Thank you Mr. Chairman for inviting me to discuss with you today some of the lessons learned from the various financial crises of the past two years and the responses to them. I will concentrate primarily on the Southeast Asian countries, but I will briefly comment on the Russian and Brazilian situations. I will close with some suggestions as to how policies might be improved going forward.
For convenience, I have summarized all of the key points I want to make in Table 1, which is attached. The table and the testimony attempt to identify where I believe consensus among economists and—more generally among G-7 policy makers—has been reached and where views remain divided. Where appropriate, I add my own views.
It is always easier to identify how crises developed with the benefit of 20-20 hindsight. This is clearly true with the Asian and Russian financial crises, which have written about extensively in the academic journals and popular media.
I find it useful to think of these financial crises—indeed of any financial crises—as fires that have raged out of control, burning many victims in the process: in this case, whole populations of Southeast Asia, Russia and elsewhere; many investors around the world; and even some workers here in the United States (although far fewer than was feared at the time). The fires started with a spark—in the case of Asia, the failure of some Thai banks in mid-1997 and in Russia with the government’s bond default and devaluation in the summer of 1998—but they spread quickly and with such ferocity because they were fueled by lots of gasoline.
In the case of Asia, it is now widely agreed that the gasoline came from several sources:
- Asian companies and banks borrowed far too much in foreign currency (mostly dollars) with short maturities. They were encouraged to do so by the pegged exchange rate policies of their governments. With a fixed exchange rate, Asians could borrow at single digit interest rates in dollars and invest the proceeds at double digit rates of return in local currency. This “carry trade”—as it has come to be called—worked, however, only so long as the exchange rates stayed truly fixed. When it appeared that central banks had insufficient hard currency reserves to maintain their exchange rate pegs, borrowers rushed to sell their own currencies short for dollars, depleting those reserves and thereby forcing the governments to let their exchange rates float—or more accurately, drop like stone.
- Totally apart from the carry trade, financial institutions within the region loaned far too much to domestic borrowers, especially for large construction projects which proved to be financially unsound. The large amounts of bank lending were fueled by the influx of foreign funds and the lax supervision of domestic regulators, coupled with the implicit guarantees of local governments that the banks and/or their borrowers would be bailed out if things went wrong. Domestic government guarantees created the well-known “moral hazard” problem (the tendency of insured parties to take greater risks because they know they have insurance). The heavy leverage of many borrowers, their relatively poor returns on investment in the years preceding the crisis, and the close ties between banks and many of their borrowers—what many have called “crony capitalism”—also helped create the conditions for the financial conflagration that followed.
I have not separately singled out in Table 1 the various factors within each of the countries that analysts have also widely agreed contributed to excessive borrowing in these countries. These include lax or non-existent systems of corporate governance, poor financial supervision, and poor accounting practices. All played their role, a theme I will return to shortly.
The borrowers were not the only ones at fault, however. For every borrower, there must be a lender or an investor. Why were foreign investors—especially banks—so eager to shower Asia with money? For example, IMF data reveal that after hovering in the $30 billion range for most of the 1990’s, foreign direct investment into five Southeast Asian countries doubled to $63 billion in 1995 and hit a peak of $73 billion in 1996, before plunging to a negative $11 billion in 1997.
Similarly, annual bank lending to the region hovered in the $15 billion range from 1990 to 1994 before jumping to roughly $40 billion in 1995-96, before taking a nosedive to a negative $32 billion in 1997.
I can’t give you a good explanation for the sharp jump in FDI before the crisis other than to say that investors must have widely believed all of the stories about the “Asian miracle.” I do not believe they were comforted by the prospect of an IMF rescue, In fact, when the Asian crisis came, equity and bond investors in the regions took heavy hits.
