The IMF's Imprudent Role as Lender of Last Resort

Throughout history, financial collapses have been defining moments for public policy. Crises promote action, embodied in new financial institutions or policy doctrines. The motives that underlie such policies are sometimes short-sighted--driven by short-run pressures rather than long-run principles--and it is easier to enact unwise policy in the midst of crisis than to reverse course after the crisis has passed, after policies become embodied in institutions or statutes.

The responses by the IMF and the U.S. government to the Mexican crisis of 1994-1995 and the recent Asian crises are examples of dangerous short-sightedness. In the wake of those crises, the Clinton Administration is promoting a new doctrine of global financial bailouts, administered through IMF largesse and conditions. If the IMF and U.S. Treasury are permitted to prevail, the efficiency of global capital markets will suffer, and the incidence and severity of financial crises will grow.

The Mexican and Asian collapses follow a pattern dating back to 1982, and are the byproduct of fundamental flaws in the incentives facing bankers in developing countries. Incentives to assume excessive risk result from the unhealthy partnerships between government and business in many countries, which manifest themselves in taxpayer bailouts of insolvent banks. International support for bank bailouts will deepen that unhealthy partnership, and thus make the preexisting problems in these countries even worse. The uses of IMF assistance and the U.S. Treasury Department’s Exchange Stabilization Fund to bail out insolvent emerging market banks and international bank lenders are not only improper (in the sense that these sources of funds were not designed to be used in this way); such assistance and the doctrine that underlies it are a threat to the stability of the world financial system.

The suggestion that the IMF’s capital and facilities should be expanded to permit it to engage in more such activity in the future is troubling. The principal lesson of the recent bailout programs managed by the IMF and the U.S. government (and the longer history of generous domestic bailouts of banks in developing economies) is the vital need for all parties (including host governments, the IMF and the U.S. government) to find a credible way to commit not to sponsor such counterproductive bailouts.

Why Are Bailouts Misguided?

A guiding principle of a well-functioning market economy is that those who undertake risks should either lose or gain according to the outcomes produced by those decisions. The idea that government, or governments acting through the IMF, should absorb losses when risky decisions turn out badly is fundamentally contrary to this guiding principle of a free-market economy. This is, regrettably, precisely what the IMF and the U.S. government are doing. While assistance is often couched as “liquidity” assistance to resolve “balance of payments” problems, in fact assistance is designed to absorb the losses of insolvent banks and their borrowers in developing economies, and to insulate international lenders from the losses that they would otherwise suffer.

What have been the costs of government absorption of financial losses? Three kinds of costs figure prominently: (1) undesirable redistributions of wealth from taxpayers to politically influential oligarchs in developing economies; (2) the promotion of excessive risk taking and inefficient investment; and (3) the undermining of the natural process of deregulation and economic and political reform which global competition would otherwise promote. I will explain each of the three categories of cost in turn.

Bailouts Benefit the Politically Powerful at the Expense of Others

The first undesirable consequence of these bailouts is the massive redistribution of wealth away from taxpayers in emerging economies, and toward the wealthy political cronies who control their countries’ industries and financial institutions, and whose imprudence precipitates financial collapse.

While bailouts entail loans from the IMF and foreign governments at subsidized interest rates to developing country governments, taxpayers in the United States and other developed economies who pay the subsidies associated with these loans are not the biggest losers from the bailouts. The IMF and the U.S. Treasury in most cases are repaid. Loans from the IMF and the U.S. Treasury, however, provide powerful justification for increased taxation to repay the loans. When the crisis has passed, the big winners are the wealthy, politically influential risk takers, and the biggest losers are the taxpayers in countries like Mexico or Indonesia.

Mexico’s financial crisis of 1994-95--often seen as a “success story” by the Clinton Administration--provides a case in point. During the resolution of the Mexican banking collapse, the Mexican government (through its deposit insurance agency) purchased more than $45 billion of bad debts from Mexican banks, half of which are the debts of bank related conglomerates. The government promised (at the time) that it would not absorb those debts permanently, and that it would hold debtors responsible for paying their obligations. So far, it has done virtually nothing to retrieve funds from borrowers liable for the debts (Financial Times 1997: 2). Together with the nonperforming loans remaining in Mexican banks, the total taxpayer exposure to loss from the bailout of insolvent Mexican banks is estimated at 16 percent of GDP. Thus, in addition to bailing out foreign and domestic bankers who had lent funds to Mexican firms prior to the crisis, the likely result of the Mexican bailout will be the transfer of billions of dollars from Mexican taxpayers collectively to the country’s wealthiest and most politically powerful enterprises and individuals. The economic result of these taxes is more than a pure transfer to the rich; taxation has also slowed recovery from the recession.

The bailout and redistribution of wealth in Mexico--like those currently underway in Asia--was blessed by the IMF and the U.S. government. In addition to lending money, the IMF and the U.S. Treasury effectively lend respectability and external political imperatives to tax-and-transfer schemes to benefit the rich.

Some proponents of IMF bailouts argue, however, that by intervening the IMF is able to promote fundamental structural reforms (in particular, reforms to domestic banking systems) that reduce the likelihood of future bailouts. The 1994-95 intervention in Mexico, however, provides contrary evidence. In 1995, I was a member of a World Bank team that provided advice to the Mexican government to assist it in implementing its promised reform of the deposit insurance system, which was part of the package of proposed IMF-U.S. Treasury-World Bank reforms. All deposits are 100 percent insured in Mexico, and this complete insurance has effectively subsidized high-risk bank lending to powerful risk-taking conglomerates. (Insured depositors have little incentive to question the use of their funds, which leaves bankers free to make whatever use of the funds they please.) In November 1995, we presented a detailed plan for reform to the Mexican government that would have introduced a small element of market discipline into the system and thus would have partially removed some of the government subsidies enjoyed by the Mexican banks.

