Amsterdam November 18-19, 2005

De Nederlandsche Bank

Summary of the Discussions

by
Augusto Lopez-Claros

Emerging Market Crises

There was an interesting and lively discussion on emerging market crises, with particular reference to the risks of contagion, the role of public/private financial flows, the adequacy of IMF resources, and the extent to which various “architectural” innovations (e.g., CACs, codes of conduct) may have played an important role in mitigating the risks of future financial meltdowns in vulnerable economies.

Whatever the many causes of emerging market crises, they tend to be defined by a sharp slowdown or, more likely, a reversal of capital flows.  Often these occur in clusters and we have come to see these as part of a process of contagion, sometimes reflecting trade and financial linkages and, often, herding behavior. One fact to bear in mind is that capital flows to emerging markets in the 1990s were much higher than in previous decades. Even if we compare them with the early 1980s, which saw the recycling of petrodollars, the flows in the 1990s were much larger. Periods of contagion tend to be asymmetric in their manifestations; that is, there are more examples of “depression” than euphoria. They also tend to be clustered around particular regions (e.g., EMS in 1992-93; Mexico in 1994-95; Asia in 1997-98; and Russia/Brazil in 1998-99). When they occur, they tend to hit weakest countries first and the characteristics that make them most vulnerable are: overvalued exchange rates, large current account deficits, high shares of short term debt, low ratios of liquid assets to liquid liabilities. Periods of contagion also tend to be short-lived. Without question, high volatility in capital flows in a central feature of periods of contagion.

Some of the surge in inflows in the past decade clearly reflects some good things. Improvements in macroeconomic management in emerging markets—highlighted by Kristin Forbes, Jose Barrionuevo and others—reflected in better performance, opening of capital accounts, trade liberalization, a more open attitude toward privatization, wider choice of assets for investors in domestic markets, improvements in credit ratings. But also the “search for higher yield” among private sector players and expectations of bailouts, a point underscored by Henk Brouwer. Contagion reflects a number of factors: changes in key economic parameters, such as world interest rates or commodity prices or key exchange rates, all of which are beyond the control of emerging countries. Or trade spillovers (devaluation in Brazil and impact on Argentinian exports, say), or financial linkages, such as exposure of Uruguayan banks to Argentinian depositors or, earlier on, Japanese banks to Thai corporations. Changes in investor sentiment also play a role. So, some volatility would appear to be inevitable and an intrinsic component of increased globalization, even if at times volatility had been excessive.

Less risk of contagion

Nevertheless, there was consensus in Amsterdam that today the risks of contagion had lessened. Ms Forbes thought that countries had taken steps to improve their economies. Reserve levels had risen substantially, making countries less vulnerable to speculative attacks. The shift to flexible exchange rate regimes (e.g., Russia, Turkey, Argentina, Brazil, many others) had certainly helped, with central banks and monetary authorities recognizing the vulnerabilities associated with fixed pegged regimes.[1]  Erderm Basci echoed this view, noting that bond spreads in Russia, Turkey and Argentina had narrowed following the move to more flexible exchange rate arrangements. Barrionuevo thought that investors had learned to differentiate among emerging markets and that, contrary to the 1990s, domestic pension funds had become an important source of stable funding, particularly in Latin America. This had led to a shift in the composition of public debt, with less external and more domestic liabilities. Reflecting these developments and an overall improvement in the policy environment, a full 30% of the asset class in emerging markets was now investment grade, compared to 3% in 1996.

Barrionuevo thought that the main risks to the economic outlook were now global in nature, including the prospects for substantially higher US interest rates, US dollar/euro volatility, and possible disruptions to oil supplies and the associated higher prices. The first two if these would have, in his view, large implications for emerging markets. Ariel Buria agreed with this assessment, noting that the risk of further dollar weakness could precipitate a move to a much higher interest rate environment, a weaker US economy, with adverse effects on those economies highly dependent on exports to the US—Latin America, China, and others.

Crises prevention

A number of interesting points were made by various speakers on the subject of crises prevention. Buria thought that asymmetries in power within the IMF governance structure limited the effectiveness of the institution in its surveillance function, which seemed largely confined to borrowing countries. Nouriel Roubini thought that “the use of the IMF as a slush fund to support G1 strategic allies seriously undermined the credibility of the organization.” He noted, in this respect, massive IMF lending to Turkey, essentially to finance a rollover of the debt, although the Articles of Agreement indicate that lending is to take place in support of balance of payments imbalances. Indeed, in many of the recent packages funding had been for fiscal purposes, in some cases involving commitments for countries to have primary surpluses for the next 10 years, suggesting IMF involvement in these countries for at least that long. Roubini argued that, in the absence of US “strategic” support to Turkey a more efficient and perhaps ultimately less costly way to deal with the crisis might have been to restructure the domestic debt—same face value but longer maturities.

