Charles Bean Bank of England

I have been asked to focus my remarks on two questions in particular. First, what macroeconomic and financial market indicators should guide the unwinding of liquidity support and funding guarantees? Second, are recent developments supportive of current plans for withdrawal? Before addressing these two questions, I would like to make three generic points about the withdrawal process.

First, the unwinding process will be easier—and indeed can be entirely automatic—if the policies have been well designed in the first place. Thus, the demand for central bank liquidity support should drop off as markets recover, if that support is provided at penal rates. Christine Cumming’s comments this morning about declining use of the Federal Reserve’s programs illustrate this point. Similarly, if funding guarantees are priced expensively, then the demand for them will drop off as risk appetite returns to unguaranteed funding markets. For instance, the insurance provided on newly issued bank debt under the United Kingdom’s Credit Guarantee Scheme is priced off credit default swap (CDS) spreads during the crisis period. Consequently demand for the guarantees has evaporated as risk appetite has returned to the market for banks’ debt.

It will, however, be more difficult to withdraw liquidity support that has been provided at nonpenal rates or against illiquid collateral that a central bank would not normally take in its operations, as the demand for support will remain even as markets return to normal. Similarly, it will be more difficult to judge when it is safe to remove nonpriced guarantees—the most obvious example being enhanced guarantees on retail deposits.

Second, support policies should be a bridge to a new, sustainable equilibrium. Some business models, such as those very reliant on short-term wholesale funding through securitization vehicles, and some asset structures, for instance for some sorts of mortgage-backed securities and complex structured finance assets, are unlikely to be sustainable without ongoing public support. Policy should not support these businesses or markets unless there is an identifiable market failure present. As the recovery proceeds, we will gradually get a better appreciation of which business models and which asset structures can survive (possibly in altered form) and which should be allowed to perish. But there is a danger that political economy considerations may delay, or even prevent altogether, the withdrawal of support for these unviable businesses and assets.

Third, and allied to this, it is important that governments and central banks as far as possible stick to commitments they have already made regarding withdrawal of support. The looming withdrawal of support should encourage banks to take preemptive action to strengthen their balance sheets by retaining profits or raising new capital. For instance, funding costs are generally lower for well-capitalized banks, so the imminent withdrawal of funding support should encourage banks to improve their capital base. I recognize, though, that this may have some adverse side effects, as it could also encourage banks to cut back lending in order to improve capital ratios through that route.

So, turning to the first of my two original questions, what indicators should govern the pace of withdrawal? To begin with, in principle one can have a strong economy and a weak financial sector or vice versa, but in practice a stronger economy leads to lower default losses and, other things equal, a stronger banking system. So the usual clutch of macroeconomic indicators, such as growth, unemployment, and capacity utilization provide a suitable backdrop.

Second, the indicators of the state of funding markets and measures of credit risk are important. In particular, the terms on which banks can issue unguaranteed debt are critical. So banks’ CDS premiums potentially provide valuable information. It is worth noting, however, that these may give a misleadingly benign impression if they reflect a belief that the public sector will always ride to the rescue of a beleaguered institution. Moreover, it would be inappropriate to delay withdrawal until the CDS premiums for all banks have fallen back to low levels, as some lenders surely should be encouraged to merge with stronger brethren or else exit altogether. The state of the securitization market will also be critical, though the problem here is to know what structures deal adequately with the underlying information and incentive problems that were exposed by the crisis and will consequently be viable.

Third, indicators of the solvency of financial institutions, such as expected losses and capital and leverage ratios, should be valuable. In particular, these considerations should be brought together through the implementation of rigorous stress tests against a range of extreme but not implausible scenarios.

Finally, general indicators of risk appetite in financial markets, including a range of asset prices, should also be factored in.

So, regarding the second question, are recent developments generally supportive of current plans for withdrawal? I have to say that progress looks promising. The worst downside risks have largely dissipated as growth has returned to most of the G-20 economies along with a reduction in the likely losses in banks’ banking books and some writing up of the value of assets in the trading book. Risk premiums on financial assets have fallen back, CDS premiums have returned to levels seen prior to the collapse of Lehman Brothers, and LIBOR-OIS spreads are back to precrisis levels. And the issuance of unguaranteed bank debt has picked up. The main problem is the continued closure of the securitization market, which raises the question of how the funding gap will be filled when the copious quantities of public support are withdrawn.

