roles for the state in terms of liquidity support, the provision of guarantees, and the possible use of direct controls. Arguably, the key challenges in this phase have been to strike the appropriate balance across measures, ensure the required degree of coordination across the relevant government agencies and the central bank, and ensure consistency within a sustainable overall macroeconomic framework. At the same time, many of the key decisions have required a robust understanding of the sources of systemic risk (which may be changing over time) and an assessment of how various measures affect the stability of the system. Two general conclusions that can be drawn are (i) that there is a tendency in most crises not to view the initial tremors as signals of a major crisis, and (ii) that domestic liquidity support and guarantees may be used very extensively as the first key line of defense. Arguably, a failure to diagnose the systemic nature of the tremors can delay the adoption of a coherent crisis management strategy and may have lead in some crises to delayed responses and an excessive reliance on liquidity support as in Indonesia in 1997–1998 (Boorman et al. 2000, Adams 2003).

Asset Write Downs and Absorption: Once the economy is stabilized (if not before), a comprehensive crisis management strategy is required to write down impaired assets, deal with troubled institutions, and, eventually, bring more capital into the financial system. Based on crisis experience, the modalities of such strategies differ widely across countries as a result not only of differences in legal and political systems but also the nature of the crisis (in particular, whether it is an external as well as a domestic financial crisis), whether the financial system is mainly bank or capital markets based, and the authorities' preferences as regards economic stability and growth. In addition, the macroeconomic health of the state and its access to domestic and foreign currency funding may also influence importantly the approach adopted.

One key shortcoming is that the state will often take a very long time to come up with a coherent and proactive crisis management strategy. This is arguably one of the key lessons that the IMF has drawn from its crisis management experiences (Hoelscher and Quintyn 2003; Boorman et al. 2000; Collyns and Kincaid 2003). And, typically, a key turning point in addressing the financial sector problems in many crises occurs when the authorities develop a comprehensive forward-looking crisis management strategy. Before such time, measures taken can be largely reactive and ad hoc in nature. Moreover, to the extent to which the problems in the financial system are not addressed, and banks and other financial firms are inadequately capitalized, there is an ongoing risk banks may “gamble for resurrection”, and that impaired asset problems may increase further.

From a narrow economic perspective, the optimal strategy to crisis resolution might appear to be one that involves taking the losses implied by the crisis as quickly as possible (hitting the floor) with a view to allowing for a speedy bounce back. For a variety of reasons, however, the process of absorbing losses typically takes several years—even in the success case of Sweden in the early 1990s⁴³—and the authorities invariably face numerous tradeoffs in resolving a number of complex burden sharing and collective action problems. Experience shows, however, that the fact that the process may take several years does not obviate the importance of coming up with a crisis management strategy as early as possible (Hoelscher and Quintyn 2003 Boorman et al. 2000).

Various government agencies and ministries are involved during the second stage and the role of the central bank in providing emergency liquidity support may start to be reduced. Alternatively, as has been the case during the current turmoil, central banks in some countries may assume a key role in allocating credit in response either to generalized credit crunches or to breakdowns in some aspects of the financial system “plumbing”. And there may be efforts by central banks to act as market makers. A number of central banks in Europe and the US have arguably played an unprecedented role in the current turmoil in seeking to circumvent the effects of perceived or actual breakdowns in credit provision. As a result, there have been large changes in the size and composition of central bank balance sheets during the current turmoil (IMF 2008a,b,c; BIS 2009).

The implementation of any crisis resolution strategy will generally take a relatively long time. Most obviously, it is difficult to move very fast in writing down impaired assets because there will typically be considerable uncertainty in the heat of the crisis about how much assets are worth (price discovery) and in finding buyers. And, at least in the case of major systemic crises, the value of virtually all assets may be impaired to some degree. Uncertainty about asset values arose during the Asian crisis experience in the late 1990s, which was predominantly a bank-based crisis, and continues to bedevil the management of the current crisis in the US and Europe where capital markets have been more important and the market (were it to be working) could play a potentially important price discovery role.

