Perspectives on External Financial Crises

The Money Marketeers of New York University

December 10, 2001



I am honored to receive the Distinguished Achievement Award from the Money Marketeers of New York University. I have interacted, and in many cases worked directly, with almost all of the previous recipients of this award.

In reading a biographical sketch of Marcus Nadler I noted that he not only was one of my many distinguished predecessors who headed the international staff at the Federal Reserve Board in Washington, but also studied at Columbia University. My connections with New York City started at Columbia more than 50 years ago. My father taught there for 19 years; my wife's physical therapy degree and our daughter's medical degree are from Columbia. I am delighted that Tracy and Tina could be here tonight. Without my entire family's support, understanding and continuing affection, I would not have enjoyed such an exciting and stimulating career.

One of the most enjoyable aspects of my career has been the opportunity to work with talented and dedicated colleagues; many became guides as well as close friends. Paul Volcker stands tallest among them, not only in physical stature, but also in his dedication to doing what is right. I am honored that he agreed to introduce me tonight.

I am also deeply moved that this occasion has brought us together in New York City three months after the tragic events of September 11. Those events impacted individuals and families in many nations. I hope that a strengthening of international cooperation, including international financial cooperation, will be a legacy of those events.

I have spent a good deal of my career pursuing international cooperation in the context of external financial crises. Such crises generally occur when a country cannot meet its external obligations. Frequently, but not always, the international community comes to its rescue via the International Monetary Fund. On such occasions, the international community acts in the collective interest of the global economy and the international financial system. That was the rationale for the establishment of the IMF in the aftermath of the Great Depression and World War II; unfortunately too many critics of the IMF today undervalue that role.

This evening I would like to share some thoughts about the role of standstills on sovereign payments obligations in the context of the provision of assistance to countries experiencing external crises. In my view, advocates of the increased use of such mechanisms are well-intentioned, but their supporting analysis is so far inadequate.

I will also outline an alternative and potentially more effective approach to dealing with some of the complex issues raised by external financial crises. I will propose the establishment of an International Financial Stability Fund to pre-position additional financing from private-sector cross-border investors for use alongside conditional IMF resources to help reduce the economic costs of crises.

First on standstills, or payments suspensions, dissatisfaction with the evolution of the Argentine situation, as well as the Turkish situation, has refocused attention on the mechanisms that are used to provide financial assistance to the countries facing external financial crises. A month ago, the Bank of Canada and the Bank of England published a revised paper by Messrs. Haldane and Kruger advocating the increased use of payment standstills in such cases and a reduction in the amount of financing that would normally be made available by the IMF. Two weeks ago, the First Deputy Managing Director of the IMF, Anne Krueger, put forward a proposal to provide conditional legal protection to a country that chooses to suspend payments on their financial obligations and to ensure that rogue creditors cannot overturn any subsequent restructuring agreement-standstills with teeth!

Advocates of the increased use of standstills generally make three arguments: First, they argue that standstills would reduce or eliminate the need for large IMF-led rescue efforts. Second, that they would help to punish investors that have made risky or unwise investment decisions, rather than bailing out the investors. Third, that they would alter incentives, in particular incentives facing investors at a time of crisis, in the direction of greater leniency for the borrower.

I have considerable sympathy with much of this argumentation, especially the third point. Payment standstills do have a role to play in dealing with some external financial crises. Moreover, too often the responsible authorities of countries delay too long in recognizing the intractability of their country's financial circumstances. Consequently, I would welcome the addition of the proposed Krueger tools to the toolkit for dealing with external financial crises. However, I am skeptical whether those tools can be added without damaging other tools already in the box, for example, the role and reputation of the IMF itself. This could have unintended adverse effects on the international financial system. I also suspect that those tools will not be as effective as their advocates claim in the difficult cases. Finally, my sense is that there is not yet sufficient consensus on the desirability of an international bankruptcy mechanism for sovereigns; the Krueger tools consequently go both too far and not far enough in this direction.

