Capital Controls and Crisis Management


Capital Account Convertibility and Capital Controls in Emerging Market Countries: Some Themes from the First Meeting

John McHale
Harvard University
Email: jmchale@harvard.edu


Section 1. Introduction

On November 6, 1998, a group of leading researchers and practitioners in the area of international finance got together in Cambridge Massachusetts to discuss capital controls in emerging markets. The backdrop to the meeting is, of course, the international financial crisis that has spread from Asia to Russia and continues to threaten Latin America. The motivating question is the role that controls might play in managing the crisis, and in preventing similar crises in the future.

Although the "tequila crisis" that followed the Mexico's devaluation in 1994 took place just a few years ago, it is striking how the crisis that erupted following Thailand's devaluation in 1997 took the world be surprise. Asia, after all, was supposed to be different. At first it was not apparent how serious the crisis would become. A few months after the Thai devaluation, at the IMF/World Bank meetings in Hong Kong, it was even felt that the time had come to add a new chapter to the Articles of Agreement of the IMF on capital account convertibility. The new powers that would give it jurisdiction over capital account in addition to current account transactions, making capital account liberalization a central part of its mandate. A little more than a year and a new round of meetings later the mood is very different, and the merits of a liberalized capital account, along with a long list of competing proposals for a new international financial architecture, are being actively debated once again.

The November meeting was the first of a series of informal meetings and more formal conferences that are being organized by Martin Feldstein on this and other crisis-related issues. The format for the first meeting was a series of short informal presentations by researchers who have written extensively on the topic, (1) interspersed with a free ranging and lively discussion with a number of invited participants. (2) The purpose of this short paper is to give the flavor of the debate by outlining the main themes of the discussion, the threads of which weaved through the day.

The discussion of the merits of capital controls did not keep within any boundaries. Nevertheless, it is useful to separate their role in crisis management and their role in crisis prevention. Thus we divide our account of the proceedings into two sections (though putting the discussion on a topics like currency boards or the role of foreign banks in one or other category is somewhat arbitrary). The discussion was also encouraged to meander, so we found it best to review it thematically rather than chronologically. Section 2 brings together the themes that relate to the crisis management, which are mainly about controls on capital outflows. Controls on capital inflows as well as competing policies for crisis prevention are then taken up in Section 3.

Section 2. Capital Controls and Crisis Management

2.1 Krugman's Eternal Triangle

Do capital controls on capital outflows have a role to play in the management of the kind of financial crises that followed Mexico's devaluation in 1994 or Thailand's devaluation less than three years later? The discussion on this question revolved around Paul Krugman's organizing device of the "eternal triangle" (which he also referred to as the "irreconcilable trilogy"). This device brings out in a stark way the constraints that have faced policy makers in the midst of crisis, so it is useful place to start.

The story of a crisis goes like this: Investor confidence in a country that had been pegging its exchange rate to the dollar suddenly evaporates. With net capital inflows falling and reserves declining, the country must raise interest rates if it wants to defend the exchange rate. With a restrictive monetary policy of this kind, the standard textbook prescription for avoiding a demand contraction is a fiscal expansion. But with investor confidence tied to the size and direction of change of the government's budget deficit, the country might be forced to pursue fiscal austerity instead, worsening the emerging output gap.

Why then not give up this recession-inducing defense of the currency? This is essentially what the United Kingdom did in 1992 when it left the Exchange Rate Mechanism (ERM) of the European Monetary System (EMS), lowered its short-term and long-term interest rates, and subsequently led the rest of the continent in economic recovery. Unfortunately, different rules seem to apply to emerging-market economies. For one thing, discreet devaluations are difficult to manage. Giving up on an exchange rate peg that the country had been promising to defend leads to a further loss of confidence. The intensification in selling of the country's financial paper and the reduced willingness to roll over short-term debt can lead what was planned as a controlled devaluation to turn into a route. Past experience also shows that exchange-rate depreciations--which can boost to price competitiveness in established market economies--can be very damaging in their emerging counterparts. (3)

There is thus, argues Krugman, a difficult dilemma. The macroeconomic policies needed to defend the exchange rate will cause deep recessions. Yet allowing the currency to fall might be even worse. For this reason, he is willing to contemplate the use of capital controls. He agrees that such controls have many problems, saying "controls are hard to implement and enforce, they disrupt normal trading relations, and they may impair confidence for a long time to come."

