Objectives for government interventions and exit

There are three leading objectives governing support measures and exit strategy:

Admittedly, in terms of exit timing, there may be a trade-off between the second and third objective.

The current setting for unwinding of public interventions

The exit from public support is currently being designed while the banking sector is still fragile. Funding profiles are short and refinancing needs over the next two years are massive, but wholesale and securitization markets have not opened up sufficiently. At the same time, central banks are expected to gradually withdraw their nonstandard liquidity support. And on the regulatory side, the Basel Committee on Banking Supervision is expected to recommend raising the liquidity requirements for banks over the next couple of years. In all, banks face severe financing challenges over the near term.

Sequencing

The design of the exit strategy should jointly consider the various support measures, as they are to some extent substitutes and should not be seen in isolation. In any event, an exit from liquidity support should be given priority to allow a return to normal monetary policy operations and to avoid residual risks accumulating on central banks’ balance sheets. Beyond this, in order to provide a back-stop, guarantee schemes could be kept open for as long as major funding vulnerabilities remain. At the same time, higher pricing of these guarantees could limit market distortions. Ideally, the schemes would temporarily continue to exist without being drawn upon. In any event, an exit from these guarantee schemes should be market driven, pre-announced, and gradual.

With regard to the unwinding of support for solvency, a flexible, tailor-made approach should be pursued according to the nature of the specific support instrument used, i.e., capital injection, asset guarantee, nationalization, or bad-bank structure. Capital injections and asset guarantees may generally be easier to exit in the short term. In contrast, unwinding a nationalization or bad-bank structure is likely to take longer, as that often requires developing and implementing a new bank business strategy. In this respect, there are several examples of supported banks in the European Union that are undergoing far-reaching structural changes in their business model, in some cases pressed by the EU competition authority.

Criteria for capital repayment

The repayment of capital injections should be assessed from both a micro-and macroprudential perspective. In the Netherlands, four explicit criteria and one implicit criterion are applied:

An additional, implicit criterion is that repayment of state support should not accommodate a bank that wishes to exit for the wrong reason (e.g., to circumvent restrictions on its compensation policies).

Lessons learned

Recent experience points to several lessons in the design and unwinding of public support schemes. A first lesson is that the schemes should provide incentives for a timely and automatic exit, for instance through exit premiums that start low and increase over time. In practice, such premiums have had a material impact on banks’ enthusiasm to exit. Second, schemes should be flexible in order to allow for a tailor-made exit across support instruments, institutions, sectors, and countries. Indeed, speed of recovery and readiness to exit have been uneven across these dimensions. Third, exit programs need to be based on a thorough assessment of a bank’s business model and forward-looking strategy.

Principles guiding exit

There are a number of unexceptionable high-level objectives and desirable features that can be set out as guides to decisions on exit. In practice, however, tensions may exist between them (or even within them), and the challenge is judging how to balance the respective objectives and the associated risks. For example, all would agree that decisions on exit should be taken to support financial stability. But a balance still has to be struck between decisions to keep policies in place as a temporary backstop and the risk that such a backstop, if not very well designed, could lead to continued support for unsustainable business models, potentially sowing the seeds for future problems.

Less difficult is the principle that decisions should support market-based exit, for example by setting pricing incentives that lower the usage of support measures as markets normalize; and the principle that exit decisions should limit market distortions, spillovers, and arbitrage across borders. But there are potential tensions between the goal that exit strategies should be transparent and preannounced to give market participants time to prepare and the goal that they should be flexible to give the authorities the capability to respond to changes in market conditions. Again, in practice, a balance needs to be struck: to gain flexibility, announcements in some cases could be state-contingent—providing market participants with the information that will condition the policy decision—while still recognizing that decisions with preannounced timetables could sometimes yield benefits, given that they are likely to be easier to understand and thus to implement.

Timing and transitional issues

Decisions on the timing of withdrawal of support also require the weighing of uncertain costs and benefits. For example, timely repayment of public capital is desirable both to reduce financial market distortions and to lower fiscal risks. But prudential supervisors need to be confident that a bank repaying such support has a sustainable capital position, taking into account the higher standards that will be required in the future, and that the capital planning of the banking system collectively does not compromise aggregate credit provision. Moreover, the incentives of the individual banks may not be fully aligned with those of the authorities. For example, a bank may be particularly keen to exit the support arrangements to demonstrate renewed strength. But it may also be more prepared to take the risk that such an exit may be premature, relying on a too-big-to-fail backstop from the public authorities if it gets into trouble again. The moral hazard this engenders and indeed the crystallization of renewed failure are clearly outcomes the authorities wish to avoid.

As the financial system gradually heals, stronger banks will regain normal market access, while weaker banks may not. In its note for the November 2009 G-20 meeting, the Financial Stability Board judged that the case for systemwide support measures is diminishing given the improvements in recent months. In dealing with weaker banks, authorities face challenges in judging which are potentially viable as a whole or in part and ensuring that nonviable operations are resolved and wound down. There is a good case for separating the impaired assets from the healthy part of the bank to provide incentives both to manage the healthy business effectively and to seek maximum value from working out the impaired assets over time. There are, of course, risks that managing such assets down too quickly could lead to fire-sale externalities on the rest of the system; that needs to be reflected in the mandate of the management of the workout.