Today, the authorities of countries can and do impose standstills on their international and domestic financial obligations; Ecuador is the most recent major example. However, this step is normally taken as a late or last resort. The principal reason for hesitation is not the associated legal risks, as was suggested by Ms. Krueger. The authorities hesitate because of concern about the consequences for their economies, and probably also their own reputations. Consequently, standstills are not likely to be effective tools to reduce the need for large multilateral packages of financial assistance in cases where it cannot be demonstrated that a country's debt profile is unsustainable. Let me elaborate briefly on these points.

First, the advocates of the more extensive use of standstills are primarily concerned that in recent years large rescue packages have created an unacceptable degree of moral hazard for investors and/or debtors. According to this line of reasoning, international investors invest unwisely in the expectation that they will be bailed out, and authorities in the borrowing countries delay the adoption of corrective economic and financial policies in the hope that they will not be needed. These are conceptually consistent arguments, but the advocates of the greater use of standstills, including those at the Bank of Canada and the Bank of England, have yet to demonstrate that moral hazard in either form has increased significantly since the Mexican rescue in 1995.1 In my experience, unsubstantiated theoretical propositions and anecdotes provide an insufficient intellectual foundation for dramatic changes in international financial policy.

Second, advocates of standstills also argue that large packages of multilateral financial assistance impose heavy costs on the international financial community that has to pick up the pieces for the mistakes of investors and policy makers. The fact is that to date no financial costs have been imposed on the international financial community in connection with the large rescue packages of recent years; the IMF has been repaid, often before maturity. Moreover, if the so-called burden of such financing were shifted from the IMF to private-sector investors, the economic costs on the borrowing countries, most importantly, in the form of output declines, are likely to increase, with associated adverse impacts on the global economy.

Third, the argument that the increased use of standstills will help punish unwise decisions by investors who reap high returns and are later bailed out ignores the reality of the evolution of most crises. In general, during the initial stages of an external financial crisis the political authorities of the country want to continue to meet its contractual obligations, in part, and sometimes in retrospect mistakenly, because they see the situation as temporary. This tendency to honor contractual obligations should be encouraged. International financial institutions and the official sector are and should be very reluctant to encourage payment suspensions or other forms of default. As a consequence of the delay, by the time a consensus is established that a country's financial situation is unsustainable, the vast majority of initial investors has either sold out or hedged; the initial investors' exposures are unavailable for capture by a standstill, and they cannot be punished financially for their unwise investment decisions.

Fourth, some proponents of the increased use of standstills, either of the soft variety advocated by the Bank of Canada and the Bank of England or the hardened variety suggested by Ms. Krueger, imply that the use of such a mechanism would have reduced the need for the extraordinary multilateral rescue packages for Mexico, Thailand, Indonesia, Korea, Brazil, Turkey, and Argentina. My judgment is that the more extensive use of such a tool would have altered the fundamental calculus in at most one of those cases, Argentina in August, and the Argentine government would not have agreed then. Only recently has more of a consensus emerged that Argentina's debt burden is unsustainable. In the other six cases, there was never any such consensus. In the absence of consensus on the sustainability of the countries' external financial position, the authorities of these countries understandably and appropriately were reluctant to incur the economic costs associated with formal payment suspensions. Moreover, the three Asian cases did not primarily involve sovereign debt obligations; therefore, standstills on sovereign debt payments would not have been effective in stemming the reversal of capital flows to those countries.

This brings me to my final observation on standstills: their use imposes substantial economic costs on the countries. If standstills are expected to be effective in substantially reducing the size of multilateral financial packages, they have to be comprehensive, which implies the need for capital and exchange controls. Such controls are costly to the country, subject to abuse, and once in place politically difficult to remove, even if there were a presumption, or even a requirement, that the controls should be temporary. Moreover, it is naïve to think that less official financing can be combined with more de facto private-sector financing, in the form of either debt service reductions or reprofiling, without imposing additional economic costs on the country in crisis in the form of deeper recessions. At the same time, standstills impose costs on other countries, in particular those likely to be subject to their own external financial pressures. Those pressures will be intensified if official representatives of creditor countries have advocated the use of standstills and international financial institutions have adopted policies designed to promote their use.