Conventionally, valuing bank loans is difficult because loans are typically held on bank balance sheets and are not traded (even if there is some potential to trade the underlying collateral) (Boorman et al. 2000). As the ongoing crisis in the US and Europe has demonstrated, however, even securities that are traded under normal conditions can be difficult to value in circumstances where market liquidity dries up and the models that have been used to value them are no longer working. In such circumstances, neither marking to market nor marking to model may be viable approaches.⁴⁴

Another reason it is difficult to move rapidly is the existence of negative feedback loops. As financial crises are inevitably preceded by excessive credit creation, there is a need for a period of deleveraging. Typically, deleveraging will require asset sales that will tend to depress asset prices, which will require further deleveraging and so on. A vicious knife-edge process can easily occur in which the process becomes destabilizing. Unfortunately, there are no easy solutions to this problem and, in practice, quick decisions based on limited information need to be made about how to navigate around the instability. In practice, it is not uncommon for crisis managers to value some assets at their medium-term fundamental levels rather than at distressed crisis prices as a partial solution to knife edge instability, and to not move too rapidly in requiring distressed asset sales. Alternatively, as during the current crisis, practices such as mark to market may be suspended with a view to helping limit the adverse feedback loops (IMF 2008a,d; BIS 2009).

The likelihood that the markdown of impaired assets will lead to some financial institutions becoming insolvent will also tend to slow down the process. Allowing all insolvent institutions to close immediately would be damaging for systemic stability, and the approach that is often taken is to only close down those whose “departure” would not create large adverse spillover effects. As I argue, below, such an approach—loosely described at too big or interconnected to fail—can create enormous moral hazard problems if the “owners” of these institutions are bailed out and are not somehow required to take losses. The difficulties underscore the importance of strengthening crisis resolution frameworks to allow for necessary interventions and allocations of losses to owners while keeping systemically important financial firms afloat until such time as they can gracefully be wound down (as discussed in the next section).

The process of valuing bank loans during a crisis can be extremely difficult and is usually assigned to the supervisors of banks working in close cooperation with bank management and staff. In many cases, asset management companies (AMCs) may also be set up to play a role in valuing bank loans as part of an effort to shift these assets off bank balance sheets (the “good bank-bad bank” approach).⁴⁵ Practice varies across countries and crises, and a key challenge has typically been to try to establish “fair” values for assets in stressed conditions. Given that banks and (private) AMCs will have different objective functions, price discovery can be very difficult when they are used to value assets (Boorman et al. 2000; Ingves (2009), Ingves, Seelig, and He 2006). This conflict can be avoided to some extent when intervened banks sell their impaired assets to government owned AMCs but it is then important to avoid situations in which the AMCs overpay for assets and implicitly help recapitalize the banks by the back door. As during other crisis resolution stages, a high degree of transparency and independence has been found to be desirable in valuing impaired assets as noted, in particular, by Boorman et al. (2000) and Ingves, Seelig, and He. (2006).

Considerable work has been undertaken to determine whether one particular approach to AMCs works better than others. With a view to having the market play as large a role as possible in asset valuation, some countries have relied on set ups in which the AMCs are private and bid for impaired assets at “market” prices.⁴⁶ Others have favored approaches in which the AMCs are set up by the state with clear bank resolution mandates. In some cases, AMCs have been centralized and in other instances, decentralized (Adams, Litan, and Pomerleano 2000). While unambiguous conclusions about which approach is preferable are generally not possible to make, it may be very difficult for fully market based solutions to be used when there are wide differences in view about the value of impaired loans.⁴⁷

As the ongoing crisis has demonstrated, the process of valuing securitized assets can be extraordinarily difficult in circumstances where there is uncertainty about underlying asset values. In such circumstances, private buyers may be encouraged to bid for such assets to facilitate price discovery through explicit subsidies on risk taking.⁴⁸ To this point, however, a proposed variant of such an approach that has been set up in the US—the private public partnership—has not yet been used and appears increasingly likely to fall by the wayside.

