Strengthening Crisis Prevention: The Role of Contingent Credit Lines -- Address by Stanley Fischer
November 15, 2000Stanley Fischer1
First Deputy Managing Director
International Monetary Fund
Given at the Banco de Mexico
Mexico City, November 15, 2000
Mr. Chairman, Ladies and Gentlemen:
It is a great pleasure to participate in this conference, to celebrate the 75th Anniversary of the Banco de Mexico, now in the sixth year of successful operation under a flexible exchange rate regime. Increasingly, it seems, monetary policy in Mexico is moving towards a formal inflation targeting framework — a shift supported by international experience.2
But we celebrate more than the seventy five years of the Banco de Mexico. We celebrate too the successful recovery of the Mexican economy from the crisis that erupted on December 20, 1994, when Mexico devalued, setting off what Michel Camdessus named the first financial crisis of the twenty-first century. The transition of the Mexican economy from deep crisis in 1995 to sustained and rapid growth in the second half of the nineties helped make possible the historic transition of power now under way in Mexico.
The credit for this belongs in the first instance to the Mexican people, whose cohesion in the face of the crisis allowed the right economic policies to be implemented and take effect. It belongs too to the policymakers — among them President Zedillo, then Finance Minister and current Governor Ortiz, then-Governor Mancera, and more recently Finance Minister Gurria — whose judgment, steadfastness, and courage during the dark days of 1994 and 1995, were essential in producing the remarkable economic performance that followed.
The first financial crisis of the twenty-first century, the Mexican crisis, was followed by five other capital-account-dominated crises — in Thailand, Indonesia, Korea, Russia, and Brazil. All six of these twenty-first century crises took place in the twentieth century, which must be proof that history, if not yet ended, is at least accelerating.
You are all familiar with the intense debate on the reform of the international economic architecture spurred by the crises. Today I want to talk about a new element in the international financial system, the IMF’s Contingent Credit Line facility (the CCL). This, I believe, is a potentially extremely significant innovation that will deploy the IMF’s financial resources more effectively to help prevent financial crises, rather than to help pick up the pieces after the damage has been done.
The basic idea is straightforward: the IMF offers a precautionary line of credit to countries that have demonstrably sound policies, but which nonetheless believe they may be vulnerable to contagion from crises elsewhere. After the recent crises, no-one should doubt the existence and importance of contagion — least of all in Latin America, which was severely affected by the massive contagion from the Russian financial crisis in August 1998. Indeed, the effects of that contagion linger to this day.
I will start by describing how the growth of international capital flows has increased the likelihood and potential effects of contagion, and how, against that backdrop, the CCL fits into the broader crisis prevention agenda. Second, I will describe the genesis of the CCL and the basic principles that underlie its operation. After describing the original design of the CCL, I will outline the modifications that our Executive Board is expected to finalize this Friday, November 17, and the valuable role that I believe the CCL can play in helping to protect individual economies and the global system.
I. Volatility and Contagion in International Capital Flows
Some history first. The expansion of private international capital flows from $40 billion in 1990 to $290 billion in 1997 (after which they declined to $170 billion in 1999) has been one of the most spectacular manifestations of globalization in recent years. These flows have brought economic benefits to borrowers and lenders alike, but — as we have seen too often over the last six years — there is an important downside. Countries have been exposed to periodic crises of confidence when inflows of capital were suddenly reversed.
Cross-border capital flows were already flourishing in the decades preceding World War I. Investors in London or Paris financed everything from American and Australian railroads to Peruvian guano. A quote from Howard’s End (1921), by the English novelist E.M. Forster, conveys the mixture of enthusiasm and wariness towards emerging markets at the time:
“She learned, to her horror, that Margaret, now of age, was taking her money out of the old safe investments and putting it into Foreign Things, which always smash. Her own fortune was invested in Home Rail[ways], and most ardently did she beg her niece to imitate her. ‘Then we should be together, dear.’ Margaret, out of politeness, invested a few hundred in the Nottingham and Derby Railway, though the Foreign Things did admirably and the Nottingham and Derby declined with the steady dignity of which only Home Rail[ways] are capable.”
Capital flows recovered during the 1920s, but were choked off again in the 1930s by the Great Depression and the closing to international trade on both current and capital accounts that accompanied it, and — ultimately — by the outbreak of World War II. When plans for the creation of the IMF and World Bank were laid during the war years, the architects of the new institutions focussed on the current account. But cross-border capital flows among the industrialized countries began to revive in the 1950s and 1960s, and were extended thereafter to what we now call the emerging market countries.