Cross-border coordination

Recognizing the differences in the strength of national financial systems, the optimum timing of withdrawal from support is likely to vary across countries. Nonetheless, there are gains from coordination, given that support measures and uneven exit decisions distort the allocation of capital across borders. There are clear gains to be realized from information exchange and stronger forms of coordination regarding support programs whose adverse spillover risks are highest, such as funding guarantees and exceptional retail deposit insurance measures. The agreement between Hong Kong SAR, Singapore, and Malaysia to set up a joint group to coordinate the exit from the exceptional retail deposit guarantees that expire at the end of 2010 is a good, practical example. Such coordination helps to resolve the potential collective action problem, in which countries may wish to end support but may be individually cautious about doing so unilaterally because of a fear of leakage of funding to other countries. Without coordination, the distortionary policies may remain in place, when in practice bringing the countries together and implementing a common decision would achieve the first-best outcome of collective exit.

Longer-term considerations

Finally, it is important to recognize that besides judging how and when to exit, authorities need to address the issue that such policies have substantially magnified the moral hazard distortions arising from institutions that are too big, complex, or interconnected to fail. Indeed, until this issue is addressed, financial institutions may well act as if they have an implicit backstop, even though the support measures have been formally withdrawn. The G-20 leaders have consequently charged the Financial Stability Board to work with member national authorities, international institutions, and standard setters to identify proposals, by October 2010, to address the moral hazard risks. That work is progressing on three broad fronts: policies to reduce the probability and impact of failure; policies to strengthen national and international contingency planning and crisis resolution tools; and measures to strengthen market infrastructure to withstand failure. There is unlikely to be a single “silver bullet” to lower the moral hazard risks associated with systemically significant institutions. But it is vital that we keep this key issue very much in mind in reviewing exit policies and the redesign of the regulatory system.

The nature of the intervention

Governments have adopted a range of interventions with respect to financial institutions. In generic terms, they might be summarized as follows:

• Supporting the capital of the bank—in some senses the “cleanest” since it does not involve any disturbance of the institution above shareholder level. This has been adopted by the United States and a significant number of countries in Europe, the Middle East, and elsewhere.

• Supporting the asset base of the bank, either by sharing the risk on certain asset classes (e.g., TARP in the United States or APS in the United Kingdom) or by taking certain toxic assets out completely (”good bank/bad bank” solutions as in Switzerland).

• Supporting the liabilities of the bank by guaranteeing deposits beyond the level that is considered conventional to support the retail market (e.g., the United States and many European countries) and providing state guarantees for debt issued by financial institutions (e.g., Denmark, Germany, Ireland, the United Kingdom, and the United States).

• Liquidity support by creating liquidity in the market, for example by governments setting up state-guaranteed special purpose vehicles (SPVs), which provide funding to financial institutions (e.g., France) and allowing riskier assets as eligible collateral for financing purposes by the central bank (e.g., the European Union, the United Kingdom, and the United States).

To some degree (particularly after the initial stages of the financial crisis), these policies were adopted as an integrated package; however, their unwinding can (and should) happen in stages across these different interventions.

Beyond interventions

In addition to their interest in unwinding interventions, governments will also wish to consider policy issues in relation to financial institutions generally, including those in which it retains an interest. Such issues might include:

• The on-going, viable structure of a particular financial institution, for example, the need to restructure it into a good bank and bad bank.

• The regulation of the sector, including pressure to enforce longer-term guidelines on governance, compensation, etc.

• The level of competition in the market place (particularly retail)—should state-supported bank X retain a Y percent share of its domestic market?

• Whether it wishes to recognize that there may well be systemically important banks that are “too big to fail” (and what does “failure” really mean?).

• The interaction between investment banking/trading and conventional banking; does any country wish to reinstate Glass-Steagall?

In addition, where the government retains a shareholding of significance, it may ask or require the financial institution to divest itself of certain noncore assets to resolve its own capital needs (in a not dissimilar way to the State Aid restructuring requirements that have been implemented by the European Union).

These are fundamental issues that need to be addressed not only on a country by country basis but also taking into account cross-border banking and capital market implications. For example, financial regulation should, ideally, be conformed across the global markets—where the IMF and the World Bank could play a role.

Unwinding

Reverting to the interventions themselves, Rothschild believes that there is a natural sequence for the unwinding:

• It is difficult to see how other unwinding steps can be taken until the financial institution is reasonably in control of, and has full understanding of, its own assets. So, the first element of an unwinding might sensibly be the unwinding of asset support packages (and, in the case of good/bad bank, the formal separation of such assets).

• Ending quantitative easing is also desirable early in the program; this is possibly best coordinated by (the relatively few) governments involved in the practice and may not necessarily involve the IMF other than as a broker of ideas.