About the big jump in bank lending there is much more controversy. In particular, critics of the IMF argue that foreign banks parted with their money so eagerly because they believed the IMF would bail them out (indirectly by channeling money to local governments first). This is the classic moral hazard problem, but on an international scale. At the same time, it is also true that much of the lending to Southeast Asia was provided by Japanese banks, which in retrospect appeared to be gambling on their own resurrection from their lending losses in Japan and elsewhere. To this extent, the Japanese banks probably were more influenced by a belief that their government would bail them out than by the prospect of a rescue engineered by the IMF.
In any event, there now appears to be a consensus on two things about the pre-crisis era. First, whatever one believes about the IMF, the international capital standards for banks set by the Basle Committee almost certainly encouraged excessive inter-bank lending to Asian banks by requiring only 1/5 the amount of capital to be held against such loans as the standards required for other types of loans. The Committee proposed in June 1999 a new system of capital standards designed, among other things, to rectify this problem.
Second, there seems to be little dispute that the prospect of an IMF bailout encouraged investors to snap up Russian GKOs issued in the summer of 1998. Indeed, these bonds—which carried extraordinarily large interest rates—were widely known on Wall Street as a “moral hazard play” because Russia was believed to be “too nuclear to fail.” The IMF proved them wrong, but at the cost of triggering a temporary panic in the financial markets that did not subside until the Federal Reserve twice lowered interest rates in October 1998.
Before leaving the subject of causes, it is important to note the fairly broad consensus about two factors that were not significant contributors to at least the Asian crisis. First, unlike Mexico—where macroeconomic imbalances helped trigger a loss of confidence in the peso in 1994—the Southeast Asian countries generally were pursuing appropriate macro policies before they experienced a financial crisis. Budgets were not deeply in deficit. Inflation was not out of control. Russia and Brazil are notable exceptions, however. In Russia, the budget has been deeply in deficit because of an inability to collect taxes, while Brazil has had a well-known problem with excessive government spending, especially generous pension costs.
Second, despite criticism in some quarters, so-called “hedge” funds were not major factors behind the currency collapses in Asia. Barry Eichengreen, formerly of the IMF and now back at the University of California at Berkeley, led a team of scholars at the IMF demonstrating this to be the case. More recently, David Hsieh of Duke’s Business School presented a paper last week at the third annual meeting of the Brookings-Wharton Papers on Financial Services, which showed hedge funds to be minor players in currency trades during 1997, the year the crisis began.
The conventional wisdom—which, in this case, I believe to be correct—is that the various currencies collapsed because domestic residents (many of whom had borrowed extensively in dollars) lost confidence in the commitments of their local governments to maintain a pegged exchange rate. When this occurred, they ran for the proverbial exits, by demanding their central banks to exchange local currencies for dollars. When the central banks ran out of money, they came to the IMF.
Responses to the Crises: The IMF
The IMF was there when the calls for hard currency loans went out. In effect, by providing loans—with significant strings attached—the IMF acted as an international financial “fire department,” using money instead of water to stop the financial contagion. Unlike the local fire department, however, the IMF did not have an unlimited amount of money to throw at the problem. Instead, as this Committee well knows, a raging debate ensued in the Congress over whether the United States should have contributed additional resources—the financial equivalent of more water—to ensure that the IMF could continue to play its fire department role effectively in the future. Eventually, the Congress agreed to do so, but not without attaching some significant conditions to authorizing the money—perhaps most importantly, that the IMF itself become more transparent in the way it operates [which it has since done].
I will not rehearse the by now well-worn arguments over the IMF and whether it should continue to exist. That battle for the time being is over. The prevailing consensus is that, at least for the foreseeable future, the Fund should stay in business.
The more constructive exercise—and the one I understand this hearing wants to pursue—is what lessons can be drawn from the IMF’s emergency lending during the recent crises. Under the “IMF Response” heading of Table 1, I begin with those lessons on which there appears to be a consensus:
- There is general agreement that the IMF lending stopped the fires, but not without a lot of pain, some of it unnecessary. In particular, even the Fund concedes that its initial budget deficit conditions were too tight and caused the affected economies to contract further, before the IMF loosened the targets.