As international capital flows have increased relative to the size of national economies, so has the potential disruption threatened by their reversal. The need to maintain investor and creditor confidence generally serves as a useful discipline. It magnifies the rewards for good policies and the penalties for bad ones. But the capital account crises of recent years drive home the possibility that volatility and contagion can on occasion become excessive.
Models of banking and financial panics demonstrate the possibility of multiple equilibria, in which institutions and economies may be vulnerable to sudden crises of confidence. The impact of these crises can be magnified by herd behavior, a point demonstrated rigorously in the recent literature, but stated more vividly by Charles MacKay in Extraordinary Popular Delusions and the Madness of Crowds (1841):
“Men, it has been well said, think in herds: it will be seen that they go mad in herds, while they only recover their senses slowly, and one by one.”
As a crisis spreads initially, most of those involved — particularly the affected countries — tend to blame excess volatility and contagion, in other words to claim that they were innocent bystanders. This is in part a natural human reaction, in part a rational response to events. Eventually, sometimes sooner, sometimes later, the policymakers in the affected countries get down to solving their problems and there is less talk about contagion, and more talk about the weaknesses in the economy.
Still, which is it — contagion or weaknesses in the affected economy? The answer is both. Establishing the presence of excess volatility and contagion in the system is complicated by the fact that a crisis of confidence can push a country from a good to a bad equilibrium: to put it simply, when a country’s institutions and policies are subjected to massive pressure from a reversal of capital inflows, they may crack, thereby appearing to justify the reversal of flows that caused the crisis. Similarly, contagion does not spread randomly: rather it hits weaker economies more quickly and more forcefully than stronger ones. But, to adopt a medical analogy, just because an epidemic strikes those in ill health more than the healthy, does not alter the fact that it is an epidemic.
Against this backdrop, the international community has confronted a twofold challenge in trying to reform the architecture of the international financial system: to prevent crises where possible and to help resolve them where necessary.
My focus today is on prevention.
One response has been to bolster IMF surveillance of national policies and international markets. We are concentrating more on potential weaknesses that might leave a country vulnerable to a crisis: poor macroeconomic policies; the exchange rate regime; unsound debt management; and weak financial sectors, to name but a few. The power of surveillance has been enhanced by the simultaneous transparency revolution at the Fund. Countries are publishing more and better information about their own policies and economic developments. And a large and growing majority are choosing to publish the advice they get from the IMF. This makes it easier for investors and creditors to take better informed portfolio decisions; no less important, by making it necessary for policymakers to explain to the public what they are doing and why, transparency contributes to better policies.
The CCL is complementary to these efforts, providing an additional financial incentive to adopt good policies. In effect the CCL allows countries that have met certain preconditions to augment — at low cost — the foreign exchange reserves they can draw on in a crisis. The knowledge that these resources are available may in itself serve to deter speculative attacks. In addition, by offering qualifying countries a public seal of approval for their policies, the CCL also makes it less likely that investors and creditors will succumb to herd behavior by pulling out their money when a crisis strikes elsewhere.
II. The Genesis of the CCL
The idea that the Fund should use its financial resources to help countries with good policies cope with exceptional capital outflows has a long pedigree. In 1972, for example, a reference was included in the “Sketch of a Reformed System” prepared by the IMF’s then Economic Counsellor, Jacques Polak, in the aftermath of the Smithsonian currency realignment.3 The paper raised a number of practical questions that would be posed by the creation of such a facility. These questions have remained relevant in subsequent discussions: Which members should be entitled to help? What policy conditions should be imposed? How much money should be available? And on what terms should the assistance be offered?
The suggestion was not acted upon at the time. But it resurfaced in late 1994 — at the time I joined the Fund, when the IMF’s executive board tentatively discussed creating a “Short-Term Financing Facility”. The staff suggested that that Fund could approve credit lines for fixed periods for countries that: (i) had strong policies that were expected to remain strong, and (ii) did not confront a fundamental balance of payments problem. With access to the credit line automatic up to an agreed amount, this was seen as a way to send a clear positive signal to financial markets.
Fund staff argued that such a facility could help “give confidence to members that the Fund would have the capacity to provide very prompt financial support in certain circumstances. Such support could help to reinforce appropriate policies and avoid resort to exchange restrictions in response to short-term financial market disturbances”.4 The facility could be used “where a member with fundamentally sound economic policies and prospects faced very short-term external pressure stemming from events largely outside its direct control, that seemed likely to reverse or to dissipate relatively quickly without major changes in underlying policies”.
The proposal proved contentious. But the debate was short-circuited by the outbreak of the Mexican crisis. Mexico had been prominently cited in the debate as a potential recipient of the facility, and its crisis undercut the arguments for the facility.