Ostensibly, as a vaguely worded condition for the World Bank to release $500 million in funds to the Mexican government, the government had to agree to consider some version of the reforms we were advocating. After a day of lip service (and the release of the funds), one of the Mexican officials in charge took me aside and expressed his appreciation for my efforts, and then told me that “of course, the banks won’t let us do any of this.” That was not a surprise; why would anyone want to give up a subsidy if they have the political influence to maintain it?

So far, little has been done to introduce market discipline into the Mexican banking system, and there is no reason to believe that anything will be done to limit the current system of subsidizing the risks of the industrial conglomerates and the banks they control. According to unofficial estimates, the overwhelming majority of domestically owned Mexican banks are insolvent. On average, 40 percent of the loans of domestic Mexican banks are not performing (even after having shed large amounts of their non-performing loans via government subsidized purchases).

The Mexican experience has made me suspicious of IMF financial sector “conditionality.” I expect similar results from the current conditions being attached to IMF assistance in Asia, where backsliding on conditions has begun even earlier than in Mexico. It is very hard to undermine the corrupt partnership between powerful industrialistbankers and governments by giving them both money in exchange for promises to reform in the future. It is even harder to do so when those conditions are specified in secret agreements--such secrecy makes it impossible for any outside observer to evaluate the wisdom of the conditions, or gauge a country’s eventual compliance with them, which further weakens the incentives of recipient countries to comply.

Indonesia is now working on its third IMF bailout agreement, after staring down the IMF on its two previous “agreements.” The Economist (1998: 37) noted that “In January, Mr. Suharto promised to dismantle many of the monopolies and cartels which control trade in some products. But he seems to have forgotten to tell the monopolists: the plywood cartel, run by his former golf partner . . . continues to issue instructions to its members. Restructuring of the financial system, awash with red ink, has barely begun.” By mid-March, Suharto had become openly defiant of the IMF. His insistence on a currency board and his announced intention to appoint Mohamad Hasan, “the biggest of the cronies,” as industry and trade minister, make it hard even for the IMF to argue that conditions will be respected (Borsuk 1998: A12).

Korea’s leadership has been much more cooperative with the IMF, but here appearances may be deceiving. As Moon Ilhwan (1998: 54-55) noted in Business Week, President Kim “wants the banks to stop lending to big companies at preferential rates. But just in the past six weeks, banks have provided nearly $1 billion in ‘emergency relief loans’ to sickly chaebol . The loans are offered at far below market interest rates, and this is a distorted distribution of limited resources,’ laments Lee Chae Kwang, head of research at Daiwa Securities in Seoul.” IMF aid is being channeled to the banks, which pass it along with impunity to the conglomerates that own them. Thus in Korea, the IMF has not even been able to prevent the immediate misuse of its funds, much less reform the long-term structure of bankindustry relationships.

That is not to say that IMF conditions of all kinds always fail. The IMF has been somewhat successful in getting countries to change tax or expenditure policies, foreign trade policies, and monetary policies. Banking policy is fundamentally different, however, for two reasons. First, real reform in the banking system takes years to accomplish because it entails new ways of measuring and managing risk, new regulations, and new supervisory procedures. These changes are both politically difficult (because the politically powerful must forego subsidies) and technically challenging. The time horizon necessary to implement successful reform is at least five years (judging from the successful examples of Argentina and Chile, which did so very aggressively and voluntarily). Building effective financial institutions, and reforming the legal and regulatory environment in which they operate, is a protracted and difficult learning process, even when countries have the political will to do so. The horizon of IMF crisis assistance and conditionality (typically two years or so) is simply not suited to achieve true reform in the banking system.

Second, banks are controlled by powerful and concentrated vested interests who are willing to fight hard to maintain their access to subsidized credit and block those reforms. A basic principle of political economy is that powerful minorities (in this case, a handful of conglomerate-controlled banks) generally will be successful in obtaining political favors paid for by fragmented majorities (the average taxpayer).

Thus in practice, crisis countries will always find it easy to promise (but never deliver) true banking reform. Instead, they will tax quickly and deeply, pay back their loans to the IMF, replenish the poker chips of their risk-loving conglomerates, and return to business as usual. Another criticism of IMF conditions in Asia, as put forth by Martin Feldstein (1998), is that they are inappropriately detailed and microeconomic. Feldstein argues that it is inappropriate for an international agency to intrude so deeply into domestic economic policy, especially since its charter provides no mandate to do so. I agree with that view, but I would hasten to add that the lesson is not that the IMF should provide bailouts with fewer conditions. Rather, I would characterize the new tendency of the IMF to intervene too deeply into the structure of borrowing country economies as the natural result of overstepping the agency’s original limits of providing assistance. The new intrusiveness reflects an unwise new goal (providing massive bailouts for insolvent financial institutions and international lenders), a noble hope (to restructure the economies of recipients to prevent future dependence on bailouts), and an unrealistic belief in the ability of such conditions to succeed.

The Expectation of Bailouts Increases the Fragility of the World Financial System

The predictable failure of government to allow losses to fall on risk takers after financial crises not only produces a one-time wealth transfer, but encourages behavior that will lead to a repeat of the same problem in the future, which brings us to the second category of costs resulting from bailouts.

If the risk-taking bankers know that future gains from taking on risk will be private, but losses will be borne by taxpayers (again), that amounts to a government subsidy for risk, which thereby encourages excessive risk taking (the so-called moral-hazard problem).