Matt Carter Royal Bank of Scotland

The question I will discuss is whether financial institutions and markets are able to deal with the unwinding as planned and what the mechanisms and incentives should be to effect such an unwinding. In my mind the first part of the question is the most important, and in order to understand it better, we need to look closely at how banks are funding themselves and how that may evolve over time.

I have spent a large part of my career as a bond syndicate manager, and the first question we would ask on any transaction was, “Who is going to buy it?” If we are to replace state sponsored and provided funding for banks, we need to be confident that private sector demand and capacity exists.

We are getting some mixed messages as we look at markets. On the one hand, banks’ use of the various government schemes is clearly coming down significantly, market measures of risk appetite among investors suggests a degree of normalization, and private sector sources of funding are reopening.

On the other hand, that is only one part of the story, and numerous challenges remain as banks seek to be fully funded on a stand-alone basis. The investor base for bank securities has changed dramatically, the future calendar for refinancing is huge as the duration of liabilities has shortened over the last two years, and banks face a number of difficulties and competing pressures as they seek to grow their deposit bases and long-term wholesale funding. Clearly the extent of these difficulties varies dramatically across countries and regions; the injection of capital by the German state of Rhine-Westphalia into the WestLB bank was a timely reminder that not every country and not every bank is at the same point in the cycle.

Use of government-guaranteed issuance schemes has fallen dramatically and on a month by month basis is down 60 percent from its peak. Nonetheless, we still see issuance under these programs—for example, the Skipton and West Bromwich Building societies’ issue in the United Kingdom; and Bank of Queensland, among others, in Australia. In fact, in some areas, issuance has been increasing over the past three months. Any discussion on incentive mechanisms will need to understand whether such guaranteed issuance is driven purely by cost relative to the alternatives.

We see a similar pattern across most of the central bank liquidity facilities—for example, the term auction credit of the Federal Reserve’s Term Auction Facility or the holdings of its Commercial Paper Funding Facility. All show material declines. But the volume of funding which needs to be replaced before utilization returns to precrisis levels is clearly huge.

If we turn to our traditional barometers of risk appetite, again we see positive trends: whether we look at LIBOR rates, repos, or bank credit default swaps (CDS). All of these suggest spread compression and normalization. But the events of the past two weeks regarding the Dubai World development are a reminder of the fragility of investor confidence. And we can see this fragility reflected in the sovereign CDS levels if we look at Greece and the growing concerns regarding its deficit, concerns clearly exacerbated by the events in Dubai.

So, how are the alternatives to state supported funding developing? The maturity of euro commercial paper issuance shows a clear, healthy upward trend. From a low of around 20 days’ maturity in September 2008 as the events around Lehman Brothers were unfolding, we now see a level of around 60 days.

We see a similar pattern in covered bond issuance, a trend supported in particular by the European Central Bank’s purchase program, which commenced earlier this year. But—and this ties in with my point on the volume of funding required—volumes are still only at pre-2006 levels and, based on our forecasts, are likely to remain there next year. Further, senior unsecured issuance by banks has similarly recovered, but, again, volumes are still not fully at precrisis levels.

We have also seen the ratio of guaranteed to unguaranteed debt shift in the right direction, but we can also see that, for most of the period since the Lehman Brothers bankruptcy, guaranteed deals have consistently been larger.

Right now the growth of the covered bond markets and unguaranteed markets is critical because, for the time being, two key markets for term bank funding are effectively gone. The first is the senior bank floating rate note market, where we can see the postcrisis collapse in volumes; and the second is the European asset-backed securities (ABS) market. European ABS issuance levels in 2008 and 2009 do not reflect sales to third parties; they are retained transactions to be used as collateral with central banks.

We have seen a significant shift in the investor demographic. The biggest buyers of longer-dated senior floating-rate notes (FRNs) were banks themselves, benefitting from the 20 percent risk weight under Basel I and using them for liquidity portfolios. An example from January 2007 is a €2 billion issue from a AA-rated European bank, in which one can see that nearly 70 percent of the order book was banks. Going forward, those bank liquidity portfolios will be invested in government bonds. Further, if we look at an order book for the AAA tranche of a UK transaction in prime residential mortgage-backed securities in late 2006, we can see that the dominant investor base was the structured investment vehicles or SIVS who are clearly no longer around. So the message is clear; banks need to cultivate a new investor base for term funding—the so-called real money community of asset managers, insurers, and pension funds—to fill this gap.

