Moderator

José Viñals, Financial Counsellor and Director, Monetary and Capital Markets Department, IMF

Panelists

Christine Cumming, First Vice President, Federal Reserve Bank of New York

Hervé Hannoun, Deputy General Manager, Bank for International Settlements

Malcolm D. Knight, Vice Chairman, Deutsche Bank

Antonio de Lecea, Principal Advisor, European Commission

Georges Pineau, Permanent Representative of the European Central Bank in Washington, D.C.

Manuel Sánchez González, Deputy Governor, Bank of Mexico

The session explored the current state of the financial system and capital markets, prospects for a systematic move toward unwinding public sector support in the financial sector, and the existence of preconditions required to balance risks to financial stability during unwinding. Discussions highlighted how these might affect (1) the timing and speed of unwinding, (2) market functioning, and (3) asset prices.

Timing and speed of unwinding

Too soon or too late? Trying to gauge what this means in practice will be much more art and judgment than it will be science.

The dominant view was that it is too soon to be implementing exit strategies, but it is not too soon to be planning them. The view that it is too soon to tighten fiscal policy was questioned by some participants. Although it is too soon to tighten interest rate policy, central banks need to gradually restore normal liquidity provision and conditions while publicly keeping open the option to tighten policy at any time so as to avoid any perception of unconditional commitment.

Exit strategies should not be used as a means of going back to “business as usual.” Basic preconditions for durable and safe exit should include at least the following three elements: (1) verification that support is no longer needed by the intermediaries, as evidenced by decreased demand for it, access to funding in the markets, and reconstructed balance sheets; (2) macroeconomic stability guaranteed in terms of sustainable monetary and fiscal policies; and (3) sustained economic recovery.

Market functioning and impact on asset prices

The goal of the exit process is to allow markets to function, as opposed to the government being the funding market for financial institutions and strongly influencing their decisions, if not actually making those decisions. Many of these interventions are winding down by themselves because the initial pricing for each program (e.g., liquidity provision, asset relief, recapitalization) was set to be less expensive than the cost of funding at the most extreme part of the crisis but well above the level of normal market rates.

The banking crises of the 1990s provide relevant reference. The rehabilitation strategy employed by the authorities at the time had three elements: (1) banking institutions identify all their problems–in a financial income statement sense–and allocate them into work-out groups; (2) they raise capital and improve their liquidity; and (3) they present a credible, forward-looking plan. In the aftermath of the present crisis, the third element is still a work in progress; to create such a plan, financial institutions need to have a sense of what the regulatory and legal landscape is going to look like as well as be able to operate under good macroeconomic conditions.

In other words, the ability of financial institutions, markets, and the private sector to adjust to the authorities’ exit from their extraordinary initiatives depends fundamentally on the ability of both the authorities and the private sector to create an appropriate environment of macroeconomic conditions and restore market competition and discipline.

Some participants were more skeptical about the collective ability to plan but rather confident about the ability to respond to whatever happens as events unfold. In that context, there was a bias in favor of getting back as soon as possible to proper functioning of the markets’ price signals and mechanisms.

Role of macroeconomic and financial indicators

Policymakers need to look at real-world data, and therefore thinking critically about the properties of various indicators is crucial. The properties of economic activity indicators (aggregate and disaggregate) and measures of financial conditions (prices and quantities) might have changed during the crisis.

Moderator

Carlo Cottarelli, Director of the Fiscal Affairs Department, IMF

Panelists

Mitsuhiro Furusawa, Senior Deputy Director General, Ministry of Finance, Japan

Simon Johnson, Professor, Massachusetts Institute of Technology and Peterson Institute

Christian Kastrop, Deputy Director General, German Federal Ministry of Finance

Nigel Ray, Executive Director, Fiscal Group, Australian Treasury

This session emphasized the poor condition of the fiscal context in which the unwinding of public interventions will take place, especially in the advanced economies, where debt levels relative to GDP are expected to be 40 percentage points higher in 2014 than they were before the crisis. Moreover, that increase, unprecedented in peace time, will occur from a fairly high starting level (about 60 percent of GDP in advanced economies, compared with 16 percent of GDP in the United States in 1929). Further, consolidation efforts will also need to take into account long-term developments, such as the aging of the population and the costs associated with addressing climate change.

In regard to unwinding fiscal measures, participants emphasized four key issues: (1) no single policy objective will fit all countries equally well, (2) communication policies are important in anchoring expectations, (3) contingent liabilities must be managed in a transparent way, and (4) reversing the alarming trend of debt growth is a major challenge.