The idea finally moved towards fruition in the wake of the August 1998 Russian crisis, contagion from which was visible to all. The concept was discussed favorably at a meeting of Western Hemisphere finance ministers and central bank governors at the IMF on September 3–4. On September 14, in his famous Council on Foreign Relations speech in New York, President Clinton said “this is the biggest financial challenge facing the world in a half-century”, and outlined a five-point plan to tackle the global market turmoil. Among them: “We need to consider ways to extend emergency financing when countries are battling crises of confidence due to world financial distress as distinct from their own errors in policy”.
The suggestion was picked up by the Group of Seven finance ministers and discussed on October 4 by the Interim Committee at our annual meetings. As the Interim Committee communique said:
“The Committee agreed to explore a strengthened capacity, based in the IMF and together with the general increase of IMF quotas and establishment of the New Arrangements to Borrow, to provide more effectively contingent finance to help countries pursuing sound policies to maintain stability in the face of difficult global financial conditions.”
The discussion was then taken up by the IMF board and staff. Agreement on the structure of the Contingent Credit Line facility was reached in April 1999. The agreement marked a watershed in the role of IMF lending, significantly extending the circumstances under which countries can prequalify for financial support from the Fund. Let me expand briefly on this point.
III. Principles Underlying the CCL
Traditionally, the IMF has offered financial support to countries already in trouble, typically facing short-term balance of payments problems on the current account. The rationale, in the words of the Articles of Agreement, is that by offering temporary support in such circumstances, the Fund enables countries to “correct maladjustments in their balance of payments without resorting to measures destructive of national or international prosperity”, such as draconian contraction of domestic demand, trade restrictions, or competitive devaluations.
When capital flows were restricted, the adjustment that a country needed to make to deal with a crisis would be set by the scale of the external imbalance, rarely more than 5 percent of GDP. This was reflected in the rules determining how much a country can borrow from the Fund under a conventional stand-by arrangement — normally up to 100 percent of its quota, its contribution to the Fund’s capital base, in any one year, with a cumulative maximum of 300 percent of quota.
Capital account crises present more of a challenge. When a country confronts a crisis of confidence over its ability to repay its foreign creditors, the scale of its financing problem is set not by the current account deficit, but by potential capital flows, related to the size of its external debt, and sometimes also to the scale of the internal debt. Of course, the official sector cannot and should not meet this entire financing need. But the official sector does have an important role in helping countries avoid the full burden of adjustment that would otherwise be needed. It can achieve this not only by providing money itself, but also through its catalytic role, by restoring confidence among private creditors and lenders.
When Mexico succumbed to a capital account crisis in 1994, followed by Thailand and Indonesia in 1997, the Fund lent them much more than the standard 100 percent of quota per year by invoking the “exceptional circumstances” clause. Other institutions and bilateral lenders contributed further support. But it was clear that a different mechanism was needed to deal with capital account crises. Two weeks after Korea was granted a stand-by loan for over 1900 percent of its quota in December 1997,5 the Fund established the Supplemental Reserve Facility (or SRF) — offering loans without access limits, but with higher interest rates and shorter repayment periods. Korea’s loan was converted to the SRF, and the facility was later used also by Russia and Brazil. Its use is now under discussion with Argentina.
Like the SRF, the CCL is aimed at capital account rather than current account crises. As with the SRF, there is a recognition that relatively large sums of money may be needed so there is no formal access limit. There is however an indicative range for access: between 300 and 500 percent of quota. Given the threat that contagion could strike several countries at once, the board also agreed that the Fund’s overall liquidity position would be taken into account in deciding the scale of access in any particular case.
With both normal stand-by arrangements and the SRF, the Fund is picking up the pieces after an accident has happened. The CCL marks a clear departure from this model. Extending credit lines to countries before problems break out means that the Fund’s lending capacity can be used to serve the cause of crisis prevention, as well as crisis resolution.
There is obviously a risk of moral hazard here. Countries have an incentive — in theory at least — to run weaker policies if they have an extra financial cushion in place. Perhaps more importantly, investors have an incentive to lend to countries with weaker policies if they believe that the presence of an IMF credit line increases the chances that they will be repaid if things go wrong. To counter this problem, the CCL is aimed explicitly at members with first-class policies, who face a potential loss of access to international capital markets because of contagion, rather than domestic policy weaknesses. The four eligibility criteria have been designed accordingly:
- First, at the time the credit line is extended, the recipient country must not be expected to need to borrow from the Fund. This implies that over the medium-term the country should be expected to be able to finance its balance of payments comfortably, based on a realistic assessment of its access to private capital and a sustainable projected path for external debt. Risk scenarios and sensitivity tests will be necessary to establish the strength of the country’s external position.