The idea of the triangle is that there are three desirable capacities: the capacity to use stabilization instruments; the capacity to defend the exchange rate; and the capacity to have free capital movement. The dilemma is that, in a financial crisis, only two of the three are achievable. If you demand free capital movement and a defense of the currency, then stabilization must be sacrificed. If you demand free capital movement and freedom in the use of monetary and fiscal policy to attempt stabilization, then you will not be able to defend the exchange rate. Since both macroeconomic policy austerity and a collapsing exchange rate are both likely to lead to deep recessions, capital controls--however undesirable in themselves--might be the only way out. Krugman's point is that given the alternatives they should be a least considered.

2.2 Multiple Equilibria or Flawed Fundamentals

The Krugman depiction of the policy constraints did not go unchallenged. One challenge that was raised in various ways by numerous participants was the Krugman presentation of the nature of the crisis as one involving self-fulfilling expectations (or multiple equilibria) as distinct from flawed fundamentals. The difference is important for how one views a recession that follows the crisis. If the crisis is caused by investors running for the exits because they believe other investors are running too, then any resulting recession probably reflects a fall in output below the economy's potential. The recession then reflects a waste of potential and reflation policies are warranted to reduce the output gap. On the other hand, if the crisis is caused by the fundamentals of the economy turning out to be worse than previously thought, then the recession reflects a move to a more realistic assessment of potential output.

The strongest challenge to the multiple equilibria view came in Michael Dooley's presentation. Dooley questioned the claim that the fickleness of financial markets is the source of the problem, rather than "governments' participation in financial markets." This participation comes in the form of implicit government guarantees, which leads money borrowed abroad to be misused, say by hiding the proceeds offshore (as in Indonesia) or through investments in unproductive capital (as in Korea). When the crisis hits the previous misuse of resources will be a serious impediment to recovery, since the "very large real loss in the financial system must be eventually allocated to tax payers." In Dooley's view, however, controls on capital outflows will not be of any help. He concludes: "Residents and non-residents 'stuck' in a financial market populated by insolvent banks and firms are not going to invest in productive capital." Nouriel Roubini observed that interest rates in Thailand Korea are now lower than they were before the crisis, yet there is a credit crunch. The reason is that the banking systems are "rotten" and the corporate sectors are "bankrupt," so that credit will not begin flowing again until there is a major financial restructuring. In these circumstances, he argued, lower interest rates than come with controls on capital ouflows are of little help.

Anne Krueger shared Dooley's view that the Asian crisis is more about bad fundamentals than self-fulfilling expectation. She contrasted two ways to have fast growth: have a high real return on investment; or have lots of foreign capital flowing in together with high domestic saving even with a low or negative real return. High growth in Korea or Malaysia, she argued, reflected the second to a large extent. In other words, the true potential output of these was less than might appear from the explosion of new construction that appeared to be rapidly lifting the sustainable productive capacity of these economies.

In response to the pessimistic assessment of the potential output of these economies, Andres Velasco cautioned against reading too much into the pervasive un-profitability in the midst of a serious recession. He pointed out that in the midst of the Chilean recession of 1982 a large number of projects seemed to be bad investments--and indeed the results of crony capitalism. Many of these turned out to have remarkable rates of return over the next decade and a half. The same might yet be true, he surmised, of many projects that are condemned today as the handiwork of crony capitalists. Paul Krugman agreed with Velasco's point, adding that a lot of the U.S. banking system was in trouble in 1991, in the midst of a relatively mild recession. He asked what kind of banking system could survive the kind of downturn now being experienced in Asia. Krugman also noted, drawing on an analogy with finding the cause of accidents on a stretch of dangerous highway, that if you look hard enough after the crisis (or crash) you can always find something that the country (or driver) has done wrong.