Following loan write-downs, banks or financial firms are often split into different groups according to their solvency and viability. Three main groupings are often identified: (i) firms that are adequately capitalized, (ii) those that are undercapitalized but which are potentially viable over the medium-term, and (iii) those that are both undercapitalized and nonviable on account either of a lack of a profitable business model or an inability to raise fresh capital in the market.⁴⁹ In some cases, as in the US and Europe during the current turmoil, the classification of banks and other financial firms might also be based on stress tests in which their solvency or need for fresh capital is determined by the government on the basis of their ability to absorb various shocks (IMF 2008b, 2009a; BIS 2009).

Generally, banks and institutions in the third “no-hope” group would be intervened quickly and, when systemic risk is not judged to be large, closed immediately, albeit with arrangements typically made to transfer their guaranteed deposit liabilities to a viable institution. Where systemic risk is judged to be high, some banks and firms in this group may be kept afloat for a while, however. Banks in the second group would typically be encouraged to raise private capital but if this is not possible may receive injections of official capital. Banks in the first group may escape official intervention. Of course, for these approaches to be adopted, the state must have authority to intervene in the troubled institutions. As illustrated by recent experience, the state will frequently have “special” powers to intervene in certain financial institutions such as banks but may not have the authority in the case of other financial institutions, even where they are judged to be systemically important. As I argue in the next section, the boundaries of crisis resolution regimes need to be expanded to allow necessary crisis resolution interventions in all systemically important financial firms.

As both the Asian crisis experience and the current turmoil have shown, there are a range of issues about the type of capital to be injected by the official sector (common or preferred equity⁵⁰), whether and under what conditions some kinds of debt may be converted into equity, and the exit strategies for the official sector from any temporary nationalizations (as discussed below). As in other areas, there is no one-size-fits-all answer. Injecting preferred equity can be optimal in situations where it is possible that common private capital may be encouraged into the sector; in some cases, preferred equity injections are made pari passu with private capital. At least until the current turmoil, best practice in the case of insolvent firms was generally to write common equity to zero before injecting any official capital. It is not clear that all countries have followed this practice during the current turmoil as national authorities in a number of countries have sought to follow private capital-friendly solutions to getting more capital into the financial system (IMF 2008a, b; BIS 2009).

Key operational questions also concern the basis on which institutions intervened by the state should be operated (and by whom) and the conditions that banks receiving public monies (preferred or common equity) should be asked to meet. Based on experience, best practice in the former area tends to suggest that intervened banks should be run on commercial grounds with a view to minimizing losses, and then privatized or closed down as soon as feasible without creating systemic risk (Collyns and Kincaid 2003). In some cases, independent managers with banking experience are put in charge of these banks, whereas in other cases, their supervisors may be put in control. Experience suggests that the intervened institutions should be protected from political influence and patronage.

Generally, injections of public capital into potentially viable banks—whether in the form of preferred or common stock—should be accompanied by time bound conditionality as regards restructuring and, in some cases, by the requirement to raise private capital. More controversially, some countries have sought to require recapitalized banks to undertake new lending. Experience suggests, however, that such requirements should generally be used sparingly as the resumption of new lending too quickly may delay the return to financial health. As evidenced during the current turmoil, however, there may be strong political pressure for banks that receive public monies to engage in new lending.

Putting these various arguments together, the development of a comprehensive crisis management strategy is the key component of the second phase of crisis management. Among the various elements of this strategy, absorption and firm intervention are, perhaps, the most difficult. The role of the state is critical at many points: creating mechanisms for valuing assets; identifying and sharing losses across assets and institutions (including in many cases using AMCs); making key decisions as to which institutions will allowed to survive, perhaps, with the support of public capital; allowing for the orderly exit of other institutions through mergers and acquisitions, downsizing, or outright closure; and in determining how, by whom, and under what conditions intervened institutions and those receiving public capital are to be run.

