Leseprobe
Inhaltsverzeichnis
1. Introduction
2. Understanding Currency Crises
2.1 The Path from the Canonical Model to Moral Hazard
2.2 Multiple-Equilibria and Financial Panic
3. At the Core of the Problem: The Exchange-Rate Regime
3.1 Currency Pegs versus Floating Exchange Rates
3.2 Dollarization as an Alternative
4. At the Center of the Problem: Liquidity
4.1 The Need for Liquidity or an International Lender of Last Resort
4.2 Insurance and Credit Facilities against Systematic Risk
4.3 Capital Controls
4.3.1 Controls on Capital Outflow
4.3.2 Taxes on Capital Inflows
5. At the Heart of the Problem: The Institutional Framework
5.1 International Bankruptcy Standards
5.2 Regulation and Transparency
5.3 Collective-Action Clauses in Financial Contracts
5.4 Importance of Foreign Ownership
6. Different Currency Crises and Remedies
7. Summary and Conclusions
1. Introduction
The last decade has witnessed a number of sharp currency depreciations, albeit the official promise never to let this happen. The trouble started in 1992 when a speculative attack on the British pound and other European currencies forced the United Kingdom, Spain, and Italy to leave the European Exchange-Rate Mechanism (ERM); France stayed only after widening exchange rate bands.1 This currency crisis differs greatly from those which followed. First, it occurred in mature industrial economies. Access to capital markets remained open during the crisis and, hence, it would have been possible to defend the exchange rate bands if governments would have chosen to do so. Second, it seems that there was no economic penalty for leaving the ERM. In fact, England and Spain experienced higher growth rates than the European average.
The subsequent crises took place in emerging markets, starting in 1994 in Mexico. Slow growth and a deteriorating political situation, for example the rebellion in Chiapas, made it impossible for Mexico to roll over its dollar-denominated short-term debt (tesobonos). A small devaluation of the peso led to a complete loss of confidence and the peso dropped to half its pre-crisis value. A rescue package by the International Monetary Fund (IMF) and the USA bailed out the Mexican government but at a price: GDP fell in the following year by 7 percent. Economic growth was regained in 1996.2
When Thailand devalued its currency, the Thai baht, on July 2nd 1997, the stage was set for the most severe and virulent currency crisis of the 90s. The value of liabilities, denominated in foreign currencies, jumped up in face of this widely unpredicted depreciation, leading to a financial crisis in the corporate sector and in the banking system. Firms, unable to meet their obligations, added to the number of bad loans in the banking sector, which itself struggled in serving its foreign debt. A resulting credit crunch precipitated a vicious circle, spreading from Thailand to Malaysia, the Philippines, South Korea, and Indonesia. The pure figures of contracting GDP, ranging from -0.6 percent in the Philippines up to -15 percent for Indonesia in 1998,3 draw only a mild picture of the magnitude. The situation for the poor was very severe. More than 13m people in the Asian region lost their jobs because of the crisis and most of them fell back into poverty.4 Primary impacts, such as unemployment and increasing poverty, adversely affect family cohesion and education expenditures, eroding social capital for future development.5 Civil unrest even triggered some open conflicts against minorities or immigrants, like in Indonesia, after the cushion of economic welfare broke away.6 Shockwaves of the Asian crisis were felt in almost every emerging economy in the form of increased interest-rate spreads or speculative attacks on currencies, that were perceived to be vulnerable in a similar manner.
In August 1998 the Russian government decided unilaterally to renegotiate payments on its foreign debt and allowed the ruble to devalue. The following reassessment of risk in emerging markets led again to widening interest-rate spreads and to a drain of liquidity in those economies.7 After months of defending its currency peg, Brazil had to devalue its real in the beginning of 1999, but with only short-lived adverse effects elsewhere in the region.8
The currency crises in emerging markets did not only correct a financial overhang, but incurred real economic losses in the affected countries, most obviously in those, who had to give in to a devaluation and experienced recession in the aftermath. But reassessment of risk for emerging markets in general put a number of other emerging countries under pressure which subsequently suffered from raising interest-rate spreads and more cautious lending procedures. Most prominent examples are Argentina and Hong Kong SAR, which successfully defended their currency boards, but at the cost of hiking interest rates.9 Table 1 offers an overview of the real output loss in countries affected by speculative pressures in the 90s, independent of their success in defending their exchange-rate peg. Preventive packages and rescue loans by the International Monetary Fund (IMF) intimated, that only enduring a recession might restore market’s faith in the exchange-rate commitment of economies in trouble.10 The near-collapse of the hedge fund, Long-Term Capital Management, in 1998 showed that even mature markets were not immune against shockwaves of the recent crises. Hence, international bodies, governments, financial tycoons, academics, and others have come up with an extensive number of proposals on how to make the financial realm a safer place.