International financial assistance efforts can be too large and too prolonged, but more often those efforts are under-funded. As a result, excessive economic adjustment burdens are imposed by external financial crises on citizens of the countries as well as on the global economy. On the other hand, the critics unfortunately are right about the limits to the availability of official financing that, in the future, can be expected to be mobilized through legislative processes in the creditor countries to be devoted to multilateral financial rescues. They are also right to look to the private sector for larger contributions. However, to wait until a crisis occurs to look to the private sector to play a role risks exacerbating the crisis itself and increasing the economic damage to the country, other countries, and the system as a whole.

Both the official sector and the private sector should look for practical, non-confrontational mechanisms to augment in advance the multilateral financial resources now available from the International Monetary Fund, the World Bank and regional development banks, and creditor governments through the Paris Club to address crises. In this spirit, consideration should be given by the private sector and the official sector to the establishment of an International Financial Stability Fund (IFSF) to help backstop cross-border finance in times of crisis. The IFSF would be a revolving trust fund administered by the IMF. Disbursements in the form of loans would be linked to IMF-supported adjustment programs.

The additional financing would be derived from an annual fee on the use of the international financial system. The fee would be assessed on cross-border investment. The fee would be assessed annually on the average annual outstanding amounts of all forms of cross-border investment-banking claims, bonds, direct investment, and financing provided by nonfinancial institutions.

It would be for consideration whether different types of investment should be subject to different fees, but a flat fee universally applied would be the appropriate starting point. The fee would be applied to cross-border investments in all and by all countries that are members of the International Monetary Fund-the United States and Japan, as well as Brazil and Mozambique. For investments in particular countries, it would be intellectually attractive to use risk weighting to set the level of the fees, for example, on the basis of sovereign credit ratings.

The purpose of the IFSF would be to enhance the stability of the international financial system for the benefit of all participants; there should be an overwhelming presumption of universal participation by all countries and all investors, especially recognizing how difficult it is to distinguish the geographical source or character of investments, ex ante or ex post.

Many technical issues would need to be debated and resolved before the International Financial Stability Fund became operational. An annex to these remarks elaborates on some of those issues.

The proposed IFSF would establish a mechanism through which external private creditors would participate financially in the resolution of external financial crises. This would reduce the need, except in extreme circumstances, to resort to mechanisms, such as standstills, in a futile attempt to penalize unwise investment decisions and to ensure that external private-sector investors bear a greater financial burden in crisis resolution. Such standstills are likely to be ineffective in achieving even-handed private-sector involvement in the resolution of crises, tend to exacerbate the crisis and its economic and financial effects on the country involved, and contribute to the spread of the crisis to other countries, all to the detriment of the global economy and the international financial system.

Again, I thank the Money Marketeers for this award. I also thank you for the opportunity to share some of my thoughts on external financial crises. I have used this opportunity to get some things off my chest. I hope I have also provided some food for thought.


Appendix: Elaboration of the IFSF Proposal

1. Would the fee be applied to investment flows?
No. The fee would be applied to the average annual gross value of all outstanding cross-border investment, i.e., the stock of investment not the flow of investment. Averaging would be necessary for banking and portfolio claims to prevent destabilizing flows around reporting or payment dates.

2. Why apply the fee to all types of cross-border investment?
Given the ingenuity of today's financial markets and their capacity to unbundle and redistribute risks, it is fundamentally impossible to distinguish different forms of investment. For example, a foreign direct investor may still own a stake in a firm, but that investor can finance or refinance that stake in domestic currency, and thereby avoid any foreign exchange risk. I question the popular view among economists that marketable debt in all circumstances is more stable than, and preferable to, bank financing; portfolio investment in all circumstances is more stable than, and preferable to, marketable debt; and direct investment in all circumstances is more stable than, and preferable to, portfolio investment. Moreover, to the extent that direct investment inflows are responsive to underlying economic conditions in the host country, the IFSF would encourage the maintenance of such inflows.

3. Which country would collect the fees?
In principle, either the authorities of the host country to the investment or the investor's home authorities could collect the annual fee; the latter approach might be more efficient though cooperation by the authorities in both jurisdictions would be essential. One by-product of the IFSF would be an improvement in the quality of information on cross-border investment.