In the United States, we learned during our Savings and Loan debacle that subsidies for risk taking could lead to large losses from unwise, high-risk investments. The losses to taxpayers from that experience (roughly 3 percent of 1990 GDP), however, were small compared to what has been happening in developing economies over the past 15 years--an era that has seen an unprecedented epidemic of high cost bank insolvency. Studies by the World Bank and the IMF have documented some 90 episodes of severe banking crisis since 1982. In more than 20 of those cases, the bailout costs to developing country governments have exceeded 10 percent of GDP. In roughly half of those cases (including the estimated losses of some of the current Asian-crisis countries) losses have been in the range of 25 percent of their GDP (see Caprio and Klingabiel, 1996a, 1996b; and Lindgren, Garcia, and Saal 1996).

These facts warrant emphasis. This string of enormous losses is unprecedented, and is occurring during a relatively stable period of positive global economic growth. Losses to depositors in the United States during the Great Depression, for example, were comparatively small. National banks (the only banks for which depositor loss data are readily available) issued 47 percent of U.S. bank deposits (FDIC 1940: 66). Losses to national bank depositors from bank failures during the worst four years of the Depression (1930, 1931, 1932, and 1933) together amounted to only 1.9 percent of average U.S. GNP over those years (GNP data are from Economic Report of the President 1968: 218). Assuming similar loss rates to depositors of national and state-chartered banks would imply an estimated loss on all deposits of roughly 4 percent of GDP. Losses during other historical periods of the most severe economic crises--the 1830s, the 1850s, and the 1890s--also pale by comparison to the experience of the last 15 years. Indeed, in many countries historically, severe recessions have not been associated with any significant losses to depositors (Bordo 1985, Calomiris 1993).

What can explain the enormity of loss since 1982? Surely not “shocks” of unprecedented magnitude (like oil price hikes, wars, or global downturns in demand), since such influences have been absent during this period. The explanation for the new epidemic of worldwide banking instability is the roller coaster of risk produced by the choices of banks in developing economies--choices that are the byproduct of government subsidies for risk-taking.

Why are banks behaving so differently now from the way they behaved previously? The answer is simple. Prior to the 1980s, banking systems did not subsidize risk nearly as much as they do now. The wave of partial economic and financial liberalization that swept through the developing world in the 1980s and 1990s has been enormously beneficial in many ways, but it should not be confused with true economic liberalization. While many countries have opened themselves to world trade, have privatized many important sectors of their economies (including their financial sectors), and have moved away from direct governmental control of domestic credit, a key flaw in the new era of liberalization has been an expanded, and unhealthy, partnership between government and private business.

Private business in many developing economies is dominated by oligopolistic conglomerates, often controlled by a small minority of wealthy, politically influential families or corporations. At the center of the unhealthy partnership between government and business in many developing countries is a new kind of bank--what I call the quasi-public bank. Quasi-public banks (typically owned and controlled by conglomerates) are private institutions with an implicit claim to public resources (which pay for their losses). They are a key instrument of domestic economic control for conglomerates, and a key vehicle for the transfer of political patronage from the government to these conglomerates.

Prior to the 1980s, banks in developing countries were often state owned institutions, or private banks subject to strict controls that limited private allocation of credit. That system was highly inefficient and limited private sector access to funds. It was replaced by a “privatized” banking system with a different set of inefficiencies--an unnaturally risky form of bank “privatization” that brought freedom without responsibility. When quasi-public banks (and their parent conglomerates) make profits, they keep them; when they suffer losses, the public pays for them (through bank bailouts). That is a formula for encouraging banks to take on extreme risk.

Quasi-public banks don’t always choose to take extreme risk, however, and that explains why they sometimes can survive successfully for years before imposing such large costs on taxpayers. But they are extremely fragile institutions, and they magnify risk for the rest of the economy, particularly during recessions. These banks turn normal economies into a house of cards--one which collapses in the face of even moderate-sized adverse shocks.

The key to understanding how quasi-public institutions magnify economic risk is to consider how they respond to initial losses produced by adverse shocks to their borrowers. Normal private banks experiencing loan losses tend to reduce their portfolio risk to restore the confidence of their depositors and limit the risk of bank failure. But quasipublic institutions need not concern themselves with the risk of failure, since bank depositors and stockholders are all insured against loss by taxpayers. In the wake of losses, these banks face opposite incentives to channel ever riskier loans to their conglomerates.

Financial crises in these economies tend to go through three stages: ( 1 ) initial losses, followed by purposeful increases in bank lending risk; (2) consequent increases in the probability of devaluation, followed by purposeful increases by banks in their currency risk; and (3) large devaluation, followed by enormous losses to banks and taxpayers. The consistency of this pattern in uncanny. It was first visible in the Chilean collapse of 1982-83 (that experience has been described in detail by the former Chilean minister of finance in de la Cuadra and Valdes 1992). The same pattern was followed in Mexico in 1994-95, and in Thailand, Indonesia, and Korea in the recent crisis.

The sequence is disturbingly predictable. First, as initial losses of borrowers and their banks mount, banks and their borrower-owners increase credit and thus pursue higher-risk resurrection strategies. The initial recessionary shock that hits these economies raises the probability of devaluation (slightly, initially). The resultant increase in bank risk makes devaluation more likely because of the link between expected bank losses and future government money supply increases.

Second, increased risk of banking collapse means a greater chance of a government bailout of bank losses. Since those losses are often in the range of 10 or 20 percent of GDP, the implications of these potential losses for government expenditure and money supply increases make drastic devaluation a real possibility.