Determining the appropriate policy objective

Establishing long-term budget projections will be helpful in setting the targets toward which a country would seek to converge in the unwinding process. In determining the specific long-term fiscal policy target, one should take into consideration existing macroeconomic parameters, such as the saving rate of the economy, the exposure to current account shocks, and the sensitivity of the economy to global financial markets. Most important, long-term demographic factors are likely to take priority, as their prospective impacts dwarf those of public interventions in the financial sector. With regard to implementation, some argued for a flexible approach, allowing for changes in strategy when warranted by economic developments.

Communication policies

The fiscal strategy for the unwinding process needs to be well communicated so as to establish fiscal credibility and foster public support and market confidence. Communications should clearly specify the fiscal risks and contingent liabilities and how these will be managed. Eventual changes in the fiscal targets also need to be well explained to support the credibility of the strategy. The communication strategy also needs to be educational to properly anchor the expectations of groups affected by the unwinding strategy.

Contingent liabilities

Contingent liabilities must be accounted for transparently even if methodologies for doing so are still under development, otherwise the public and policymakers may have the impression that the risks are insignificant. For instance, assuming it is not fully addressed, the authorities may need to clearly recognize the too-big-to fail issue, have an assessment of its potential size, and possess sufficient fiscal flexibility to respond if the risks are materialized. Other solutions were also proposed, such as reform of the bank resolution regimes and guarantee fees that are commensurate with the systemic importance of the covered financial institutions.

Reversing debt trends

All panelists acknowledged that the rate of accumulation of public debt is presenting a challenge for unwinding and for long-term fiscal policy. In particular, the primary balance adjustment may need to be unprecedented in some cases to restore debt sustainability. However, there was some debate regarding the level to which debt ratios need to converge in the medium term. On one hand, restoring debt ratios to precrisis levels may have adverse implications for economic recovery given the substantial fiscal adjustment implied. Aggressive debt reduction may also not be politically feasible. On the other hand, solely ensuring debt sustainability may not be sufficient to provide flexibility to deal with future contingencies.

Moderator

Olivier Blanchard, Economic Counsellor and Director, Research Department, IMF

Panelists

Charles Bean, Deputy Governor, Bank of England

Matt Carter, Managing Director, Head of Sovereigns and Agencies, RBS

Global Banking and Markets

Dino Kos, Managing Director for Equity Research, Portales Partners, LLC

Haruyuki Toyama, Director General of Financial Markets Department, Bank of Japan

Edwin M. Truman, Senior Fellow, Peterson Institute

This session considered market factors key for a successful unwinding process. Among other issues, it explored (1) the role of market indicators that could guide an active process of unwinding, (2) technical aspects relating to the unwinding of liquidity provisions and guarantees, and (3) factors that could have an effect across markets and therefore would require close agency and cross-border coordination. The session also discussed interdependencies among various intervention measures.

How to unwind public support to banks

Unwinding public support to banks will be easier where policies were well designed in the first place. The demand for liquidity support from central banks is going to decline as markets recover, if it is provided at penal rates. In fact, there has already been some drop-off in the use of such facilities. Similarly, if funding guarantees are priced relatively expensively, the demand for them is going to wane as risk appetite returns to unguaranteed funding markets and unguaranteed funding starts looking more attractive.

It will be more problematic to withdraw liquidity support that was not provided at penal rates or was provided against illiquid collateral that central banks do not normally accept. Such support includes nonpriced guarantees, like those on retail deposits, that were introduced by many countries during the crisis. No obvious signals may appear to withdraw the support even as markets recover, and banks receiving it will have an incentive to say its continuance is essential.

How to deal with risky assets purchased by central banks

Different views were expressed on how to deal with risky assets purchased by central banks. Some argued that dealing with them should be treated as a fiscal operation: assets with uncertain market values could be transferred to the Treasury and replaced in the balance sheet of the central bank by government bonds. That approach could lead to a clearer demarcation of risk and allow the central bank to operate in a more independent way. However, there was no consensus on this approach.

How to raise interest rates in the presence of large excess reserves

Two broad strategies are available for exiting from monetary ease while large reserves are still in the system. The excess can be drained and then interest rates raised, as the Bank of Japan did in 2006. Alternatively, excess reserves could be kept in place while interest rates are raised along with the interest rates on reserves.

In the United States, one approach to draining excess reserves would be to sell the mortgage-backed securities purchased by the central bank during the crisis, but that could undermine the housing recovery. Another approach could be to execute reverse repos, although there may not be enough counterparties with adequate balance sheets for the amount of such transactions needed. Finally, there has been some discussion of the Federal Reserve’s issuing securities to drain liquidity, but the Federal Reserve currently does not have the legal authority to issue such securities.

What will the new financial landscape look like?

The current support facilities offered to the financial sector should be seen as representing a bridge to a new, sustainable equilibrium. Some business models developed before the crisis will certainly not be sustainable afterward, including those heavily reliant on short-term wholesale funding associated with securitization. Similarly, certain asset structures, such as some types of mortgage-backed securities and complex structured financial assets, may not be sustainable in the future without ongoing public support. It is important that policy does not end up supporting unviable business models or unsustainable asset structures unless there is a very clear market failure that needs to be corrected.