We can also see this manifested in the difference between the investor pattern for guaranteed and unguaranteed issuance. Consider the case of the order books for two U.S. dollar issues by the same large bank in the United Kingdom. What you may notice is that the guaranteed deal is dominated by banks, but also that there is huge concentration in the order book, with the top 10 allocations accounting for more than 80 percent of the deals. The average order size is large at 67 million. In some respects what is happening here is that the banks that have benefitted from an inflow of deposits and liquidity in the crisis have recycled that to other banks through the mechanism of the guarantee schemes.

In contrast, the unguaranteed deal is primarily driven by fund managers and, to a lesser degree, hedge funds. But notice also how granular the demand is, with an average order size of 23 million and more than 160 orders in the book for a transaction that is 40 percent smaller.

This is a key point—the investor bases for guaranteed and unguaranteed issuance are different and we need to be confident that as we switch off the guarantee schemes that the broader investor base is deep enough and can provide the volume of funding required.

It is worthwhile at this point to consider the incentives and disincentives for banks to use the guarantee schemes. Consider France, where the government guaranteed issuance was via the SFEF agency, and Spain. We can see that in France banks were issuing unsecured funding even at points where guaranteed funding came at much tighter spreads; but perhaps more interestingly, in Spain some banks are still issuing guaranteed paper even though covered bonds are pricing at much tighter spreads. The reality is, of course, that the covered bond market is not yet open to all issuers, and therefore an increase in the guarantee cost in and of itself is unlikely to change things. So, although markets are reopening, not all financial institutions enjoy the same degree of access.

Finally, a couple of words on the refinancing burden and other challenges. The redemption profile of bank securities is sizeable, and unsurprisingly, the average maturity of liabilities has gone down. We estimate that for European financial institutions, we see 2010 redemptions approaching €700 billion. This represents only outstanding securities and does not reflect usage of central bank facilities. Average duration has fallen from around six years to nearer to four.

Banks have numerous incentives to restore stand-alone funding. One that I have not touched upon is the proposed new liquidity buffers that will penalize short-dated funding—the negative carry of a large liquidity portfolio will be a significant cost to the business. But banks face other external challenges in trying to build wholesale and retail funding. Governments are competing issuers in wholesale markets and can also compete for retail deposits, as we recently saw in the United Kingdom with National Savings.

As highlighted earlier, insurance companies are an important target investor base. The new insurance regime in Europe, Solvency 2, will make it more costly for insurance companies to invest in longer-dated bank bonds. And finally, the fungibility of funding and liquidity between jurisdictions face potential constraints arising from the likely requirements regarding self-sufficiency of liquidity.

Dino Kos Portales Partners, LLC

What indicators are being used by the market to judge whether monetary authorities are approaching an exit from their ultra-accommodative policies? I will focus on the United States because of the Federal Reserve’s influence and, more important, because it’s the situation I am most familiar with.

To set the stage, what are central banks—and the Federal Reserve in particular—exiting from? Four instruments are currently being utilized:

• zero short-term interest rates;

• assorted liquidity facilities;

• high levels of excess reserves; and

• asset purchases.

Thus, I would define a true exit as one in which the Federal Reserve reverts to targeting a nonzero interest rate and exits the remaining mechanisms. The liquidity facilities, designed to become unattractive as money markets normalize, have been contracting for months and are slated to wind down. However, asset purchases have more than offset the reserve contraction that the reduction of liquidity facilities would otherwise have implied.

Exiting the remaining aspects—zero interest rates, high levels of excess reserves, and asset purchases—will be a significant event for the Federal Reserve and for markets more generally. What signposts are investors and traders using to assess whether the exit is approaching?

Answer: The ones the Federal Reserve has told investors to focus on. In its November 2009 statement, the rate-setting Federal Open Market Committee listed the following elements:

• measures of resource utilization;

• inflation trends; and

• stable inflation expectations.

Resource utilization: Capacity utilization is at very low levels, while unemployment has reached 10 percent. In short, the output gap is very large and should not press against capacity constraints anytime soon.

Inflation trends: Headline inflation has been negative for much of the past year. Core measures are running at around 1.5 percent.

Inflation expectations: Breakeven rates from inflation-protected treasury securities (TIPS) suggest that longer-term inflation expectations have reverted to levels observed before the crisis. Despite worries about “quantitative easing” and “printing money,” the fixed-income market has not, to this point, priced in a high likelihood of inflation that will rise beyond the Federal Reserve’s implicit target.