Buria pointed out to Greece as an interesting case study in crises prevention. For Greece, EU membership meant that despite weak fundaments (debt to GDP ratio in excess of 100 percent, large budget deficits, high inflation) there were no sustained speculative attacks against the drachma. Speculative attacks against the French franc in the early 1980s—the “Socialist” start of former president Mitterrand—were, likewise, largely ineffective. In both cases the presence of strong institutional mechanism of support provided these countries with an effective shield against market speculation. Buria contrasted this situation with IMF advice which, in essence, tells countries “to resort to self-insurance, which is costly and inefficient.”

An interesting implication of Buria’s observation’s about the utility of formal mechanisms of international cooperation is that, in the absence of such mechanisms emerging markets will have to continue to rely on an improved policy environment. Chile is able to tap the international markets at unusually low spreads partly because its public debt levels have come down from over 100 percent of GDP in 1986 to 12 percent of GDP in 2004. For the foreseeable future crisis prevention is likely to be very much a responsibility of emerging countries themselves and their private creditors. But, central banks and the IMF also have responsibilities to prevent or contain financial crisis and the spread of contagion. The rationale for a lender of last resort is to have an institution that can lend to solvent but illiquid borrowers when no one else is prepared to do so and in volumes sufficiently large to end financial panic. The IMF fulfills this role only partially.[2]

Lessons from Indonesia

Miranda Goeltom gave a thorough overview of key elements of the Indonesia crisis, providing a useful practical complement to some of the other interventions. A number of factors were instrumental in precipitating the crisis, including a weak prudential environment for the financial sector, un-hedged short-term borrowing by the corporate sector and, more generally, a short-term debt overhang. While there was no apparent budget deficit problem and the current account deficit appeared small and manageable, there was a problem of widespread corruption which, at the outset of the crisis, introduced a political dimension to management of the crisis. The crisis brought about a large drop in private consumption and a sharp increase in the incidence of poverty.

As seen from the perspective of the authorities, the IMF was to perform a dual role during the crisis: provide backing and signaling to the markets and to be used by policymakers as the justification for the passing of tough reforms. In her view, however, an overly ambitious structural program ultimately undermined its own credibility, resulting in a loss of confidence rather than a gain in it. The sharp depreciation of the currency—much larger than would have been the case in the event that confidence had been quickly restored—caused widespread defaults in the corporate sector. The closure of 15 banks precipitated, in the absence of a system of deposit insurance, panic in the domestic financial markets. Ms Goeltom thought that more important than the size of the IMF package itself was whether the program was focused on the handful of really important problem areas where success was essential, if confidence was to be restored. She thus argued for programs that were pragmatic in design, deliverable in terms of implementation and opportune in timing. She would favor a larger role for mechanisms of regional cooperation, such as those existing in the EU, as an effective mechanism to deal with the confidence issue. Richard Portes thought that the United States would remain opposed to regional monetary arrangements because they would diminish its influence in the IMF.

Crises Management

Roubini provided a compelling synthesis of some of the key issues affecting emerging market crises and the international community’s response to them. He asked whether the crises of the last 10 years were an aberration, noting that the last one had taken place in Brazil in 2002. The improvements recently seen in the macroeconomic environment—globally and in the emerging markets themselves—could change, however. Already average levels of public debt in emerging markets were around 70 percent of GDP, worryingly high and virtually assuring future crises. Indeed, changes in the level and structure of emerging market debt (maturity, currency of denomination) would keep many of these countries vulnerable, although it was also certain that some countries would gradually “graduate to credibility.” In light of this, a legitimate question to ask was whether the IMF needed more funds. Roubini thought no, but that it was necessary to use existing funds more efficiently. This was desirable, not just to reduce the number of large debtors to the Fund absorbing an inordinate share of its resources, but also as a precautionary measure, to have an adequate cushion of resources available in the event that crises in places like China and India might require an international response. A similar response in these countries to the assistance provided to Mexico in the mid-1990s might call for packages of some US$150-250 billion, clearly well beyond the scope of resources available to the Fund at this time.