• Ideally, too, liability guarantee schemes would have moderated toward normalized policies (e.g., retail deposits only up to just a modest level), although the greater risk probably lies in a premature phasing-out and historically the complete unwinding of such liability guarantees has typically taken five or more years.

• After the above steps, the sale of stakes in the capital of the institutions may begin.

These are general guidelines; indeed the next two paragraphs expand on how these last two might be seen in parallel.

Implementation

Rothschild strongly supports a policy initiative whereby the taxpayers of a country who have supported their financial institutions should be given an early chance to reinvest in the unwinding. In addition, we observe very strong appetite from retail investors for assets that have traditionally been seen as secure dividend generators. For example, in France, the 2005 initial public offering of EDF generated €6.7 billion in orders from 4.9 million private individuals. Similarly, the final sell-down of Telstra shares by the Australian government in 2006 generated A$7.8 billion in orders from 2.5 million individuals. Modern distribution techniques and communication media have made accessing such demand a cost effective option.

A common concern among senior politicians is that the targeting of the general public as investors relatively early in the unwinding exposes a financially less sophisticated audience to too much risk. At the same time, history teaches us that early sales in any unwinding generally bring significantly higher returns than the later ones. Rothschild therefore strongly supports the idea that retail investors might be invited to acquire government-backed securities exchangeable into the bank’s capital. Effectively this might take the form of a, say, three-year government bond, yielding no interest (bank deposits yield little anyway) but convertible into the financial institution’s shares at a premium to current market. No downside (if held to maturity), but potentially significant upside. Such investments could well be issued before certain other unwinding has been fully implemented, particularly the guarantees.

A program

Unlike the private sector, governments do not sell individual assets, they sell programs. It is arguable that a major offering of a well-known retail bank (or other financial institution) could create a swell of enthusiasm that could, of itself, do a lot to reinstate confidence. To address this and to make sure that each sale/unwind fortifies the next, governments should consider their assets as a package as well as looking at the individual characteristics of each financial institution.

If successful, it could be that the unwinding resuscitates a large element of the confidence that has been so seriously degraded by the crisis over the past few years. This, in turn, could help influence the question of timing referred to at the outset of this paper.

Measuring effectiveness

Regulators need an appropriate framework for measuring the effectiveness of public interventions in private companies. Too much of the discourse on this subject has centered on investment return analysis. More attention should be paid to a proper assessment of the net impact of the interventions on overall social welfare, as measured in terms of the net economic costs and benefits to society. The problem with the investment return approach is twofold.

First, because there was no market source of capital to fund some of these private company interventions (e.g., AIG on September 16, 2008), the imputed “subsidy” was arguably the entire amount of the assistance rendered. Put another way, the private cost of capital for these firms was infinite, since no one would invest on any terms. So comparing the investments made by governments with private market securities does not capture the whole story.

Second, the exclusive focus on investment returns has enabled the large Wall Street banks to take the position that they have repaid their debts to society by returning the government’s capital with a positive return (IRR) to the government. This misses the external costs to society of unbridled risk-taking and widespread financial failure. The public has borne this cost, restored these institutions, and is now rightly frustrated that bankers propose to enjoy the fruits of the public intervention through high profits and equally high bonuses. (Public monetary policy favoring low interest rates in the aftermath of the crisis also may have the effect of exaggerating this profitability.) The financial sector needs to be held accountable for its overall impact on the economy, which requires a broader analysis of the costs and benefits of public intervention.

Measuring the effectiveness of public interventions is a difficult econometric task of course, but important. It probably requires both a top-down and bottom-up analysis, looking at the impact on GDP and employment on the one hand, and interconnections between institutions on the other.

Impact on competition

European authorities seem to be much more concerned than U.S. regulators about potentially disruptive competitive effects of public interventions. The European Commission has acted on the concept of “too big to fail” by breaking up institutions and imposing penalties to compensate for receipt of public assistance. In the United States, the focus has been almost exclusively on addressing systemic risk and short-term job losses, regardless of second-order competitive impact.

Importance of public policy

Public interventions and exits must be executed within a coherent public policy regime rather than on a deal-by-deal basis. Perhaps too many bankers and not enough economists have been on the front lines of company-specific interventions. For example, efforts to shrink or split up AIG and Citibank have seemingly contradicted the simultaneous encouragement for the combination of Merrill Lynch and Bank of America. This apparent incoherence is explainable in light of the need to make decisions quickly in a crisis, particularly when existing statutes do not provide all the right tools for regulators to intervene effectively.

Nevertheless, regulators and legislators need to determine what the future regulatory environment looks like as soon as practical so that they and private companies can build a bridge from here to there. For example, it is clear that financial institutions that benefit from the public safety net should be required to hold a larger capital buffer. During the crisis, however, they were permitted to operate with relaxed requirements. Since private companies are making real-time decisions now in planning for the future, they need to know what target levels they will need to satisfy. Similarly, they need to know what the standards or limitations on size will be relative to the concept of “too big to fail.” It may be that institutions in that category are simply too big and need to be regulated more aggressively or forced to shrink in size or scope.