In a related vein, Velasco asked at another point just what was it that changed in Asia between 1994 and 1996. The conventional wisdom had been, he pointed out, that the Asian economies were models of good management. Anne Krueger responded that the problems had been there all along. Offering a different explanation, Carmen Reinhart suggested that it was macroeconomic policy that had changed as these countries attempted to sterilize inflows as a counter-cyclical measure, which kept relative rates of return high, and attracted "hot money." If this was their intent, Michael Mussa joked, it turned out to be counter-cyclical beyond their wildest expectations.

2.3 The Wisdom of Currency Boards

The option of using a currency board arrangement (4) (a la Hong Kong or Argentina) as a way relaxing the constraints was raised repeatedly through the day. Carmen Reinhart went a step further and argued for the benefits of full dollarization of the economy (a la Panama), although objections to this more radical step were raised on the grounds of seigniorage (Richard Cooper) and sovereignty (Paul Krugman and Peter Garber). These arrangements do not avoid the irreconcilable trilogy, of course, since monetary policy--and even the capacity for a central bank to be lender of last resort--is given up. Noting a tendency of monetary policy to be destabilizing rather stabilizing, Greg Mankiw wondered if countries would actually be better off surrendering monetary control.

The possible attraction of currency boards is that the exchange rate commitment is potentially more credible than a commitment to a fixed but adjustable exchange rate peg. As a consequence, interest rates might not have to rise so high to keep capital from flowing out (Felipe Larrain observed that Argentina now has interest rates than Mexico). It was pointed out by Kathryn Dominguez, however, that although Argentina has fared relatively well in the recent crisis, it had to endure a sharp contraction after the Mexican crisis. Credibility is not guaranteed by just adoping the arrangement. Anne Krueger was concerned that we give due weight to the different political economy situations in different countries. The will of Argentineans not to return the hyperinflations that racked the country before the currency board gives credibility to their vow to defend the arrangement. Lacking the same will to defend, she conjectured that a commitment to a currency board would be less credible in Brazil. Jeffery Frankel agreed with this point, emphasizing that the wisdom of a currency board depends on the circumstances--circumstances that were not right in Indonesia, for example. Felipe Larrain made the point that even if you believe that flexible exchange rate regimes are better as a general rule, that does not mean that you should recommend moving toward flexibility to a country such as Argentina that has already committed to a highly fixed regime.

2.4 Evidence on the Effectiveness of Outflow Controls

In their presentations, Felipe Larrain and Carmen Reinhart reported on empirical and theoretical evidence on the effectiveness of capital inflow and outflow controls. We will refer to the evidence on inflows in the next section, but the evidence on the effectiveness of outflows bears directly on effectiveness of such controls as a crisis management tool.

Filipe Larrain's discussion focused on Chile, a country that actually liberalized outflows during the 1990s while simultaneously increasing controls on inflows. Somewhat surprisingly, the liberalization of outflows led to larger net inflows--a finding that is consistent with a number of other episodes. A possible theoretical explanation explored by Larrain is that liberalizing outflows reduces the irreversibility of investments, and thereby reduces the option value of waiting to invest. For the debate about restricting outflows in the midst of a crisis, Larrain puts his finding in reverse and concludes that such controls might actually cause net capital inflows to fall.