The high probability of devaluation provides banks and their borrower-owners a new opportunity for profitable risk taking in the form of currency risk. As the risk of devaluation grows, the interest rate difference between local currency-denominated debt and dollar denominated debt rises (reflecting the expectation of a devaluation). Now banks and their borrower-owners face the choice between domestic-denominated borrowing--which has a high current cost, but no currency risk--and foreign-denominated borrowing--which has a low current cost, but a risk of loss following a devaluation (when the value of hard-currency debts can rise astronomically). Because the banks and their owner-borrowers know they will be bailed out by the government if a devaluation occurs, they prefer to borrow via low-interest rate, dollar-denominated debt, and need not worry about the enormous losses they will suffer from a devaluation.

Third, the more the economy increases its dollar-denominated borrowing, the more likely it will be unable to meet those hard-currency obligations. Thus devaluation becomes more and more likely over time.

That, in short, is how quasi-public banks have turned many developing economies into the riskiest financial systems the world has ever seen. Indeed, several former government officials in these economies have issued what amount to “public confessions” that document exactly this pattern (notably the central bank president of Venezuela, Ruth de Krivoy (1995), and the finance minister of Chile, Sergio de la Cuadra). Consider de la Cuadra’s discussion of the Chilean collapse.

As in many other countries, the adverse macroeconomic consequences of the initial exogenous shocks to the Chilean economy made it politically difficult to impose the necessary discipline on banks. As de la Cuadra and Valdes ( 1992: 75) argue, “The superintendency could not include in its loan classification procedure a truly independent assessment of the exposure of bank debtors to foreign exchange and interest rate risk because such an assessment would have interfered with official macroeconomic policies.”

De la Cuadra and Valdes go on to trace how excessive risk taking by banks and firms, and eventual losses from those risks, produced economic devastation by 1982 and increasingly perverse incentives for lenders. Their discussion warrants recounting in detail:

In 1981 most banks saw their effective capital plummet further as soon as optimistic debtors became less willing to pay when the net worth of their corporations fell. This reluctance reinforced the previous perverse incentives to banks, so that banks became even more willing to assume credit risks derived from exchange rate and interest rate risks.

By 1981 financing decisions by Chilean firms and banks reflected a de facto government guarantee to the private sector for foreign exchange risk. Our analysis has identified the superintendancy’s lack of penalization of credit risk in its loan classification criteria as the channel for the guarantee.

The outcome of this structural contingent subsidy was that many small and medium-sized businesses got deeply into debt in 1981. Debts to banks increased during 1981 from 37.6 percent to 50.4 percent of GDP in response to the rise in real interest rates. . . .

By mid-1982 the fall in GDP was so steep that it took on the character of a depression. In June 1982 the government finally decided to devalue the exchange rate by 14 percent. . . . By the end of 1982 the losses that the devaluations had inflicted on the holders of dollar-denominated debts had created insolvency among firms of all sizes. . . .

The sorry state of most debtors caused delinquent loans to rise from 2.34 percent of loans in December 1981 to 3.83 percent in February 1982 and 6.31 percent in May. Most delinquent loans turned out to be 100 percent losses, so they reduced the net worth of banks. . . .

. . . On July 12, 1982, the central bank decided to allow banks to defer their losses over several years, so it began to buy the banks’ delinquent loan portfolios at face value. The banks, however, had to promise to repurchase the portfolios at face value over time with 100 percent of their profits, so the scheme did not improve bank solvency by itself. It solved a liquidity problem but also set the stage for making good the implicit contingent subsidy that the government had offered to speculators in 1981 (De la Cuadra and Valdes 1992: 79-80).

De la Cuadra and Valdes emphasize that loans to industrial firms that were linked to banks via conglomerates were especially forthcoming from banks as a consequence of the government subsidization of risk. Thus despite its free-market orientation and stated commitment to private discipline in banking, Chile ended up insuring “uninsured” claims on banks, subsidizing high-risk resurrection strategies on the part of its banks, and passing on enormous risk-encouraging credit subsidies to industrial firms with close links to banks.

The Chilean pattern was repeated in Mexico in 1994-95. Initial bank loan losses were aggravated by currency devaluation’s effects on Mexican firms that had undertaken dollar-denominated debts. Furthermore, Mexican banks, like those in Chile in 1982, had bet heavily against devaluation. Despite the fact that Mexican banking regulations prohibited banks from assuming currency risk, as the peso devalued Mexican banks suffered large losses from illegal “structured note” agreements they had entered into with American banks (discussed in Garber 1997).

In the recent Asian crises, Thailand, Korea, and Indonesia also have seen enormous increases in their dollar-denominated debts over the past year--after recessionary shocks and bank losses were widely known. In the case of Korea, as early as the end of 1996, its chaebols had averaged debt-equity ratios of 400 percent. As exports fell in 1997, debt rose even further (Woodall 1998: 6).

By June 1997, foreign bank debt had grown to 45 percent of GDP in Thailand, 35 percent of GDP in Indonesia, and 25 percent of GDP in Korea, most of which was short-term debt. By June 1997, South Korea’s short-term debt was three times its foreign reserves (ibid.: 6). The devaluations of recent months (in combination with the prior pursuit of low-interest dollar-denominated funds) have produced an enormous burden on the taxpayers of these countries who now must repay the dollar-denominated debts at inflated exchange rates.

Ironically, some supporters of international bailouts--notably, Jeffrey Sachs et al. (1995)--see the high dollar-denominated short-term debt burdens of developing market economies as the “cause” of “unwarranted” runs on their currencies which they claim produce financial crises. According to that view, large amounts of short-term foreign debt expose countries to the fickle preferences of foreign speculators. That view is misleading for at least three reasons.