Table 1: Currency Pressure and Real GDP Change
Abbildung in dieser Leseprobe nicht enthalten
Source: Sachs (1998, p. 11)
In this paper, I will concentrate the perspective on emerging markets. For simplicity reasons the labels emerging market or developing country will be treated as a synonym in the following. Capital account liberalization during the 90s often flushed enormous amounts of capital into the emerging economies, whereas capital market development was neglected. Restated, the build-up of large foreign debt coincided with weak banking regulation and supervision. The surge in foreign financing contributed to the so-called Asian Miracle and economic progress elsewhere. Hence, the recipients benefited from them before a sudden reversals of capital flows triggered the currency crises.11
In principle, international allocation of capital improves welfare. However, financial markets suffer from asymmetric distributed information and thus from adverse selection. The value of financial assets depends not only on economic fundamentals, but also on the behavior of other players in the market. Runs on banks or currencies can happen whenever confidence is lost. In mature markets, regulation and supervision as well as self-imposed codes of conduct, tame excessive behavior of participants. A reliable monetary policy of the central bank contributes to this fact. In emerging markets the government itself is a source of uncertainty. They lack the reputation to pursue credibly a certain monetary policy, particularly in times of economic or political distress. And with protective market institutions missing, a collapse can hardly be avoided. Establishing appropriate institutions and creating a reputation of sound economic policies should therefore be a strong priority for emerging markets. Brushing these problems away by arguing that proper regulation and enforcement would have prevented recent crises is not enough, as it would not help today’s already liberalized economies, although, those countries that deregulated more prudently, such as India or Chile, fared much better in recent currency turmoils. Proposals to improve the financial landscape should therefore aim to give emerging countries in need the possibilities and the breathing space to develop appropriate institutions to regulate and supervise their capital markets.
A great variety of recommendations has evolved hitherto, many of them contradictionary or mutually incompatible.12 They can be distinguished after those which can be followed by emerging countries unilaterally and those which need a multinational consensus. The latter are more ambitious and refer to things like global standard setting and establishing supranational organizations like a true lender of last resort, but are less likely to be quickly implemented. The former, like generating more liquidity or switching the exchange-rate regime, can be done by developing countries individually and thus on short-notice. If at all, they can only improve the status quo and rule out moderest confidence crises. There may always be - by accident or misbehavior - pressures that surmount newly erected protection barriers. Hence, respective improvements will make currency crises less probable, but cannot prevent a big bang.
With regard to the sequences of a currency crisis, three different starting points for a reformed financial landscape can be distinguished. Predicting currency crises is crucial if appropriate countermeasures shall be used in time to prevent it from becoming worse. But recent attempts to do so have been futile. Crisis prediction suffers from a basic obstacle. When a crisis becomes predictable, a speculative attack will be launched immediately, usually too fast for far-reaching policy decisions. Thus it will be hard to find any reliable leading indicators.13 Crisis prevention addresses proposals that aim to make currency collapses less likely to happen and represent the majority in the reform discussion. In the wake of a financial meltdown, procedures to deal with the inevitable become necessary. Crisis management is therefore the third approach in reforming the international financial architecture. Proper measures to manage crises may also discourage financial pressure to convert into a speculative attack and, thus, contribute to crisis prevention.14
A caveat is, that good workout-procedures in the face of currency or financial crises can prove to be counterproductive. If borrowers can walk away with their debt too easily, investors lose an effective disciplining measure and might react by raising borrowing costs or credit rationing, that is, reducing the overall engagement in emerging markets. For example, improving domestic bankruptcy procedures is highly sensible and vitally necessary in most developing countries. But the first one doing so will be confronted with investors who fear that the only reason why the respective country has eased up debt rescheduling and alike is, because it will or must reschedule its debt the next time. They will regard it as a bad sign and flee the market. Hence, there is a trade-off between a facilitated restructuring process and its adverse signaling effect.15
Crisis prediction, prevention, and management are strongly connected with each other. By way of, controls on capital outflows are a way of dealing with crises. To allow them to work properly, accurate timing of introduction is crucial, giving rise to the need for reliable prediction of currency crises. Under ideal conditions, capital controls can prevent a grab race on a country’s domestic assets, thereby preventing the crisis.
Outline of the Paper of the Paper
Attempts to explain currency crises have evolved since the 1979 canonical model by Paul Krugman. Subsequent approaches became more sophisticated and tried to cope with the new aspects brought to light by recent financial collapses. In part 2 of this paper, I will simplify these approaches down to two major flaws. In 2.1, the path from the basic Krugman model to its modern application with moral hazard is characterized. Overinvestment in suspicious domestic projects, enabled and protected by crony capitalism, are thus key to the understanding of currency crises. The subject of moral hazard due to fixed exchange rates is again discussed in part 3.1.