He noted that there were already in place restructuring mechanisms for sovereigns facing unsustainable debt burdens. In many cases debt restructuring would be the most realistic way out. Indeed, we should not shy away from a laisser faire approach to crisis resolution; in Argentina there may be no practical alternative to the authorities sitting down with their creditors and hammering a deal out. Too much emphasis, in any event, had been placed on sovereign debt restructuring when, often, the problem was to be found elsewhere—for instance in a weak and undercapitalized banking system. The problem, in his view, was that countries like Brazil, Turkey, and Argentina, already highly indebted, had received IMF lending mainly for expenditure support, potentially aggravating the problem down the road. Roubini regretted the ad hoc approach to crisis management seen over the past decade. A 25 percent GDP drop in Argentina was not necessary—it reflected bad use of IMF resources, a confusion of the role of the IMF in liquidity crises (when lending was warranted) with that involving cases of insolvency of the state, when debt restructuring was the desired course of action.

Access limits, signaling, and the role of the Fund

These inefficiencies notwithstanding, Roubini thought that an active IMF/G7 support in future crises was inevitable and, in this respect, argued that the 300 percent of quota IMF access limits were too low—the constant need to break them in recent financial market meltdowns had undermined the organization’s credibility. Mahmood Pradhan echoed this view, suggesting that it was not advisable for the IMF to be constrained in times of crises. In his view, however, most large programs in recent years (Turkey, Argentina, others) had not involved countries posing systemic risks to the international financial system. It was thus necessary for the Fund to better justify exceptional access limits—the lack of clarity in this area was self-evident and added a degree of adhockery to management of emerging market crises. Doris Grimm thought that greater Fund clarity on access would be helpful to the market in the event of emerging market bailouts.

Mark Allen agreed that the Fund should not lend to unsustainable debt cases. However, there were cases when “we simply did not know.” Crises often did not have a balance of payments origin, in the traditional sense of the term. Should the Fund have left Uruguay and Turkey “to their own devices”, particularly at a time when G7 statements had signaled that the IMF would remain the only channel of official emerging market lending? He conjectured that the answer to this question was not that simple. For his part Alexander Swoboda pointed out that in the midst of all this discussion about highly indebted countries, massive bailouts and moral hazard, one should remember that there was still a place for “small” country programs to tide countries over in the traditional balance of payments support of the past, as per the Fund’s original mandate.

Jack Boorman noted that the IMF’s access limits were a “political” heritage from the past—they have little to do with the realities of the present world. Would it not be dangerous to tie our hands by a rigid interpretation of these limits? Was the linking of these to quotas still relevant as a concept and, if so, what was the right size? More generally, what was the role of the Fund in times of crises? Beyond providing finance, was it mainly that of technical advisor to countries, or a signaler vis-à-vis the markets? In his presentation Kees van Dijkhuizen made a distinction between IMF surveillance, which he regarded as a form of “light” signaling (multidimensional, no Fund resources on the line, low frequency, Article IV reports not endorsed by the Board) and IMF programs, which could be seen as a form of “strong” signaling (clear and unambiguous, backed by Fund resources, albeit with a number of drawbacks, such as their tendency to add to the debt burden, to be perceived by countries as intrusive). He argued that an intermediate approach (or solution to this quandary) might be increased used of precautionary arrangements, involving upper tranche conditionality, Board endorsement, Fund resources (potentially but not necessarily), and a largely favorable historical track record.

Allen thought that there was a much better understanding of the “mechanisms of crises” at the IMF than in 1997. Yes, the IMF had failed miserably in Russia and in Argentina. There was increasing concern at the IMF Board about overexposure to certain countries. There had been, undoubtedly, too much funding of crises and not enough effort expended in preventing them. The CCL had been an early attempt to adapt the Fund to the role of crisis manager, but countries had been unwilling “to put their fates in the hands of the Fund” and the possibility of international stigma associated with access to the facility. The high qualification standards, combined with the large voting power necessary to keep the facility operational had led to its early death. The Fund had recently ventured into so-called non-borrowing programs, wryly characterized by some as providing “all the benefits of Fund conditionality without the burden of funding.” There was also some on-going thinking about the structure of precautionary arrangements, in particular the nature of their underlying conditionality. On Boorman’s point about the Fund’s role as signaler, Allen thought that there was a conflict of interest between the signaling and the lending role of the Fund. Signaling should be a by-product of other activities of the Fund, not an end in itself; the Fund remained a cooperative insurance mechanism. Allen thought the Fund was at its most effective when it acted through persuasion rather than the threat of punishment.