First, the run-up in short-term foreign debt is a symptom of weakness (a characteristic of an economy that cannot attract long-term debt) and indicative of the perverse risk-taking incentives of banks and conglomerates that are willing to absorb massive amounts of foreign currency risk at taxpayers’ expense.

Second, fickle foreign speculators are not the source of devaluation pressure. Developing economies in which government and business are too closely linked tend to suffer two kinds of fundamental problems which underlie devaluation pressures: low productivity growth, and off-balance-sheet fiscal pressures resulting from weak banking systems. Because crony capitalism is highly inefficient, it produces low long-term factor productivity growth, which can threaten the long term maintenance of a fixed exchange rate against the dollar. As Paul Krugman (1994) and Alwyn Young (1995) noted of the “Asian tigers” years ago, their impressive growth resulted from combining large amounts of savings with inexpensive unskilled labor, not high factor productivity growth.

Fiscal pressures are also important. The off-balance-sheet liabilities associated with costly bank bailouts imply the need to monetize government debts. Because these potential costs and their monetary implications are anticipated by markets, they can undermine the credibility of the fixed exchange rate. A common error of many macroeconomic analyses of exchange rate collapses in Asia and Mexico is the tendency to focus only on the official government deficit, ignoring the enormous costs of bank bailouts.

The Sachs view is also wrong to identify foreign funds as the primary sources of balance-of-payments outflows. In many cases (as the IMF’s report on the Mexican crisis made clear) foreigners are not the ones who initiate the run on the currency. Well-informed domestic market participants often are the first to flee once it becomes clear that devaluation is imminent.

To What Extent Are the IMF and the U.S. Government Magnifying Moral Hazard?

My review of the moral-hazard consequences of bailouts over the past 15 years has emphasized that domestic governments have often been the most important source of perverse incentives for their banks. Where does the IMF fit in? The main influences of the IMF and the U.S. government in the 1990s have been to aggravate the problem in two ways: (1) to lend legitimacy to (and thus facilitate) domestic bailouts by providing conditions that call for taxation of the domestic middle class to repay the bridge loans from the IMF and the U.S. government; and (2) to insulate foreign creditors (especially banks) from losses during these crises.

Of the two influences, the second is the more pernicious. Insulating foreign banks from loss (by ensuring that bailout packages also rescue them) removes the incentive for foreign banks to avoid lending to high-risk countries. That aggravates the moral-hazard problem by promoting the flow of dollar-denominated “hot money” during the second and third stages of the financial crises outlined above.

In this regard, consider the contrast between what creditors learned from the Mexican crises of 1982 and 1994. In 1982 (in the wake of a decline in oil prices and the rise in U.S. interest rates) foreign lenders to Mexican firms suffered enormous losses as the peso depreciated to 1/3 of its pre-crisis value, making foreign-denominated debt unsustainable for many firms. For example, the workout of one of the largest Mexican conglomerates, Grupo Alfa, entailed eventual losses to some of its creditors in excess of 50 percent. Painful lessons were learned by some of these creditors. Citibank in particular learned important lessons about managing risk on its commercial lending in Mexico, and managed its exchange and credit risk much better during the 1980s and 1990s. During and after the crisis of 1994-1995, its losses were far smaller than in 1982.

What will foreign banks learn from the 1994-95 Mexican crisis, or the recent Asian crisis? I fear they are learning that they can lend without fear of default because of the implicit protection of the IMF and the U.S. Treasury. And it does not help matters that the IMF and the Treasury are signaling their intent to provide future bailouts by calling for ever-increasing amounts of IMF capital and new IMF lending facilities. That, of course, will add fuel to the fire of risktaking in developing economies.

Undermining the Process of Economic Reform in Developing Economies

The cost of insuring foreign lenders against loss runs even deeper, however, which brings me to the third category of costs from bailouts. By insuring foreign creditors who fuel developing economy risk taking, the IMF and U.S. government are undermining the natural process of reform in many emerging economies.

For developing economies true reform is a big step--one that requires the fundamental political transformation from a domestically oriented, rent-seeking society to one willing and able to participate in the competitive global economy. Powerful local oligarchs often can successfully block liberalization if they choose to do so. But the oligarchs may prefer true liberalization if they profit from it. The attraction of participating in the competitive global economy is that globalization offers greater access to foreign markets and foreign sources of capital. That may lead powerful special interests to permit true liberalization if it is the necessary path to globalization. Entrenched oligarchs may choose to liberalize in order to trade a large slice of a small pie for a small slice of a much larger pie.

The incentives for oligarchs to liberalize can be strong if foreign sources of capital are only willing to provide funds to economies with appropriate capitalist infrastructures--that is, those which are based on the rule of law, the protection of creditors and stockholders rights, a predictable means of laying claim to title, an orderly bankruptcy procedure, an intelligible system of accounting principles, a nonconfiscatory tax system, and fair competition in markets.

But IMF and U.S. government assistance can undermine the incentives that encourage the liberalization process. If oligarchs can avoid true liberalization but still maintain access to foreign capital, where is the incentive for them to relinquish the rule of man in favor of the rule of law, or to allow competition and democracy to flourish? If foreign investors are protected by the IMF and the U.S. government, foreigners will be less discriminating about where they place their funds, and thus provide less of an incentive for reform in developing economies.