- There is also a fairly strong consensus—which former Treasury Secretary Rubin announced as policy shortly before he left office—that the Fund should no longer support pegged exchange rate regimes, which as I’ve just noted, encouraged much of the excessive foreign currency borrowing in Asia that was a precondition for the crisis. In fact, most of the developing world now has adopted some type of floating exchange rate system.
A distinct minority view, however, urges smaller, emerging market countries to establish currency boards, or even to go all the way and dollarize. The idea here is that countries can purchase the reputational value of the dollar, and the lower interest rates it brings, by giving up their local currencies. Doing so, however, puts countries in a macroeconomic straightjacket, potentially depriving their central banks of the ability to act as a lender of last resort. Furthermore, countries that adopt currency boards or that dollarize in effect export their monetary policy to another country—specifically the United States—which is not eager (at least formally) to be the central banker for other countries. The subject of the appropriate exchange rate regime for various countries is a complicated subject, however, and one which I will not dwell on at length.
- There is also an emerging policy at the Fund—which I believe has general support—that not every country will receive a large lending package as a matter of right. Leading examples are the Fund’s refusal to lend to Russia in August 1998, following that government’s default on its bonds, and more recently the refusal to support Ecuador after it announced suspension of payments on its Brady bonds. In each of these cases, creditors—including banks—took losses. By taking a hard line and refusing to lend in both situations, the Fund introduced into the market what central bankers call “constructive ambiguity”—or some doubt whether rescues will follow, which helps reduce any moral hazard associated with Fund lending packages. A question I address later is whether the IMF in the future should be more explicit about forcing bank creditors in particular to take losses in all cases in which the Fund extends loans, or should instead continue to assess this on a case by case basis.
A number of things the IMF has done in response to the crisis have generated much controversy, however, and I believe it is fair say, have not drawn a consensus. Table 1 lists these areas as well:
- Perhaps the most intense debate is over whether the Fund was correct in
insisting that each of the borrowing countries follow the orthodox policy of defending a falling exchange rate: by maintaining tight money and thus temporarily high interest rates. To some extent the issue has become academic because interest rates in all of the countries involved—even Russia—basically have returned to pre-crisis levels, as shown in Chart 1. Similarly, with the exception of Brazil and Russia that deliberately devalued and Indonesia, which has been racked by civil conflict and political uncertainty, Chart 2 shows that the currencies of each of the Asian countries have since recovered most of the ground they lost during the crisis.
Nonetheless, argument continues over whether the tight money policies recommended by the Fund caused an unnecessary degree of pain during the transition. The critics claim that raising interest rates in economies full of over-leveraged firms caused an excessive number of bankruptcies, which in turn depressed confidence among local consumers and firms. The resulting decline in spending and output arguably had the perverse effect of weakening, rather than strengthening, the exchange rates. The Fund and its defenders respond that easier money would have made the free fall in exchange rates even worse and thus greatly magnified the local currency debt burden of local firms and banks that had borrowed extensively in foreign currency. In effect, many firms were going to fail either way, but at least with higher interest rates, it was more likely that the currency would strengthen than collapse.
I have two points to add to this debate. For one thing, it is unlikely to be resolved definitively any time soon. For another, it is easy to blame the Fund with 20/20 hindsight. The right question to ask is whether at the time, given all of the uncertainties involved, the Fund made a reasonable choice in insisting on orthodox tight money policies. I lean slightly in favor of an affirmative answer, but believe the issue is a close call.
- Debate also continues over the various “structural conditions” the Fund imposed on its lending: improvements in domestic financial regulation, accounting practices, corporate governance, bankruptcy regimes, and the like. Critics say these conditions intrude on national sovereignty, are difficult to enforce, and encourage a backlash in these countries against the IMF and the West more broadly. Defenders respond that the firms and banks borrowed so heavily in foreign currency because of these structural defects and that it is exactly when the Fund is lending money that it has the most leverage to effect necessary change.