Part 2.2 addresses the problem of financial panic. Small events can have large consequences, for example the shift from a stable equilibrium to an equilibrium of huge and sudden capital outflows. Information cascades, herding behavior of investors, and an insufficient banking system in emerging markets are some of the phenomena identified as being responsible for financial panic.
For reason to be explained in the mentioned parts, this paper is based on the assumption that, in principle, the two mechanism contribute to currency crises at the same time and, hence, successful reform proposals must address both. This can be exemplified by the traditional value-at-risk (VAR) modeling of the risk management systems in financial institutions. In short, VAR models are used to choose those investments, that will generate with a high probability, say 95 percent, a profit or at least no unacceptable loss that could endanger the entire bank. The residual probability represents market failure with severe losses but is often disregarded when assessing investment projects.16 Based on economic fundamentals, countries can be more or less vulnerable to financial panic - the higher the vulnerability, the higher the probability of market failure. But moral hazard can offset this mechanism. Bailout expectation can replace sophisticated work on market risk, making investors regularly underestimate the latter. In addition, standard VAR models tend to contribute to financial panic themselves. They regularly neglect the comparative history when assessing a country’s financial vulnerabilities because they often concentrate on the immediate past of 24 to 36 month.17 Indeed in the wake of a crisis, their risk indicators turn sharply upwards, forcing firms by their own methodology to retreat from the exposed positions, even if they believe that prices have been overshot.18
With financial panic and moral hazard being identified as the core ingredients of a currency crisis, part 3 addresses a more subtle origin of financial distress. As a matter of fact, most of today’s emerging markets are unable to borrow in domestic currency abroad, or long-term at home. This so- called “original sin” hypothesis highlights the need for prudential sequencing in financial liberalization if these adverse effects are to be avoided. For those which apparently deregulated too fast, and are now unwilling or, more likely, are unable to reverse this trend, the issue of dollarization enters the discussion.19
Parts 3,4, and 5 review the most influential reform proposals of the financial landscape in the current discussion. The starting point is the exchange-rate regime in part 3. Because an explicitly or implicitly pegged currency lures domestic and foreign investors to take on the risks of a currency mismatch, limited liquidity in foreign funds during times of distress may trigger creditors’ grab race, as only the first demanding repayment of their claims will be served, whereas the late ones receive nothing in the worst case. Hence the problem of liquidity, its costs, and possible substitutes are discussed in part 4. Institutional factors have been made responsible for the vulnerability of many emerging markets to currency crises. Weak regulation and supervision as well as an intransparent banking system foster information asymmetries and can lead to information cascades, when a minor event reveals or intimates that investor’s expectations in general have to be changed. Part 5 therefore describes approaches for improving the institutional framework, including those which facilitate a necessary restructuring process after the outbreak of the crisis.
The distinction according exchange-rate regime, liquidity, and institutional framework is of course somewhat artificial. Other approaches, some of which have been presented in the introduction, would be possible. But this way sheds light on the reform proposals with regard to the respective underlying fields of problem.
In part 6, the recommendations so far will be applied to the currency crises of the 90s. The examples described there are no full grown case or country studies. But in a simplified presentation, the practicality and usefulness of the given strategies is tested. In table 6, the different causes of crises and the variety of remedies are summarized. Concluding remarks can be found in part 7.
2. Understanding Currency Crises
2.1 The Path from the Canonical Model to Moral Hazard
Currency crises refer to a sudden depreciation of the exchange rate of more than 10 percent.20 Economic models that describe how a proper currency crisis has to work have evolved over several generations. Since 1979 so-called first generation models explain how a budgetary deficit, financed by monetary policy, can lead to a speculative attack on the hitherto pegged currency when foreign reserves are limited. A speculative attack is a process were investors abruptly change the composition of their portfolios in favor of foreign currencies and against the domestic money, thereby, trying to acquire the remaining stock of governmental reserves.21
The approach is rather deterministic. The expansionary monetary policy is not in line with peoples’ preferences, hence, they will reduce their amount of money by increasing their demand for goods and services, that is increasing the absorption. Foreign reserves will then be depleted in either of two ways. First, domestic prices might rise due to the increased demand. The interest rate will not be allowed to raise, as this would offset the growth in money supply intended by politicians. Hence, inflation will reduce the domestic rate of return on invested capital giving an incentive to transfer the money to higher yielding areas abroad. Second, the increased money supply might directly raise the demand for foreign goods and services and the resulting trade deficit leads to an outflow of foreign reserves.