4. What might be the size of the IFSF?
A reasonable target figure for the size of the IFSF would be $250-300 billion. An approximate but reasonable estimate of world external assets is $25-30 trillion.2 An annual fee of 10 basis points, or an average of risk-weighted fees on investment in different locations at that level, would raise $25-30 billion a year.3 For the first few years, the IFSF might have to "borrow" from the IMF to cover its share of joint operations. After the initial target amount is reached, the annual fee could be reduced to cover only the growth in the outstanding amount of cross-border investment.

5. What would be the relationship with the IMF?
The IMF would administer the IFSF as a trust fund that would be invested and earn interest.4 Loan disbursements from the IFSF would supplement IMF financial support of adjustment programs. Guidelines on the proportionality of drawings on the IFSF and on regular IMF resources would need to be established; the guidelines would have to be firm, but they might be stated in terms of ranges to provide some flexibility. The IFSF might only be drawn upon in cases of extraordinary access to the IMF's own resources, it might be drawn upon proportionately in all cases, or a progressive formula could be established, i.e. the larger a loan from the IMF as a percentage of a member's quota, the larger would be the proportion provided from the IFSF.5 The IMF might be obligated to consult with a private sector advisory committee as a matter of course or in wake of any activation of the IFSF. Policy conditionality associated with IMF lending involving the IFSF would be affected only to the extent that more financing would be made available through the IFSF in the short run than otherwise would be the case; this would permit less drastic adjustment of the real economy and consequently less drastic adverse implications for the global economy.

6. Should countries be willing to participate in funding the IFSF?
It would be reasonable to expect that the major industrial countries should be willing to participate in the IFSF by collecting the fee on their residents' cross-border investment because those countries now provide, directly or indirectly, the great bulk of multilateral financing in support of adjustment programs. Other countries-Brazil and Mozambique-would have an incentive to apply the fee to cross-border investment by their own residents because their potential receipt of financing from the IFSF would be conditional on their participation in collecting the fee.

7. What would be the impact on incentives for investors?
It is unlikely that the establishment of an IFSF would affect incentives of international investors significantly one way or another. If anything, investors collectively might be somewhat more inclined to monitor the activities of individual investors, since the mistakes of those individual investors would increase the need to replenish the IFSF. With respect to disincentives for investors, the size of the annual fee, for example, 10 basis points per year, would be small enough that it would not be expected to impact adversely expected returns on investments or the flow of international investment. Moreover, in contrast with private-sector contingency financing arrangements, such as those arranged by Mexico and Argentina with foreign banks, investors would not have an incentive to offset any increase in country exposure when the arrangement is drawn upon because payment into the IFSF would have been in advance.

8. What would be the impact on incentives for the borrowing countries?
As long as the potential availability of additional crisis financing through the IFSF is conditional, which it would be, the incentives facing policy makers borrowing from the IMF would be essentially unaffected.

9. What would be the impact on the economic costs of crises?
The potential availability of additional, conditional, crisis financing through the IFSF holds out the promise of somewhat less deep or protracted adjustment processes for the countries facing crises. The reduced need to search for ways to promote private-sector involvement in crisis financing at the time of a country's crisis, for example, through a suspension of payment of external obligations, would reduce the economic costs on other countries, i.e., reduce contagion somewhat.



1 For example, Haldane and Kruger write, "Moral hazard is clearly a question of degree. Every insurance contract possesses some degree of moral hazard. And the empirical evidence on the moral hazard effects of official lending is not conclusive. Nevertheless, anecdotal evidence of the importance of moral hazard is widespread."

2 This estimate is derived from 1999 data on the external assets of 23 countries that collect and provide those data to the IMF to publish in the IMF's Balance of Payments Statistics. The 23 countries account for about 80 percent of global GDP.

3 A fee that size would be in the same ballpark as FDIC deposit insurance assessments although the analogy with deposit insurance is not applicable to this proposal. Deposit insurance is intended to give confidence to individual depositors; the IFSF is intended to give confidence to the international financial system.

4 The trust fund would have to have investment guidelines, presumably providing for investment only in triple-A sovereign obligations. It would be reasonable to reinvest earnings, net of administrative expenses, back in the IFSF, but some might want to divert them to other uses.

5 Requiring consultation prior to activation would defeat the purpose of having resources potentially available to deal quickly with emerging crises.