Brouwer made a number of interesting points concerning aspects of crises resolution, reforms undertaken, and contingency planning. He thought that the move, in the past decade, to excessive reliance on large packages of official support had contributed to heighten the risks of moral hazard, something that called for a return to more cautious lending policies. This was particularly the case in light of the mixed results associated to large scale official rescue packages, with many of the recipient countries nowhere near regaining market access and thus likely to remain IMF clients for the foreseeable future. Indeed, these developments had raised questions about the financial position of the Fund and substantially undermined its leverage vis-à-vis its borrowers, which, unlike the 1970s, were no longer large, solvent, industrial countries. Brouwer saw virtue in more thorough debt sustainability analysis to settle the issue of whether the Fund faced a liquidity crisis or insolvency on the part of the state, a critical judgment in helping determine the role that the Fund was to play in a particular crisis. In this regard, he thought that there was much value in the design of alternative strategies for a given country, in the event that the evolving situation suggested that what appeared, ex-ante, as a liquidity crisis turned out, ex-post, to have been a solvency crisis. This would led to better crisis management overall, but would also tend to reduce Fund exposure.

The IMF’s financial position

John Chambers spoke about the financial strength of the IMF, saying, in essence, that it is strong. In addition to its credit portfolio, Fund assets include 10 percent of the world’s official gold holdings, which it cannot sell. It also has access to a potential pool of some SDR 34 billion under the GAB (General Agreement to Borrow) and the NGAB. At present, there are four member countries whose debts exceed the Fund’s “precautionary balances.” There is a credible plan in place to raise these to SDR 10 billion by 2010, a figure intended “to insulate its income statements from adverse credit performance.” A larger figure would “enhance the IMF’s ability to provide exceptional financing to its larger members in the developing world.” Chambers thought that the IMF should not lend to a single country more than its precautionary balances. Nevertheless, the Fund’s one-year forward commitment capacity was now some SDR 58 billion, compared to SDR 20 billion after the Asian crises. Adding flexibility to management of the Fund’s financial position was the fact that it could change its interest rates at will (“burden sharing”) and could, thus, for instance, sustain an Argentina default.

Issues in IMF governance

James Haley focused attention on the consensual nature of the IMF—a cooperative institution between sovereign states in which each felt the need to gain something through active participation. It was increasingly evident, however, that the governance structures that were appropriate in 1944, at the outset of the institution’s creation, were no longer valid today.[3] Its quota structure was no longer consistent with its lender of last resort role, which had acquired prominence in recent years, at a time when the institution had moved to having permanent debtors in its balance sheet. He noted that in a world of wild capital movements, there would be issues of credibility, access, predictability and that the IMF would remain the focus of much public attention in the context of future crises management. However, the recent accumulation of reserves in Asian countries may very well reflect fears that the “Insurance Co IMF had changed the terms of the contract” and that, henceforward, countries would do better to fend for themselves. Ms Goeltom’s earlier remarks reflected this lingering sense of frustration with the role played by the Fund in its role as crisis manager.

There was a lively discussion—started by Boorman—on the issue of whether the funding of the IMF’s administrative expenditures by the spread it charges on its lending operations was a sensible approach. As currently structured, the salaries of the Fund’s managing director and of its entire staff (as well as other administrative expenditures) are financed by the interest paid by tax-payers in Argentina, Turkey, Russia and other users of Fund resources. Whereas IMF lending operations have no budgetary implications for members such as the US and the EU; (indeed they earn a return on their SDR reserve assets); a country such as Russia, by contrast, has paid, since August 1998, well over $3 billion in interest charges on previous Fund loans. In the view of many, particularly in the developing countries, such a circumstance alone might go some way to counter the existing notion that, because the large shareholders “contribute” more to the organization, they are in some manner entitled to oversee its operation as well (including, in the case of the EU, the appointment of the Fund’s managing director), particularly since they have already the largest voting shares at the IMF Board.

In his discussion of the international and domestic politics of IMF programs James Vreeland made a number of interesting points. First, in the issue of whether there is evidence of systematic use of the IMF by the US for foreign policy purposes, the evidence suggested that countries that voted in the United Nations along the US had higher likelihood of US support in IMF programs. Second, the more a country received US aid, the more lenient the IMF was likely to be on program implementation. Both of these, in Vreeland’s view, undermined the credibility of conditionality since there was evidence that conditionality actually had some effect in countries not usually allied with the US and where there was partial ownership of programs. He also suggested that there was no credible evidence that entering into IMF programs played a catalytic role in terms of private sector financial involvement—in fact the reverse was more likely true.

���