- Second, our Board must have made a positive assessment of both the country’s economic prospects and its progress towards meeting international policy standards:
A positive assessment of the country’s economic prospects obviously requires that it should not face a significant risk of a self-inflicted balance of payments crisis; this implies that its policies should be sufficiently sound to meet the conditions that would be attached to a standard post-crisis loan. The policies under scrutiny would include the health of the financial sector, which was an important factor in the recent Asian and Russian crises.
Four standards are key in judging progress towards meeting international standards: (i) the Special Data Dissemination Standard, (ii) the Basle Committee’s Core Principles for Banking Supervision, (iii) the Code on Fiscal Transparency, and (iv) the Code on Transparency in Monetary and Financial policies. Countries must be making satisfactory progress towards meeting the data standard if they wish to qualify for a CCL; the other standards are judged as a group and a country must be making adequate progress towards meeting a “critical mass” of them if it is to qualify.
- Third, the country must both enjoy constructive relations with its private creditors and be taking appropriate measures to limit its external vulnerability.
The former means that the country has in place a good framework for debtor-creditor relations. It also means that the country has made arrangements to secure its access to private credit in the event of a crisis — or that it is making credible efforts to put those arrangements in place. Such arrangements could include: effective debt-management procedures, strong domestic bankruptcy regimes, private contingent credit lines, renegotiation-friendly bond contracts, frameworks for debtor-creditor discussions, and the issuing of debt with put-options that make it possible to extend the period for repayment.
Several factors need to be looked at in judging whether a country has taken appropriate measures to limit its external vulnerability. These include the nature of the exchange rate regime and whether the currency is overvalued. They also include the size, maturity and currency composition of the country’s external debt. Other indicators include the country’s stock of net international reserves (including in relation to short-term debt), and the net foreign asset position and asset market exposure of its commercial banks.
- Fourth, the country must outline to the Board the policies that it intends to pursue during the one-year period covered by the CCL. These have to be supported with a quantified quarterly program that covers in detail the outlook for fiscal policies, monetary policy and the balance of payments. However — and this is an important difference — unlike a post-crisis program supported by an IMF loan, there would be no formal performance criteria or benchmarks against which performance would be assessed.
The insistence on first class policies as a precondition for the granting of a CCL, and the absence of formal performance criteria or benchmarks, as well as the size of the CCL, mark a contrast with the precautionary stand-by arrangements that the Fund has offered to countries with potential current account difficulties since the 1950s. A country with a precautionary standby can draw on it at any time provided it has passed all the necessary reviews. It is thus much closer to a regular standby than would be a CCL, where the basic assumption is that the country already has a strong economy and is pursuing good policies.
One criticism of the CCL is that it would be very difficult to withdraw a country’s access to its credit line if the quality of its policies slipped after the CCL was in place. The assumption is that the IMF Board would be reluctant publicly to withdraw its stamp of approval, for fear of triggering a crisis that would then be expensive to resolve. Equally, countries might be concerned to enter a CCL for fear their access might later be withdrawn. To some extent, this problem is inherent in the certifying and incentive function of the CCL: it is not possible both for access to the facility to be a seal of approval, and for its withdrawal to send no signal at all. However, this problem should not be exaggerated: it is unlikely that the markets would not have noticed the policy slippages that would cause withdrawal from the CCL; and we should also recall that there are many occasions on which negotiations on the renewal of an existing Fund program have been suspended and the market reaction has been relatively subdued.
If countries do have good policies — and these are as clear to the private sector as they are to the Fund — then it is legitimate to ask why the task of providing contingent credit lines should not be left to private institutions? Indeed, Argentina, Mexico, South Africa, and Indonesia have all arranged lines of credit with private banks that they can draw upon in times of difficulty.
In practice, countries have had difficulty using these credit lines. Mexico decided to draw on its credit lines in the wake of the Russian crisis. The drawing was clearly in line with the terms of the arrangement, but the banks put strong pressure on the authorities not to use it — arguing that it was unnecessary and that it would hurt Mexico’s creditworthiness. In the event Mexico went ahead. Yields on the country’s debt rose by 100 basis points in the immediate aftermath, but quickly fell back.
Private credit lines can play a valuable role, but their effectiveness could by offset by dynamic hedging on the part of the banks extending them. Banks could offset the increase in exposure implied when a country draws on its credit line by reducing their exposure to that country in other forms — taking with one hand what they give with the other. And if the banks are concerned by their exposure to emerging markets as a whole, then they may also reduce their exposure to similar countries as well — thereby spreading contagion.