Thus, bailouts undermine the natural process of reform that global competition would otherwise promote. They do so not only by taxing (and thus weakening) the emerging middle classes in developing countries (the segment of society most likely to push for real reform), but by undermining the incentives of the existing oligarchs to permit liberalization.

From some quarters one hears praise for IMF bailouts as a means to political “stability” in developing economies. If the pursuit of stability means tilting the balance to preserve corrupt rulers and undermine democratic forces within developing economies, it becomes harder to defend policies solely on the basis of the political stability that accompanies them. I for one am very thankful that stability was not the overriding objective of Americans in 1776.

Distinguishing Liquidity Crises from Solvency Crises

Supporters of the IMF sometimes refer to its assistance as an infusion of “liquidity.” While liquidity has nothing to do with financing bank bailouts, liquidity assistance was the motive that gave rise to the IMF as part of the Bretton Woods system. Traditionally (under the pre-1973 Bretton Woods system) IMF intervention was supposed to help bolster central bank reserves to preserve a fundamentally sound exchange rate regime buffeted by “destabilizing speculation.”

After the collapse of the Bretton Woods system and throughout the 1980s, the IMF’s role changed. It assumed the role of helping mainly developing countries devalue in an orderly way, and establish credibility in private markets. The IMF offered technical advice, and monitored compliance with macroeconomic policy objectives. During that period, one could argue that the IMF provided “liquidity” assistance (rather than simply wealth transfers) in the sense that its policies sometimes helped to restore credibility by reversing adverse trends in the fundamental macroeconomic determinants that drove exchange rate depreciation.

In the 1990s, the IMF has stretched the notion of “liquidity” assistance beyond any reasonable definition. IMF programs in Mexico and Asia are now microeconomic bailouts that restore the solvency of clearly insolvent financial institutions. That objective has nothing to do with bank or government liquidity, or with temporary imbalances in the balance of payments.

These bank bailouts also have nothing to do with “panic prevention.” In particular, there is no connection between current IMF programs and the historical interventions by central banks or private coalitions of banks to stem banking crises. (The history and theory of banking panics, and the proper role of the lender of last resort, are reviewed in Gorton 1985, Bordo 1990, Kaufman 1991, 1994, Calomiris 1990, 1993, 1994, 1997, Calomiris and Gorton 1991, Calomiris and Schweikart 1991, and Calomiris and Mason 1997.) The current IMF bailout policies are bridge loans in support of the large wealth transfers from domestic taxpayers to recapitalize clearly insolvent financial institutions and related parties. Historical lender-of-last-resort assistance during banking panics, in contrast, was geared to prevent the failure of solvent banks which were temporarily in need of cash to prevent their unwarranted failure.

Given some of the recent concerns of a threat from “irrational financial contagion” voiced by policy makers in the popular press, it is worth emphasizing that the literature on the history and theory of banking panics (cited above) demonstrates that panics have been “rational” phenomena. Bank panics result from reasonable concerns on the part of bank depositors, and are predictable historical phenomena. Random, irrational attacks on financial systems are not evident in financial history. Thus concerns of “irrational contagion” spreading from one country to another without any fundamental explanatory link connecting the countries are unwarranted. Such concerns should not be used to justify financial bailouts. For example, there are clear fundamental economic connections (notably export product competition) that have produced “spillover” effects across countries within Asia during the recent crisis. As during the Mexican crisis, not all countries suffer from the fallout; the spillovers can be traced to economic and financial linkages, not irrational contagion.

IMF bailouts cannot be justified by panic prevention, as that term is properly defined. Nor could the IMF serve as an effective lender of last resort to the banking systems of developing economies. A lender of last resort (whether private or public) must be in the position to observe and control the uses of the funds it provides. Historically, bank clearinghouse coalitions or central banks have been the lenders of last resort.

There is no reason to believe that legitimate lender of last resort protection to stem financial panics would be best achieved via IMF or U.S. government intervention. Runs on banks are either the consequence of fears of impending devaluation (which central banks control via monetary policy), or the consequence of confusion about default risks within the banking system. In both cases, local authorities are the proper institutions to deal with the problem (by resolving the exchange rate uncertainty in the former case, or by deciding on the appropriate lender-of-last-resort policy in the latter case).

In cases where lender-of-last-resort assistance is warranted, the local central bank (unlike the IMF) has the information and legal authority to enforce the necessary conditions on the behavior of banks receiving such lending. Furthermore, those conditions may involve long-run reforms of banking practices. As I argued before, the brief time horizon of IMF involvement makes any attempt by the IMF to achieve meaningful reform of the financial sector, as a condition for assistance, virtually impossible. Financial sector reform is a process that requires many years to design and implement. Countries that have achieved successful banking reform have done so over many years and as the result of a strong domestic commitment to improve banks’ incentives, not in response to IMF conditions (Calomiris 1997).

The Current Asian Crisis and the Proposed Increase in IMF Capital

If there is no respectable intellectual justification for the current direction of IMF policy (as illustrated in Mexico and Asia), then why are so many people coming out in support of expanding the IMF’s capital and lending facilities? If IMF-sponsored bailouts are weakening democracy, strengthening corruption, aggravating inequality and poverty, and fostering systemic financial instability and industrial inefficiency, why are they so popular?

I think part of the answer lies in the short-run fears of American banks and businesses, which have led them to equate support for expansion of the IMF with support for its programs in Asia. Many U.S. banks and businesses would stand to lose if the current IMF Asian bailouts were undermined. For many, the “long-run” appropriateness of IMF policy is not the issue; their current exposure in Asia is their overarching concern.