Clearly, and not surprisingly, the Federal Reserve is placing a strong focus on inflation. However, readings of actual inflation are likely to lag. Therefore, other indicators will have to supplement this list of inflation signals. One candidate is the behavior of commercial banks. Despite high levels of excess reserves, bank lending has contracted over the past year. However, at some point, a reversal of this trend should emerge and could be an important signal by suggesting that both the demand for credit and banks’ willingness to lend are reviving. This may signal a healing of the credit intermediation process (one of the important reasons for undertaking the interventions) and indicate that the economy is improving, since stronger credit demand will signal a need to finance growing receivables, inventories, and capital expenditures. In short, it may suggest the need to begin removing accommodation before inflation signals begin to turn amber.

What about logistical and operational issues posed by the exit?

The Federal Reserve’s balance sheet grew by virtue of its asset purchases. Hence, the simplest means of exit would be to sell those assets. What is the probability of that happening? It borders on zero. Why? The Federal Reserve has acquired more than $1 trillion of mortgage-backed securities (MBS). The purchases have pushed down longer-term rates in general and have compressed the MBS spread in particular. During 2009, the government (mostly the Federal Reserve, but also the Treasury) has been the only buyer of MBS. Selling would likely push both the risk-free rate and MBS spreads much higher. The adverse impact on the housing recovery could be significant. The authorities will not wish to risk that outcome.

The more likely approach for exit will be to use the securities as collateral in reverse repos. However, there are only 18 primary dealers, and they have balance sheet constraints. A broader set of counterparties will be necessary. Money market funds are an obvious choice. But this also has complications. The money market segment grew from less than $2 trillion in 2005 to about $3.9 trillion earlier this year—before shrinking by $650 billion over the past eight months. In other words, the money market fund business is subject to wide fluctuations, which suggests that the Federal Reserve may want to limit its reverse repos to this segment.

An alternative for the Federal Reserve would be to increase the “interest on reserves” (IOR), the interest rate it pays on excess reserves. That would push market rates higher. This method has the advantage of avoiding an abrupt drain, such as that in 1937, and gives the banks a longer period of adjustment to a world of lower reserve balances. However, there is a threshold question that has not been answered.

Assume the Federal Reserve raises the IOR to, say, 3 percent and leaves several hundred billion dollars of excess reserves in the system. Is that the same monetary policy as a 3 percent policy rate with minimal excess reserves? Are financial conditions the same even though in the first scenario the system stays awash with significant amounts of reserve balances?

Put differently, the IOR mechanism has surely affected the demand curve for reserves among commercial banks. What is the new equilibrium? How will the Federal Reserve know when to stop the draining process? How it navigates this process will have huge implications for markets, and more broadly, for the economy.

Haruyuki Toyama⁸ Bank of Japan

Given Japan’s experiences of exit from measures that dealt with the last Japanese financial crisis—including quantitative easing, blanket deposit guarantees, and capital injections—I would like to stress that there should not be any predetermined timetable or specific conditions for unwinding of public interventions. The timing, sequence, and modality of unwinding should be up to pragmatic judgments, including whether the financial sector can stand on its own without the measures in question, how probable it is that the financial sector will get into double-dip problems that would require resurrection of the measures, and whether the adverse impacts of the measures would outweigh their benefits. An exit strategy should differ from country to country, but it is my sense that the current stage of working out the problems of the European and U.S. financial sectors is where Japan stood in 1999 and 2001, respectively. For Japan, it took two to four more years to have a turnaround in its working out process.

For unconventional measures directed at acute symptoms, unwinding should not be postponed once the symptom disappears. In this spirit, the Bank of Japan has already decided on dates to terminate outright purchases of corporate instruments and other unconventional measures.

For unwinding directed at more basic policies, such as raising interest rates and repayment by banks of injected capital, the decision should take more time. On both fronts, resurrection of confidence in the financial sector should be a prerequisite for unwinding. One of the most important lessons from Japan’s financial crisis is that without a healthy financial sector, the normal transmission mechanism of monetary policy does not function. While the financial sector was in intensive care, monetary policy could do no more to underpin the economy than to support the financial sector’s recovery. Repayment of injected capital should require careful thinking, too. Merely dropping a hint regarding the expected timing of unwinding would accelerate deleveraging by banks, possibly resulting in another round of adverse feedback between the financial sector and the real economy. Basically, before the financial sector can emerge from the vicious cycle of deleveraging and the incessant revelation of credit losses, the completion of balance sheet adjustments in the corporate and household sectors should be in sight.