What indicators should govern the pace of withdrawal?

The pace of exit should be guided by several factors, including (1) macroeconomic indicators, (2) measures of credit risk, (3) the solvency of financial institutions, and (4) general measures of risk appetite. The macroeconomic indicators, such as growth, unemployment, and inflation, will provide the background to the exit process. Indicators on the state of funding markets and measures of credit risk, in particular the terms on which banks can issue unguaranteed debt, will also be critical. Credit default swap spreads are going to be important indicators; but CDS spreads may give a misleadingly benign impression if they reflect the market’s belief that the public sector will provide support if institutions get into trouble. Indicators related to the solvency of financial institutions, such as measures of expected losses and capital leverage ratios, would also be a key element in deciding the right time to withdraw support. Finally, general measures of risk appetite as reflected in financial markets will be important.

Exit from quantitative easing: The experience of Japan

The Bank of Japan had three types of tools to drain excessive reserves: sale of central bank bills, reverse repo operations, and outright sale of treasury bills. But it did not have to make substantial use of these tools, as its quantitative easing (QE) had been carried out through accumulation of short-term assets. Rather, the most difficult part of the exit for the central bank was determining how it would ensure that funds would be efficiently allocated among market participants after the direct provision of funds had receded—it was not sure whether the market would smoothly resume functioning after a five-year QE period in which the money market remained inactive. Moreover, market conditions had significantly changed: a series of mergers and acquisitions had significantly reduced the number of large financial institutions; market participants scaled back resources for engaging in money market transactions; and the central bank had changed its settlement mechanism from designated settlement to real-time gross settlement.

Moderator

Reza Moghadam, Director, Strategy, Policy, and Review Department, IMF

Panelists

Aerdt Houben, Director, De Nederlandsche Bank

Nigel Jenkinson, Advisor, Financial Stability Board

Simon Linnett, Executive Vice Chairman, Rothschild

Thomas D. Stoddard, Senior Managing Director, The Blackstone Group

Panelists extensively discussed (1) the need for a coherent exit strategy, (2) the sequencing of exit, and (3) the problem of systemically important institutions. Participants agreed that the current environment remains very fragile: short maturities, huge refinancing needs, poorly functioning secondary markets, and closing liquidity windows. Exit and unwinding therefore must be cautious and gradual, with the costs (political and economic) they will impose clearly calculated before the moves are initiated. Panelists also discussed the need for an appropriate framework for measuring the effectiveness of public interventions, acknowledging that it is not straightforward to balance their benefits and costs in terms of social welfare. It was noted that support measures in the crisis have added substantially to moral hazard.

Need for a clear and coherent exit strategy

Panelists emphasized that exit requires a coherent public policy regime rather than a case-by-case approach. Before it is executed, the blueprint for exit must reflect agreements on broad structural issues (for example, which types of banks will be saved?) and be transparent and credible. One of the most critical aspects of unwinding is the need to push the residual risk of acquired assets away from the central bank’s balance sheet. The remaining objectives need to be clearly spelled out.

By way of example, one of the panelists discussed prudential considerations related to privatizing nationalized banks, repaying government shares, and unwinding public asset protection schemes. He raised some practical criteria for financial institutions repaying public capital:

• The private capital in the operation must be of at least the same quality as the public capital it will replace.

• The capital ratio of a supported bank should be comfortably above the regulatory minimum both before and after the repayment.

• Financial institutions must have shown their ability to access equity and debt markets.

• Financial institutions should have taken measures necessary for their restructuring.

However, the apparent heterogeneity across countries highlights the need for flexibility in exit strategies across countries and limits the possibilities for developing a common rule for exit.

Sequencing of exit

Regarding sequencing, panelists suggested the need to take account of the interrelationships among different facilities and noted that the sequencing of exit from various public support measures is unlikely to be clear-cut. One problem is that the tools used for interventions are largely substitutes for each other, which makes it difficult to think about sequenced exit from them. Other panelists stressed the importance of a clear sequencing for issues regarding financial structure, unwinding asset support, and capital sales.

Also stressed by panelists was the need to factor in potential retail demand when planning the sequencing of exit. Doing so could create competition for the sale of assets. It was also suggested that market-motivated exits, where possible, should be encouraged.

Problems of dealing with too-big-to-fail institutions

Panelists admitted that no clear-cut policies exist for dealing with systemically important financial companies. One panelist indicated the need to set up a system in which individual institutions bear all of their burdens in the case of failure. But all agreed that no “silver bullet” is available to resolve the problem of systemically important institutions, and the issue remains in the forefront of the international regulatory agenda.