Of course, deciding not to expand the IMF’s capital would not in any way undermine the IMF’s existing commitments in Asia. Rather, it would only limit the IMF’s ability to expand such commitments in the future, in Asia and elsewhere. Thus, I think much of the support demonstrated for the IMF on the part of U.S. banks and businesses is not only myopic, but misguided. There is no immediate threat to Asia from limiting the ability of the IMF to expand in the future.

Policy Recommendations

The following are four specific recommendations that follow from my analysis.

First, policymakers should recognize that IMF bailouts like those provided in Mexico and Asia are counterproductive. The IMF can best contribute to global financial stability by committing not to insulate foreign or domestic creditors from loss. The more that developing countries are forced to handle their own financial insolvencies, and the more foreign investors are forced to bear the costs of their investment decisions, the more developing countries will be attracted by the benefits of true liberalization. International “coordination” of assistance to insolvent creditors is counterproductive to the stability and efficiency of the global financial system.

Second, consequently, there is no reason to expand the IMF’s capital or to develop the new proposed lending facility to provide bailouts to financial systems in distress. Indeed, an expansion of IMF capital or facilities would do real harm by signaling an intention to strengthen and expand the IMF’s commitment to provide bailouts in the future. The U.S. government should at the very least try to limit the IMF to its pre-1994 goals of advising countries on their macroeconomic policies (to improve exchange rate stability) and serving as an international delegated monitor charged with tracking those policies and providing credible information to global capital markets. The IMF has more than enough capital to achieve those ends. It currently has $45 billion to allocate, and within three years (after its Asian loans have been repaid), it will have much more.

The IMF and World Bank sometimes have been successful in helping to identify and give credibility to regimes that are honestly pursuing the path of reform. In my view, their expressions of support for these regimes have been more important than the funds they have contributed in support of those reforms. The IMF does not need to use funds to bribe countries to restore balance to their macroeconomic accounts or proper incentives to their banks. The IMF should place its trust in global competition, which gives the most reliable encouragement to true liberalization. Wise economic policies will be rewarded by prosperity, and by global inflows of “unprotected” capital. In the case of banking reform, bribery is not only unnecessary but ineffectual. The IMF should recognize that it cannot control (and should not try to control) the banking regulations of developing economies.

Third, denying the IMF its desired increases in capital and facilities, and working to restrict its purview, are not enough to stop the trend toward unwise expansion of global bank bailouts. Other means of promoting bailouts must also be forsworn. Along with refusing to expand the IMF’s capital, Congress should abolish the Exchange Stabilization Fund--a legacy of the Great Depression which has no legitimate role in U.S. monetary policy today. The Exchange Stabilization Fund--originally created to “stabilize the exchange value of the dollar” (Schwartz 1997: 135)--was the source of a $12 billion loan to Mexico in January 1995. No one could plausibly argue that the loan to Mexico was a form of exchange intervention in support of the dollar. This was not the first time the Exchange Stabilization Fund was used inappropriately. Indeed, as Anna Schwartz (1997) documents, the history of the Exchange Stabilization Fund--contrary to its stated purpose--is rife with similar examples of abuse by previous administrations.

The World Bank--which has been partially successful in providing advice and support for long-term financial sector reform--should also be prevented from serving as a substitute vehicle for bailouts. Whatever assistance the World Bank provides should be limited to gradual support promoting long-run reform of the financial sector. For that purpose, sudden large flows of credit subsidies are unnecessary and counterproductive. Subsidizing privatization of banks requires only small annual flows of credit. Such funds should be distributed only in response to credible government reforms--that is, only after reforms have been initiated, not before.

Fourth, IMF secrecy is contrary to its proper role as a source of independent, objective, and informed opinion about the economic performance and financial risks of member countries. In pursuit of its appropriate mission, any policies or conditions for assistance advocated by the IMF should be revealed publicly. That will encourage a lively debate about their merits, and permit critical evaluation of their effectiveness.


Bordo, M. D. (1985) “The Impact and International Transmission of Financial Crises: Some Historical Evidence, 1870-1933.” Revista di Storia Economica 2: 41-78.

Bordo, M. D. (1990) “The Lender of Last Resort: Alternative Views and Historical Experience.” Federal Reserve Bank of Richmond Economic Review (January/February): 18-29.

Borsuk, R. (1998) “Likely Suharto Nominee Called Rebuff to IMF.” Wall Street Journal, 12 March: A12.

Calomiris, C. W. (1990) “Is Deposit Insurance Necessary? A Historical Perspective.” Journal of Economic History 50 (June): 283-95.

Calomiris, C. W. (1993) “Regulation, Industrial Structure, and Instability in U.S. Banking: An Historical Perspective.” In M. Klausner and L. J. White (eds.) Structural Change in Banking, 19-116. Homewood, Ill.: Business One-Irwin.

Calomiris, C.W. (1994) “Is the Discount Window Necessary?” Federal Reserve Bank of St. Louis Economic Review (May/June): 31-55.

Calomiris, C.W. (1997) The Postmodern Bank Safety Net: Lessons from Developed and Developing Economies. Washington, D.C.: American Enterprise Institute.

Calomiris, C. W., and Gorton, G. (1991) “The Origins of Banking Panics: Models, Facts, and Bank Regulation.” In R. G. Hubbard (ed.) Financial Markets and Financial Crises, 109-73. Chicago: University of Chicago Press.

Calomiris, C. W., and Schweikart, L. (1991) “The Panic of 1857: Origins, Transmission, and Containment.” Journal of Economic History 51 (December): 807-34.

Calomiris, C. W., and Mason, J. R. (1997) “Contagion and Bank Failures During the Great Depression: The June 1932 Chicago Banking Panic.” American Economic Review 87 (December): 863-83.