Another condition for regaining confidence in the financial sector is the existence of safety nets for depositors and the basic functions of financial institutions. In the long list of safety-net measures, effective supervision is of utmost important. On the other hand, although toughened regulations may be significant in calming the public’s rage over the government and the financial sector, it may risk accelerating deleveraging.

The injection of taxpayer money into the financial sector is unpopular everywhere in the world. In Japan, severe criticism of the injection of public money to resolve the Housing Loan Companies in 1996 caused undue delay in policymakers’ decision to move ahead with expenditures of public money to resolve the banking sector. As a result, another couple of years had to be wasted until 1998, when failures of large financial institutions left the government no choices.

A healthy condition of the financial sector affects a large number of sectors in the economy. In particular, premature unwinding will deal a blow to small and medium-sized enterprises, which do not have access to the capital markets.

A rise in interest rates will adversely affect the sectors that have been helped by the ultralow interest rate policy. In Japan, a substantial portion of mortgage loan borrowers have selected floating interest rates rather than fixed rates in view of the prevailing low short-term interest rates. If rates rise, many of those borrowers may find it difficult to repay the mortgage, which will put further downward pressure on residential real estate prices.

A rise in interest rates will also cause a reversal of capital flows into emerging economies and the commodity markets. Investors will feel safe in continued risk taking until major central banks, in particular, the Federal Reserve, decide to make an exit. It is unfortunate that industrial countries and emerging countries blame each other for shocks that materialize when the direction of capital movements is reversed. The problems are twofold. First, a disparity in the correction of exchange rates emerges. The burdens of a U.S. dollar depreciation are put on currencies that are not pegged to the dollar or managed through intervention. Those currencies essentially assume the role of an anchor, with adverse impacts on the export sector and price stability when deflationary conditions exist.

Second, risk taking in emerging market equities and commodities may have pushed up their prices beyond a level reasonably justified by their fundamental values. How these irregularities or imbalances will play out when major central banks change monetary policy is uncertain. The recent incident of the Dubai World dramatically revealed that markets were already nervous over excessive risk taking. I am not optimistic that major industrial and emerging market countries can agree upon a coordinated action to prevent further excessive risk taking or carry out orderly exits, as countries primarily calibrate their policies according to the conditions in their domestic economies. However, it is important for an institution such as the IMF to give warnings about movements in the markets that could give rise to another bubble and its subsequent bursting.

With respect to the external impacts of unwinding financial sector policies, I would stress only that recovery of confidence in financial institutions takes longer in foreign markets than at home.

Edwin M. Truman Peterson Institute for International Economics

There is a singular lack of consensus about what caused the economic and financial crisis of 2007–09. That there is a similar lack of consensus about how best to exit from the crisis should not, therefore, be surprising. Before I address the specific questions put to me by Olivier Blanchard, I would like to offer some general reactions to what I have heard at this conference.

First, the overall objective should be strong, sustained, and balanced worldwide growth. Financial sector repair is part of that process, but it is only one part. Nor is it the most important element for every country. The circumstances of individual countries differed in advance of the crisis. Their actions as the crisis unfolded differed as well. It is attractive to think about phased, coordinated exit plans, but I do not think that will be the most likely outcome, nor should it be the guiding principle. It is desirable for national plans to be phased, for partner countries to be as informed about those plans as is practicable, for antisocial behavior to be minimized, and for plans to be coordinated in that sense. However, reality will fall short of even that modest ideal.

Second, on the treatment of nonconventional assets purchased by central banks, my view is that this is not an issue of high importance, at least for the United States. The Federal Reserve appropriately took extraordinary actions during the crisis. As a result, the Federal Reserve has suffered criticism from many who should know better. Those critics are not going to be silenced by a quick restoration of the Federal Reserve’s balance sheet to the status quo ante.