Based on crisis experience, there are several key lessons that can be drawn about the second phase. The first is the importance of coming up early with a consistent forward-looking crisis management strategy. Given the complexity of the challenges faced, the implementation of this strategy will necessarily take time but developing the strategy early has typically been critical. The second lesson is that many of the decisions in the second phase have significant implications for which entities will bear the costs of the crisis and the required taxpayer support. Accordingly, the processes and the criteria used to assess asset values, solvency, and viability should be transparent, fiscal authorities should be involved to protect tax payer interests, and the agencies running intervened banks must be protected from narrow political and financial interests. In the absence of transparency and independence, distributing losses may be seen as rewarding the politically connected at the expense of the unconnected, and the credibility of crisis management may be seriously compromised. Finally, a third lesson is the need for the authorities to have the legal authority to intervene as necessary in systemically important financial firms to both keep them afloat, if this judged necessary, while at the same time imposing losses on their owners and, perhaps, other claimants. In short, crisis resolution regimes must provide the authorities with the ability to intervene and manage any systemically important financial institution, whether it is a bank or nonbank financial institution.

Rehabilitation and Normalization: The third and final phase involves the ongoing restructuring and strengthening of financial firms and the gradual unwinding and withdrawal of the special measures taken during the earlier phases of the crisis (the exit phase). Even though the state will continue to play a critical role in this stage—determined, in part, by the speed and manner in which it pulls back—the nature of its role will tend to evolve and become less hands on as temporarily nationalized financial firms are reprivatized, nonperforming loans and assets held by state run AMCs are slowly divested, and the central bank steps back from its liquidity and market support operations.

Even though the structure of the newly emerged financial system will generally be determined by the market, it will also be influenced importantly by changes in financial structure as a result of state actions during the crisis. In addition, as was the case following the Asian financial crises in the late 1990, the state may develop financial action plans to help guide the direction in which financial systems develop including, for example, with regard to the degree of concentration in the banking sector as well as seeking to grow local currency financial markets and the supporting infrastructure.

Two key operational issues in the third phase concern the speed with which the state should pull back from its management of supported financial firms and how quickly impaired assets should be sold by state run AMCs. Even though there is a presumption that the moves in these areas should be rapid—related, inter alia, to the fact that governments are not generally good at managing financial firms and assets—there is generally a need to balance several competing objectives.⁵¹ And, in those cases where nonfinancial corporate restructuring is a key component of the effort, it has had to be recognized that such restructuring generally takes a long time to complete even under London Club approaches. Ultimately, it is the “quality” of the adjustments made that will determine the success of the restructuring efforts and “good” operational restructuring will necessarily take time (Adams, Litan, and Pomerleano 2000; Pomerleano and Shaw 2005).

Generally, states have sought to divest themselves of stakes in financial firms relatively rapidly. In those cases where firms have been temporarily intervened ahead of their orderly closure, the key steps have typically involved transferring some of their assets and liabilities to other institutions and/or mergers and acquisition of parts of their businesses with stronger firms. The argument for moving quickly in the case of these institutions is that their continued operation may create unfair competition for viable firms and may involve substantial taxpayer cost. Provided the firms can be unwound in an orderly manner, they are typically allowed to exit as soon as possible. Nationalized and viable banks will also need to be privatized, but the speed with which this occurs is also influenced importantly by how quickly their balance sheets can be cleaned up and by the availability of private capital. Moving rapidly can imply that these firms will be bought up at fire sale prices and can generate political backlashes.⁵²

How rapidly the assets held by state AMCs are sold to the market involves balancing several competing objectives. With a view to maximizing recovery value (and minimizing taxpayer cost) governments may choose to avoid selling assets quickly and wait for markets to recover (Ingves, Seelig, and He 2006). Such approaches were quite common in the wake of the Asian financial crises and led to many AMCs being in operation several years after the crisis.⁵³ On the other hand, to the extent to which the state AMCs are sitting on a large stock of assets, the effect may be to delay the bottoming out of asset prices and their eventual recovery. Some balance is necessary between competing objectives.

Download this Paper [ PDF 189.4KB| 29 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.