Carmen Reinhart also reported on some preliminary research related to impositions of controls in such countries as Venezuela, Argentina and Mexico. She emphasized that these were harsh controls involving such measures as suspension of convertibility and forced conversions of foreign exchange. Also, she points out that the measures were supposed to be temporary but tended to persist. (Greg Mankiw had raised this danger earlier in the day, offering the example of New York rent controls. These were supposed to be just a war-time measure, but remain to this day.) Her evidence indicates that the controls were effective in limiting reserve losses, but there was also evidence of considerable evasion. Overall, however, taking into account the "signalling element," she shares Felipe Larrain's belief that controls on outflows are probably not effective in increasing the net inflow position.

Although the crisis management discussion was mainly about controls on capital outflows, Carmen Reinhart raised the importance of counter-cyclical controls on inflows, where controls are tightened during an inflow cycle and loosened during an outflow cycle (Chile and Brazil have used this flexible approach). On this point, Felipe Larrain reported that Chile has reduced its non-renumerated deposit requirement to zero to attract inflows during the current difficult period, but the instrument remains in place for future use. Reinhart's emphasized that, given Paul Krugman's description of how little scope there was for engaging in counter-cyclical policy by other means, counter-cyclical controls on inflows could be a valuable instrument.

Section 3. Capital Controls and Crisis Prevention

As a tool of crisis management, capital controls were viewed with skepticism by most of the participants--despite Paul Krugman's spirited arguments. When viewed as a tool of crisis prevention, however, there was more willingness to consider efficacy of at least certain types of controls. The discussion ranged widely, but three themes serve as useful organizing devices: Are capital controls needed to prevent crises? Is there a better way? And if capital controls are used, what type of controls might be appropriate?

3.1 Are Capital Controls Needed?

As noted in the previous section, how a participant viewed the crises--multiple equilibria or flawed fundamentals--conditioned to some extent their view of capital controls. In the context of crisis prevention, the multiple equilibria view was most strongly advanced by Jagdish Bhagwati. Professor Bhagwati stressed the differences between the case for free capital movements and the case for free trade--a case for which he has been an articulate defender. He expressed agreement with Charles Kindleberger's view of financial markets as being prone to cycles of manias, panics and crashes. On the benefit side, he finds little evidence that the gains from non-FDI capital inflows are large in non-crisis situations. Moreover, he argued that there is a reasonably high probability a crisis will occur when there is free capital movement, and that the costs of the crisis tend to be amplified by bad IMF imposed policies when they do occur. His conclusion is that countries need to move cautiously in removing controls. (Although, interestingly, he expressed the belief that countries that have already liberalized should "stay the course" during a crisis.)

A number of participants who view the crises as being determined by flawed fundamentals, or more generally, the poor use to which capital inflows were put, also expressed reserved support for certain controls on capital inflows. The moral hazard problem loomed large in this discussion. There is a common outline to story told: Domestic financial or corporate institutions with explicit or implicit government guarantees received foreign loans, but then put the money to poor use financing projects with low or negative real rates of return under "crony capitalism." The foreign investors also pay inadequate attention to their loans believing that the domestic governments will not let the institutions fail. Pegged exchange rates further lull the investors into believing that they do not face exchange rate risk.

The empirical evidence on the impact of controls on inflows reported by Felipe Larrain and Carmen Reinhart presented a consistent picture: controls have little effect on the overall volume of capital flows, but they do alter the composition in favor of longer term flows. Reinhart also found that efforts to sterilize the effects of inflows on the domestic money supply and interest rates (5) did increase the volume and composition of flows, in this case towards the short end. Such sterilization efforts were common in Asia leading up to the crisis in an attempt to stop their economies overheating, keeping rates of return relatively high, encouraging significant flows of "hot money" and making the region vulnerable to the kind of sudden reversal that ultimately took place. Michael Dooley made a similar point in his presentation, concluding that "controls can generate interest rate differentials but not prevent crises." The reason is, Dooley argues, the prospect of discreet devaluations generate profit expectations that swamp these interest rate differentials.