Again, I have two points to add to this particular discussion. First, as I understand it, the Fund has not attempted to imposed its own standards in these various areas on its borrowing clients. Instead, where they are available, the Fund has used internationally accepted norms or standards (such as the Basel standards for bank capital and financial supervision).
Second, as Tom Friedman of the New York Times has eloquently written in his new book The Lexus and the Olive Tree, countries that want to grow economically are going to have to put on some form of the “golden straightjacket”—by which he means some of form of Western-style, transparent capitalism—if they want to attract foreign capital from the private markets. The IMF, in effect, has attempted to short-circuit this process by not waiting for private capital to fully return to the countries affected by crises, and instead conditioning its loans on adherence to global financial standards.
This is not an unreasonable position to take in the wake of the crisis, but I do not believe in the longer-run this is the best approach to follow. Instead, the objective should be to return to private market discipline. This, in turn, will require accurate and timely information about the extent to which countries are pursuing reforms—a subject I will address below.
- Another area of controversy surrounds the Fund’s new “Contingent Credit Line”, or “fast money” for countries judged to be following suitable macroeconomic and structural policies but susceptible to a contagious attack on their currencies. The idea behind the CCL is commendable: local residents and foreign investors presumably will not test a country’s currency if they know that plenty of foreign exchange reserves are available to support it.
Nonetheless, the CCL remains largely an untested idea. More importantly, the CCL contains a well-recognized moral hazard problem: once countries secure the credit lines up front, their commitment to policy reform can lapse. At that point, they have the Fund over the proverbial barrel: if the IMF yanks the credit line, it runs a severe risk of precipitating the very crisis its advance lending was designed to prevent.
Responses to the Crises Within Emerging Market Countries
Economists and G-7 policy makers, as well as many policy makers in emerging markets, also have reached a consensus on what policies emerging market countries themselves should follow to avoid crises in the future. These lessons, too, are shown on Table 1:
- In retrospect, it clearly was a mistake for the West to have encouraged countries with weak financial systems to liberalize their restrictions on the influx of short-term foreign funds, such as inter-bank loans and portfolio investments (bonds and equities traded in the open market). Too much money in the hands of parties that are not properly supervised or encouraged to behave carefully is literally like pouring gasoline on the floor and waiting for the proverbial match to be lit. There is now widespread (but not universal) agreement that the IMF and the West should tolerate, if not encourage, emerging market countries to maintain some types of disincentives to discourage the excessive inflows of “hot money.”
- At the same time, however, there remains a consensus—with one notable exception—that emerging market countries should not resort to restrictions on capital outflows in an effort to staunch a crisis. The exception, of course, is Malaysia, which imposed such controls and whose economy, like others in Southeast Asia, since has bounced back somewhat. But Malaysia also began to relax its capital restrictions shortly after they were introduced. Furthermore, Malaysia runs a real risk of having discouraged for some substantial period of time far less foreign direct investment from coming into the country than it might otherwise have received. Indeed, the international policy making community remains united in the belief that FDI is crucial for enhancing the growth and stability of emerging markets, a topic I will return to shortly.
The Asian Recovery
Chart 3 shows that GDP growth the Southeast Asian countries where the crisis began already has rebounded in a classic V-shaped recovery pattern. It is difficult to be as optimistic about Russia, if only because the official statistics there are notoriously incomplete and essentially miss the huge underground, barter economy.
Brazil is a mixed story. The good news is that the devaluation of the real earlier this year did not trigger another round of financial crisis, as some had feared. In my view, this was due in large part to the IMF lending package that helped cushion the fallout from the devaluation. At the same time, Brazil’s fiscal deficit remains large and its interest rates are high (although not as high as before the crisis when the central bank tried to maintain the value of its currency).