The balance sheet of every central bank is ultimately a part of the balance sheet of its government as a whole, even if some central banks would like to pretend otherwise. The fact is that the United States more effectively shares a consolidated balance sheet between its central bank and treasury than do most other countries. Federal Reserve profits and losses flow through to the Treasury on a weekly basis. (In the case of international assets, this treatment extends to paper gains and losses, as holdings are marked to market.) It follows that it is of limited significance whether the Federal Reserve or the U.S. Treasury holds the unconventional domestic assets acquired in the crisis. If the Federal Reserve takes losses on its holdings, or on sales of its holdings, there will be an immediate loss of revenue to the Treasury from the Federal Reserve just as if the Treasury had held the assets. Perhaps the Treasury would be more likely to hold certain assets to their maturity, and perhaps the resulting losses to the taxpayers would be lower, but this is not obvious, as the assets would have to be financed in the meantime.

Third, in thinking about unwinding monetary stimulus, a focus on so-called excess reserves is not the right place to start. Today the U.S. banking system holds willingly—indeed demands—more reserves than is normal. This fact is inconsistent with the view that there is a huge monetary overhang that will soon lead to a renewed credit boom. It is true that if the Federal Reserve raises its policy rate (the federal funds rate) significantly relative to the rate it pays on what are technically excess reserves, then the central bank may face overly rapid growth of the money supply. But that is not the most important issue right now.

Let me now answer the questions Olivier Blanchard posed to me before this conference:

Will the process of withdrawing public support in the financial sector be influenced by the ongoing regulatory and supervisory reform in major countries? There will be unpredictable and unintended consequences of phased repair, reform, and recovery of the financial sector. But the aim of the financial sector reform is not to go back to business as usual. There will be some hiccups as we approach a new normality. The system should be able to absorb a few bumps in the road.

Which country circumstances should most importantly affect the unwinding approach? The truth is that each country is going to unwind its extraordinary support activities for the economic and financial system in its own way. I am concerned about the fixation I hear at this conference on putting fiscal recovery first. While that may be desirable in principle, it may not happen. What if the fiscal authority does not get its act together? Does that mean the monetary authority should just sit on its hands? I think not. But even if the fiscal authority initiates unwinding in a timely manner, it might take 18 months. Does that mean monetary policy should also be unchanged for the same period? Not necessarily. In the more likely event that fiscal policy in one or more countries is less than ideal, that also does not mean the monetary authorities should stand by until the fiscal authorities finally act. The risk is that political pressure on the monetary authorities will increase under these circumstances. It should be resisted.

Finally, if we think about monetary policy not in terms of its impact on the individual country but in terms of its impact on the global financial environment, we should be especially cautious. National monetary policy authorities mistakenly kept interest rates too low for too long in the past decade in the European Union, Japan, Switzerland, the United States, and many emerging market economies. This was a major contributing factor to the crisis via mechanisms such as the carry trade. History does not repeat itself precisely, but we should learn its lessons.

What should countries most affected by financial crises be most watchful for? There is an understandable concern that countries should avoid a premature exit from their support activities before financial repair is well underway. At this point, however, what I worry more about, at least in terms of monetary and financial policies, is a transition that is much too late rather than much too early. Countries should expect aftershocks from the crisis, much like the events we saw in Dubai in November. The seeds were sown years ago. There will be many more such aftershocks. That likelihood should not limit timely exiting.

Which countries are most at risk from suffering distorted capital flows? The countries that are most exposed will be most at risk from distorted capital flows; in other words, it will be those countries that have the greatest imbalances, broadly defined to include much more than current account and external debt positions. A country has a problem if it entered the crisis period with a high inflation rate and comes out of the crisis period with a high inflation rate—it therefore now has a relatively high nominal interest rate. If its exchange rate is pegged, its problem will be exacerbated because it is at additional risk of large capital inflows. Is that the fault of the central countries and their easy monetary policies or is it the fault of the imbalances in the peripheral countries? Bygones should be bygones and not become excuses for not addressing imbalances.

What are the costs and externalities if countries unwind in an uncoordinated manner? The answer to this question depends on what is meant by a coordinated manner. In my view, we are going to have differentiation in the timing and content of postcrisis policy adjustments. That is inevitable because the original interventions themselves differed along with countries’ ex ante circumstances. So we are going to have some inherent differentiation as countries exit. This will produce some adverse external consequences—negative externalities. The best we can hope for is shared objectives, open information flows, a minimum of free riding and deliberately antisocial policy actions (such as competitive nonappreciation of currencies), and international support for those countries caught up in the backwash of events. As we emerge from this crisis, it would be dangerous and inappropriate to try to run a convoy system in which the weak hold back the strong to the detriment of obtaining the goal of strong, sustained, and balanced worldwide growth.