3.3 Is There a Better Way?

Although this misuse of funds was seen as providing a rationale for controls, most participants who accepted this believed that there was a better way of dealing with the problem. Nouriel Robini, for example, argued for the importance of prudential regulation of domestic financial institutions given the existence of deposit insurance and a lender of last resort. He agreed, however, that capital account liberalization should proceed slowly until a reasonably strong banking system was in place, but he was optimistic that this does not have to take a very long period of time provided the political will is there. Barry Johnson noted the difficulty of distinguishing prudential regulations from capital controls. He asked, for example, whether Chilean reserve requirements on foreign liabilities are capital controls or prudential regulations. Assuming that they can be distinguished, he suggested that the next question should be whether the capacity exists for implementing the prudential regulations or not. It is only if this capacity does not exist that the controls should be reached for. Peter Garber made the interesting observation that much of what investors have done in the crisis follows standard risk management practice. Indeed, they have followed practices that are imposed by industrial country regulators. These methods have worked to prevent more serious problems in the industrial country banking systems, but they have intensified the liquidity problem in both emerging and industrial countries. Efforts to impose these types of regulations (e.g., the requirement that assets are valued on "mark-to-market" basis) on emerging markets economies would also make them even more prone to sharp liquidity crises.

In his presentation, Barry Eichengreen argued that an all or nothing is not the right way to think about capital controls, but instead put the issue in terms of multiple lines of defense. The first line of defense is prudential risk management of risks on the part of banks and other financial institutions. This, in turn, should be backstopped by domestic regulatory supervision that sets limits on exposures, concentrated lending and so on. Only when the first two lines of defense are inadequate should capital controls be considered. Even then, it must be recognized that not all controls are created equal, with some more market friendly (e.g. Chilean controls) than others (e.g. Malaysian controls).

Barry Eichengreen also referred back to the fact that emerging market economies get treated differently to more established market economies when they have to devalue. In an economy without central bank independence and a history of weak discipline in fiscal policy making, a devaluation sends a quite different signal than it would send in an economy with better institutions. The conclusion he draws, however, is not that capital controls are needed but that monetary and fiscal institutions should to be improved. For example, there is a case for an independent fiscal authority. Michael Mussa followed up on this by contrasting the cases of the U.K. and Italy after they both left the ERM. The U.K. was able to lower interest rates aggressively, but Italy did not feel that it had the scope to lower interest rates because of their weak fiscal position. What has happened in emerging markets is an extreme version of the Italian situation.

A more radical solution to the problem of weak institutions and poor track records was proposed by Michael Dooley--allow no foreign debt. In this way countries could process domestic savings without running into costly crises. On the debt management theme, John Campbell noted that there are credibility problems both with domestic currency as well as with foreign currency debt, in part because the value of domestic debt can be eroded through inflation. A solution to this credibility problem, he suggested, might be inflation indexed domestic debt, although he noted that even here the unreliability of price indexes could pose a problem. Andres Velasco agreed with the importance of indexed debt, and reported on Chile's success with these instruments. Anne Krueger once again raised the issue of the importance of political institutions. The Chileans might have the institutional infrastructure to be successful with indexed debt and capital controls, but many other countries do not share the Chilean advantages.

The issue of exchange rate regimes came up again in the context of this discussion. A number of participants felt that pegged exchange rates encouraged large but volatile capital inflows. But rather than trying to control the inflows a better response might be to let the exchange rate float. Aaron Tornell's observation that the largest corporations in Mexico have built up substantial short-term dollar denominated debt, despite Mexico now having a floating rate regime gave rise to some surprise and concern. Micheal Dooley felt, indeed, that the exchange rate regime does not make all that much difference, stressing again the importance of the moral hazard problem created by the implicit government guarantees.