IV. Experience with the CCL
In the first year of the CCL’s existence, a number of countries discussed with Fund staff whether they should apply for one. But in the end none did. Potential users pointed to several weaknesses in the original design. These were discussed in September, 1999 by the Executive Board, which agreed to important changes to make the facility more attractive. The Board is expected to formalize its decision on these changes this Friday, November 17.
Let me discuss in turn the four main factors that may have deterred countries from seeking a CCL: the pricing of CCL loans; the mechanics of drawing on the facility, the global economic environment, and; the possible reaction of financial markets to the decision to apply for one.
- First, the pricing of the facility. In the original design, CCL loans, if drawn, would have been priced at the same rate as an SRF — 300 basis points over the Fund’s standard rate for the first year of the loan, rising by a further 50 basis points every six months up to a maximum surcharge of 500 basis points. But since the international community should support countries that adopt policies that limit their vulnerability to crises, it is reasonable to impose a lower charge for countries whose policies meet the crisis-preventing standards of the CCL. The Board has now agreed in principle to lower the surcharge over the standard loan rate to 150 basis points, rising to 350 basis points. The commitment fee countries pay when securing a CCL will also be cut. This will increase the financial incentive for countries to adopt sound policies.
- Second, there were concerns about the process of drawing on a CCL. Under the original rules, it was far from automatic that a country that with a CCL would be able to get its hands on the money when contagion struck. Before it could do so, the Board had to determine: (i) whether the crisis was the result of contagion; (ii) whether the country had lived up to its policy promises; and (iii) whether additional policy measures were required.
There is an obvious trade-off here between the desire to insulate countries from crisis and the need to ensure that there are adequate safeguards on the use of Fund resources. In light of the strong track record required of countries that successfully qualify for the CCL, the Board has agreed to move in the direction of greater automaticity in the release of the first third or so of the resources committed under the facility.
In particular, it will give countries “the strong benefit of the doubt” as regards their future policy intentions. In determining whether the country had lived up to its policy promises, the Board will simply verify that the balance of payments problems are not a direct result of the country’s own policies.
- Third, the global economic environment. One reason countries may have chosen not to apply for a CCL is that did not feel threatened by contagion. Until recently, gross inflows of private capital into emerging markets have been recovering — albeit unevenly — since the beginning of 1998. This reflects improving macroeconomic fundamentals in the emerging market countries and favorable liquidity conditions for most of the time in the advanced economies. The EMBI spread — the heart monitor for financial conditions in the emerging markets — also dropped significantly from its levels in the midst of the crises. But conditions have recently become a bit more difficult. In any case, calm waters are no cause for complacency, contagion is unfortunately a very real threat in the global capital markets, and the CCL would still be a useful insurance policy to have in place.
- A fourth factor that may have deterred countries from asking for a CCL is the fear that financial markets would regard an application as a sign of weakness, as a signal that the country has good reason to believe itself vulnerable to a crisis. There are two reasons this should not be a concern for the markets: first, the stringent prequalification requirements should signal that the country’s economy is strong; and second, access to the CCL is in effect a way of adding to reserves at low cost — and the markets generally regard higher levels of reserves with favor.
The introduction of the CCL is a potentially highly valuable addition to the armory with which the IMF helps protect its members from the enormous economic and social costs of financial crises. It means that the IMF’s financial resources — as well as its surveillance efforts — can be used more effectively for crisis prevention as well as crisis resolution. Problems are less likely to occur if countries have a financial incentive to adopt sound policies and measures that will make them less vulnerable to sudden reversals of capital inflows.
The fact that countries have so far chosen not to take up the CCL may in part reflect the relative calm that was returning to the international financial system since the invention of the facility late in spring 1999. But a more important reason is that the original design had serious flaws. Those flaws have been corrected, and the changes which our Board is set to finalize later this week will make the facility more attractive and more user-friendly.
I hope and believe that we will in the coming months see several countries enter the CCL facility — and that they and the world economy will be the better for it.
1 This paper was prepared for presentation at the 75th Anniversary Conference of the Banco de Mexico, Mexico City, November 14–15, 2000. I am grateful to Robert Chote and Tessa van der Willigen for assistance.
2 I was originally asked to talk about the challenge of designing IMF-supported programs in countries that have adopted inflation targeting. An excellent paper on this topic by Mario Blejer and other colleagues from the IMF was presented to the conference on “One Year of Inflation Targeting” in Rio de Janeiro on July 10, 2000. I will return to the question later this month at a conference on inflation targeting to be hosted by the Central Bank of Chile.