Caprio, G., and Klingabiel, D. (1996a). “Bank Insolvency: Bad Luck, Bad Policy, or Bad Banking?” In M. Bruno and B. Pleskovic (eds.) Annual World Bank Conference on Development Economics, 1996. Washington, D.C.: World Bank.

Caprio, G., and Klingabiel, D. (1996b) “Bank Insolvency: Cross-Country Experience.” World Bank Policy Research Working Paper, No. 1620. Washington, D.C., July.

De la Cuadra, S., and Valdes, S. (1992) “Myths and Facts about Financial Liberalization in Chile: 1974-1983.” In P. Brock (ed.) If Texas Were Chile: A Primer on Banking Reform, 11-101. San Francisco: ICS Press. Economic Report of the President (1968) Washington, D.C.: Government Printing Office.

The Economist (1998) “Once Again, Indonesia Starts Living Dangerously.” 21 February: 37-38.

Federal Deposit Insurance Corporation (1940) Annual Report. Washington, D.C.: Government Printing Office.

Feldstein, M. (1998) “Refocusing the IMF.” Foreign Affairs (March/April): 20-33.

Financial Times (1997) “Debt Burden Drags on Economy.” 16 December: 2.

Garber, P. (1997) “Managing Risks to Financial Markets from Volatile Capital Flows: The Role of Prudential Regulation.” Brown University Working Paper. Providence, R.I.

Gorton, G. (1985) “Clearing Houses and the Origin of Central Banking in the U.S.” Journal of Economic History (June): 277-83.

Ihlwan, M. (1998) “Kim’s War on Two Fronts.” Business Week, 2 March: 54-55.

Kaufman, G. (1991) “Lender of Last Resort, Too Large to Fail, and DepositInsurance Reform.” In J. R. Barth and R. D. Brumbaugh Jr. (eds.) The Reform of Federal Deposit Insurance: Disciplining the Government and Protecting the Taxpayer, 246-58. New York: Harper Business.

Kaufman, G. ( 1994) “Bank Contagion: A Review of the Theory and Evidence.” Journal of Financial Services Research 8 (April): 123-50.

De Krivoy, R. (1995) “Lessons from Financial Crises: Evidence from Venezuela.” In Proceedings of the 31st Annual Conference on Bank Structure and Competition. Chicago: Federal Reserve Bank of Chicago.

Krugman, P. (1994) “The Myth of Asia’s Miracle.” Foreign Affairs (November/December): 62-78.

Lindgren, C.J., Garcia, G., and Saal, M.I. (1996) Bank Soundness and Macroeconomic Policy. Washington, D.C.: International Monetary Fund.

Sachs, J., Tornell, A., and Velasco, A. (1995) “Lessons from Mexico.” Mimeo. Harvard University, March.

Schwartz, A.J. (1997) “From Obscurity to Notoriety: A Biography of the Exchange Stabilization Fund.” Journal of Money, Credit and Banking 29 (May): 135-53.

Woodall, P. (1998) “Frozen Miracle: A Survey of East Asian Economies.” The Economist, 7 March: 1-18.

Young, A. (1995) “The Tyranny of Numbers: Confronting the Statistical Realities of the East Asian Growth Experience.” Quarterly Journal of Economics. 110 (3) (August): 641-80.

Charles W. Calomiris is the director of the Project on Financial Deregulation at AEI.

Also Visit
AEIdeas Blog The American Magazine
About the Author


Charles W.
  • Charles W. Calomiris is the Henry Kaufman Professor of Financial Institutions at Columbia Business School. He is also a research associate at the National Bureau of Economic Research, a member of the Shadow Financial Regulatory Committee and the Financial Economists Roundtable, and the coordinator of the Bank Performance and the Economy program at the Center for Financial Research at the Federal Deposit Insurance Corporation. Until 2007, he was the co-director of AEI's Financial Deregulation Project. His research at AEI spanned several areas, from banking and corporate finance to financial history and monetary economics. Calomiris also served on the 2000 International Financial Institution Advisory Commission. Known as the Meltzer Commission, this congressionally mandated group recommended specific reforms of the International Monetary Fund, the World Bank, the regional development banks and the World Trade Organization to the U.S. government.
  • Phone: 2128548748

What's new on AEI

image The Census Bureau and Obamacare: Dumb decision? Yes. Conspiracy? No.
image A 'three-state solution' for Middle East peace
image Give the CBO long-range tools
image The coming collapse of India's communists
AEI on Facebook
Events Calendar
  • 21
  • 22
  • 23
  • 24
  • 25
Wednesday, April 23, 2014 | 12:00 p.m. – 1:30 p.m.
Graduation day: How dads’ involvement impacts higher education success

Join a diverse group of panelists — including sociologists, education experts, and students — for a discussion of how public policy and culture can help families lay a firmer foundation for their children’s educational success, and of how the effects of paternal involvement vary by socioeconomic background.

Thursday, April 24, 2014 | 12:00 p.m. – 1:30 p.m.
Getting it right: A better strategy to defeat al Qaeda

This event will coincide with the release of a new report by AEI’s Mary Habeck, which analyzes why current national security policy is failing to stop the advancement of al Qaeda and its affiliates and what the US can do to develop a successful strategy to defeat this enemy.

Friday, April 25, 2014 | 9:15 a.m. – 1:15 p.m.
Obamacare’s rocky start and uncertain future

During this event, experts with many different views on the ACA will offer their predictions for the future.   

No events scheduled this day.
No events scheduled this day.
No events scheduled this day.
No events scheduled today.
No events scheduled this day.