This led to some lively exchanges. Micheal Mussa agreed that moral hazard is an important part of the story but that Michael Dooley was taking its importance too far; bad luck also played a part a large part in the crisis. To bring out the importance of moral hazard Mussa used the example of the Great Depression in the U.S At that time a large number of banks did fail, but not as many as you would have expected given that output was down by a third. The reason, he argued, was that the banks had no expectation of being bailed out and so had been prudent in their lending. To maintain this discipline, he argued, there must be losses along the way: everybody must not be bailed out all the time.

The importance and difficulty of having a good domestic banking system also led to a discussion of the possible role of foreign banks in the economy. This was strongly advocated by Linda Goldberg, who argued that this was a way to get more prudential banking as well as solving the problem of lender of last resort where there are foreign currency liabilities. David Lipton agreed that a greater role for foreign banks would help, be recounted his experience of how big an obstacle the issue of sovereignty was at the height of the Korean crisis. Steve Radelet seconded this, pointing to the very different attitude to the acceptability of FDI in the goods sectors and FDI in the financial sector. Jeffery Shafer pointed to the length of time it took for interstate banking to emerge in the U.S. Even if it was possible to have a bigger foreign banking presence, Peter Garber did not think it would help much in a crisis. What is a crisis after all, he observed, but investors taking their money out of the country. He did not see why foreign banks would be any more willing to extend credit than would domestic banks.

Responding to the various alternatives to capital controls that were advanced, Dani Rodrik agreed that there are better ways of dealing with problem of badly directed money than resorting to capital controls. He stressed, however, that these measures--better prudential regulation, eliminating crony capitalism, relaxing the excessive rigidity of exchanges rates, and so on--all take time to implement. And time is not a luxury you have in a crisis. His conclusion is that you should "work on all the margins that you have," adopting a "portfolio approach" to the problem. He believes that given the practical limits on the other solutions, capital controls cannot be dismissed out of hand.

2.3 What Types of Capital Controls Are Effective?

For most of the day, participants discussed capital controls in broad terms, passing over fine distinctions between controls of different types. An important theme in Jeffery Frankel's presentation, however, was that controls come in very different types and we need to be clear about what type we are talking about. Moreover, to determine which controls might be helpful, we need to consider the aims of controls.

Although most of Frankel's presentation focused on different aims for inflow controls, he began by discussing outflow controls as a way of managing a crisis. These, he argued, have the least support, although he agreed with Paul Krugman that one must ask what the alternatives are. The main objections he raised were: they typically easy to evade; they reduce the pressure of the international financial markets to pursue good policy; and they can led to contagion (e.g., Russia's unilateral rescheduling and Malaysia's controls hurt affected investor confidence in other countries as well).

What are the different aims for inflow controls? And which of these aims are likely to be met? First, there is the aim of discouraging the volume of inflows. These do seem to be effective, if only because of their signalling effect to foreign investors. Second, there is the aim of changing the composition of inflows. Frankel reported evidence that short-term flows do increase the probability of a crisis, so that restricting flows at the short end them could make an economy less crisis prone. He agreed with Felipe Larrain and Carmen Reinhart that controls that discriminate between flows of different types can be effective in changing the composition of flows. Third, there is the aim of reducing exchange rate volatility with a small Tobin tax on all types of capital transactions. Here he agreed with Michael Dooley that small taxes of this type are not effective in avoiding currency crises. Finally, there is aim of retaining or regaining monetary control and the decoupling domestic from foreign interest rates. This can be achieved by limiting both inflows and outflows. Can this be effective without completely turning your back on international capital markets completely? Here he finds that the verdict is mixed. A number of the Asian economies now in crisis (e.g., Thailand and Korea) thought they had achieved this, but obviously this did not turn out to be the case. On the other hand a country like China seems to be weathering the crisis reasonably well. The waters are muddied even more, however, in that Hong Kong and Singapore with quite open capital markets have also weathered the crisis reasonably well. So again there is no clear conclusion about the best capital regime for insulating an economy from financial market disruptions.