The big mystery is why the Asian economies in particular have bounced back as so quickly, especially when so much of the debt of the corporate sector seems not yet to have been restructured. This spring, Brookings is co-sponsoring a major conference with the World Bank and the IMF to explore this question, and maybe by then I can give you some answers. All I can say now is that the relatively rapid recovery in GDP growth in Asia demonstrates, at the very least, the underlying resilience of the economies in the region, coupled with several other factors: the restructuring of the banks and the gradual improvement of financial regulation, the return of some foreign direct investment to the region, and the hard work and energy of the Asian people.
Suggestions Going Forward
It is tempting to say that little about the “global financial architecture” has changed since the crises. But this is a false view. Significant progress has been recorded. It is significant that the IMF has become more transparent, less willing to impose fiscal austerity on countries that don’t need it, has taken a u-turn on capital mobility and indicated a willingness to tolerate countries with weak financial systems to discourage short-term borrowing in foreign currency, and has been willing to say “No” in at least a couple of cases to large lending packages. Furthermore, it is significant that emerging market countries are moving away from fixed exchange rates and appear to be making at least some progress toward reform of their financial infrastructure. Indeed, the greatest danger now is complacency: with recovery in many places under way, it is easy to say that no further work remains to be done.
This would be unfortunate because the reality is that the world has not and probably will never come up with a magic bullet to end all financial crises. The policy challenge remains: how to minimize both their frequency and severity. So, what remains to be done now?
First, radical reform—whether international financial regulation at one extreme, or abolition of the IMF at the other—seems to be off the table. I will discuss neither in further detail here.
Second, many economists and policy makers nonetheless remain concerned that the IMF’s implicit policy of constructive ambiguity is too new and sufficiently untested to convince bank lenders to emerging market banks that their money is really at risk. It is true that the banks have had to extend the maturity of their loans to the Asian countries, but at least in the case of Korea they received sovereign guarantees for doing so and continued to collect some interest. If bank lending to emerging markets is going to be more disciplined in the future—especially once the pain of this crisis recedes into distant memory—bank lenders are going to have to face the risk of the much larger “hits” that investors in emerging market debt and equity now confront. This is much easier to say than do, however, and many of the approaches that have been advanced for doing that have their own complications—and for that reason, neither has attracted what I would call a consensus.
One idea that I and some academic colleagues who specialize in financial institutions have favored is for the Fund to establish an automatic policy of refusing to lend to countries (more than their quota of hard currency reserves held with the IMF) unless they find some way of penalizing foreign currency lenders to their banks. Alternatively, the Fund could charge lower interest rates on loans to countries that follow such a policy.
For example, one such penalty might be to say to the bank lenders: as long as an IMF package is in place, you must roll over your loans or else suffer some some kind of “haircut.” In addition, the Fund could insist on countries not issuing sovereign guarantees to creditors, as Korea did in late 1997. A related idea has been advanced by Willem Buiter, a distinguished economist in the United Kingdom, who together with a colleague (Ann Sibert at Birbeck College in London) has suggested that bond and loan contracts contain automatic roller provisions, to be exercised at the discretion of the borrower. Yet another proposal is the one suggested by the Council on Foreign Relations Task Force Report, which I will discuss in greater detail in a moment. Specifically, the Report urges the Fund include as part of its conditions a requirement that those countries judged to have unsustainable debt burdens, “engage in ‘good faith discussions’ with their private creditors with the aim of reaching a timely agreement on a more sustainable debt and debt-servicing profile.” (The Report does not mandate, however, that creditors take a loss or extend the maturities of their loans, as would be the case with the other two proposals).
I heard several objections to these kinds of arrangements. One is that the prospect of having to suffer some kind of penalty may provoke “anticipatory runs” by creditors of emerging market country banks who fear that an IMF rescue is about to be needed, debt restructurings are about to take place, or that the bank may invoke the rollover clause. I personally do not think this to be as serious an objection as it is made out to be because if such runs occur, it is highly likely that the country would have approached the IMF for assistance in any event; the sooner that happens and the country is put on some kind of conditionality program, the better off it (and the international community) will be.