Nouriel Roubini agreed that we need to be precise about the nature of controls when we talk about limits on capital inflows. There are a number of hard questions to be answered. For example, are controls placed only on the banks or are they placed on the non-bank sector as well? He gave the example of Indonesia where the corporate sector borrowed heavily in foreign currency. It is hard to say that you only care about the banks because of their unique systemic importance, since the Indonesian banks were themselves heavily exposed to the corporates. If controls are only on the banks, he queried, are they only to be placed on the borrower banks (say higher reserve requirements on foreign currency deposits) or are the lender banks targeted too (say by higher risk weights on cross-border short-term lending)?

With various cases having been made for capital controls of various sorts, Aaron Tornell raised the question of implementability given the "fungibility" of capital. Peter Garber also raised this concern in relation to inflow controls, noting that if you don't control outflows the same position can often be taken using derivatives. Responding to Tornell, Michael Mussa stressed that the ability to implement depends very much on the circumstances--evidently, a recurring theme. As an example, he pointed to the openness to trade of the Malaysian economy as a factor making controls less effective. On the more postive side, he noted that they work better in China given the smaller share of trade in GDP, but even there they "leak like a sieve."


1. The presentations were given by: Jagdish Bhagwati (Columbia University) by videotape, Paul Krugman (MIT and NBER), Michael Dooley (Federal Reserve System and NBER), Jeffery Frankel (Council of Economic Advisors and NBER), Barry Eichengreen (UC, Berkeley and NBER) and Michael Mussa (International Monetary Fund and NBER), Felipe Larrain (Harvard University); Carmen Reinhart (University of Maryland).

2. In addition to the presenters, the other participants were: Andrew Berg (International Monetary Fund), John Campbell (Harvard University and NBER), Richard Cooper (Havard University), Kathryn M.E. Dominguez (University of Michigan and NBER), Peter Garber (Deutsche Morgan Grenfell), Linda Goldberg (Federal Reserve Bank of New York and NBER), Simon Johnson (MIT), Barry Johnson (International Monetary Fund), Anne O. Krueger (Stanford University and NBER), David Lipton (Carnegie Endowment for International Peace), N. Gregory Mankiw (Harvard University and NBER), Don Mathieson (International Monetary Fund), John McHale (Harvard University), Jean Pisani-Ferry (Ministere de l'Economie, des Finances et de l'Industrie, France), Steve Radelet (Harvard University), Dani Rodrik (Havard University and NBER), Nouriel Roubini (Council of Economic Advisors and NBER), Jeffery Shafer (Salomon, Smith Barney), Hal Scott (Harvard University), Aaron Tornell (Harvard University and NBER), Andres Velasco (New York University and NBER), Shang-Jin Wei (Harvard University and NBER). The meeting was chaired by Martin Feldstein (Havard University and NBER).

3. Possible reasons include the high share of imports and exports in GDP, which leads to a fast pass through of deprecations to domestic inflation. Moreover, the high share of imported goods in the consumption basket leads to a fall in the feasible real consumption wage, making adjustment difficult when consumption wages are rigid downward. A relatively new dimension to these crises is the importance of foreign currency liabilities in the balance sheets of domestic banks and corporations. Exchange rate depreciations thus increase liabilities in the domestic currency, making banks less willing to finance working capital and long-term investments, and leading to a debt overhang in the enterprise sector.

4. With a currency board the entire monetary base (and possibly even all of M1) is backed by dollar assets. In other words, the central bank cannot issue currency backed by domestic government bonds or loans to domestic banks.

5. Capital inflows tend to increase the domestic money supply and decrease domestic interest rates. The expansionary effect on the money supply comes about when the recipient of the inflow exchanges the foreign currency for domestic currency. The central bank can offset--i.e. sterilize--this effect by engaging in contractionary monetary policy. For example, the government can sell bonds to the domestic public, taking money out of the economy in the process. One consequence of this is that the government debt held by the public increases, thereby increasing the consolidated government's debt servicing cost.