A second objection to any plan requiring actions by the IMF is that when push comes to shove the IMF may back down from insisting on the bank penalties—or, in the case of the Council’s proposal, the Fund may avoid finding that a country’s debt burden is unsustainable. This is a problem of credibility, one that the Fund has in any event. But the virtue of announcing a clearer presumption in favor of burden-sharing by all segments of the private sector (including bank lenders), is that it makes it easier for the Fund to resist pressure from international banks to be made whole than is true when the Fund decides its policies on a case-by-case basis.
The Council’s thoughtful Task Force report, prepared under the chairmanship of Carla Hill and Peter Peterson and directed by Morris Goldstein, suggests yet another approach to the moral hazard issue. The report urges the Fund to place strict limits on the size of its conditional lending packages to individual countries (at some multiple of the country’s quota). This would limit the ability of borrowing countries to bail out creditors of their banks or other firms. At the same time, the CFR report urges the creation of a special “systemic risk fund” that would be paid for by developed countries, such as the United States, through a one-time contribution of Special Drawing Rights or SDRs, which the report claims would not need Congressional authorization (but consultation instead).
Countries would be eligible for this money without strings attached if they could demonstrate they were about to be victims of contagious runs on their currencies but were otherwise following good economic policies. In effect, this systemic risk fund would be like the IMF’s new CCL, except that it would be separately paid for, and ideally, have more funds on which to draw. This is an interesting idea but, it too, has some drawbacks, as the report candidly admits. I think the most significant one—although not necessarily a discussion-ending one—is that the developed countries administering the systemic risk fund may be tempted to classify a country’s situation as posing a “systemic risk” in order to circumvent the tight limits on the conventional funding packages that the CFR would like to see imposed. The response is that since the developed countries are paying for the facility they wouldn’t abuse it. But one doesn’t know in advance that this would be the outcome.
In the end, if these more far-reaching proposals are not accepted but national policy makers nonetheless want less moral hazard from IMF lending, they can urge the Fund, as it is now structured, to “just say No” to more large lending packages on a case-by-case basis. Over time, this would convincingly demonstrate to the markets that a constructive ambiguity policy really does exist.
Third, more remains to be done about cleaning up the wreckage from financial disasters after they happen, both to make the transition toward recovery easier and to provide better assurances to creditors in advance that such a process exists. The desire to accomplish both these objectives is behind the Fund’s insistence that countries adopt and implement better bankruptcy systems. But realistically, this cannot and will not be done overnight. Effective bankruptcy requires not only an appropriate set of rules, but an independent, well functioning judiciary. Even we in the United States continue to wrestle with our own bankruptcy laws, so we should be patient with any progress emerging market countries make in this area.
In the meantime, Barry Eichengreen has been foremost among thinkers suggesting that changes in bond contracts would accelerate the restructuring of debt following major crises. He notes that virtually all bond clauses now require unanimous agreement among creditors if the bond payments are restructured. Furthermore, these clauses do not require those creditors who nonetheless get some repayment to share their good fortune with other creditors. These aspects of bond contracts inhibit the orderly resolution of troubled debt. In particular, the unanimity requirement encourages “holdouts” by minority creditors, who in their desire to obtain the most favorable terms from any restructuring can and often do dramatically slow down the process of reaching agreement.
Accordingly, Eichengreen recommends that both sovereign and private international bond contracts be amended in various ways to address the holdup problem. Among the ideas is to introduce majority voting and sharing, and to establish permanent creditors’ committees to facilitate renegotiation if it is required. All good ideas, but a major problem is in getting some countries to go first by imposing such requirements on their borrowers. A good place to begin is with sovereign debt of the G-7 countries, most notably the United States. In fact, various “G” reports (G-10, G-22 and G-7) have all recommended that creditors’ committees be established in advance to facilitate workouts. But as I understand it, our Treasury department has been reluctant to subject our own bonds to this requirement, possibly for fear of having to pay slightly higher interest rates on our government debt (alth