By assumption, the level of foreign reserves is limited and cannot be raised by foreign lending or alike. Thus the steady net outflow of foreign currency must ultimately deplete the reserves at some point in the future. Then the central bank will have no choice but to let the currency devalue or float. International investors anticipate this mechanism and will withdraw their funds as early as the path to devaluation is stepped on. As soon as it becomes evident that the current monetary policy is not compatible with the exchange-rate peg in the long run, a speculative attack on the currency will be triggered, depleting the central bank’s reserves immediately. The original idea stems from the 1979 Krugman model, whereas Flood and Garber (1984) added a more precise timing of the speculative attack.22
In the subsequent second generation models, this deterministic mechanism has been replaced by a discretionary decision of a country’s government, based on the trade off between the costs and benefits of keeping the exchange-rate peg. A long lasting commitment to a fixed currency will enhance the authority’s credibility to maintain this regime in the future, thus making price stability policies less expensive. The costs derive mainly from adverse effects of hiking interest rates on the domestic economy. Defending the currency with the appropriate measures will tighten the money supply, thereby raising the interest rate. Alternatively, fiscal austerity could be used to reduce absorption in order to improve the balance of payment. Both, higher interest rates or a reduction in public spending might damage the domestic economy and may lead to recession.23
The magnitude of the damage depends on the economy’s capabilities to adapt to adverse shocks. A high debt burden of the fiscal authority will make any increase in interest rates painful. If loan contracts have a short maturity time, the loss of demand will be precipitated, because liabilities adjust quickly to the higher interest rates. Private corporations and financial intermediaries will face trouble if investments are financed under a maturity mismatch, that is, long running projects financed with short-term debt. When revolving the outstanding liabilities, firms have to accept the raised interest rates, perhaps driving the whole undertaking into unprofitability. Defaulting firms will add to the number of non-performing loans further endangering the banking sector.24
Inflexible labor markets let the number of unemployed surge, whenever demand is reduced as a consequence of adverse shocks. Hence, structural factors determine how vulnerable an economy is towards changes in interest rates and thereby how much this would add to the unemployment rate. The impact of the decision to withstand or to give in to pressures depends on the government’s assessment of unemployment. If the unemployment rate is already high and perceived as a problem of society with public responsibility, politicians might not be eager to carry the additional burden.
In second generation models, international investors cannot rely on the mechanical loss of foreign reserves which will inevitably bring down the exchange-rate peg, but have to anticipate government’s valuation of magnitude and severity of the consequences to defend the currency. This explains why even countries with a currency board got attacked. It is not possible to force a depreciation by withdrawing foreign funds because the domestic money base is, by definition, backed with foreign currencies. But investors might think that recession and unemployment, caused by a tightened money supply, is a price too high to pay for maintaining this exchange-rate regime. Speculative attacks in 1998 on Hong Kong SAR proved to be unsuccessful in the end, but showed the market’s loss of confidence in the currency board.25
Some conditions of a domestic economy significantly increase the likelihood of speculative attacks. Theses include the state of the financial system, public deficit, growth prospects, unemployment, and a number of less-easily measured factors, such as the political support for the government. Hence, self-fulfilling prophecies and multiple equilibria become possible.26
The Asian crisis of 1997 came to a surprise to many representatives of the mentioned models. The economies of the later affected countries, Thailand, Malaysia, Indonesia, South Korea and the Philippines (Asian-5), were in general thought to be in good shape. Official data showed public debt to be sound, the prospects of economic growth were fine, and the Asian governance style was perceived to be a different, maybe superior way of doing business. The shiny picture of the East Asian Miracle27 covered concern about growing trade deficits that were steadily accumulated in the region during the 90s. The situation was considerably heterogeneous among the Asian-5, with Thailand being perceived to be the weakest part and, hence, the first country to fall. But even the critics of the Asian economies, among others Paul Krugman and The Economist, did not expect a crisis of the magnitude experienced in the aftermath of the Thai devaluation in July 1997.
Subsequent explanations by Corsetti, Pesenti, and Roubini (1998) or Krugman (1998) describe the picture of the Asian economies as much gloomier than perceived before. Domestic firms and banks were equipped with explicit or implicit guarantees by the government in the case of default. Investors did not face the full risks of their projects, as a negative outcome would have to be carried by the taxpayer, whereas the profit would remain at the initiator. Hence, this is a typical example for moral hazard leading to overinvestment in the protected sectors. In exchange for these guarantees, private enterprises and financial intermediaries had to undertake tasks on behalf of the government in the field of economic development but also to the personal profit of politicians. This close relation has gained prominence as crony capitalism. The real-estate bubbles in many East Asian cities can be regarded as an example of overinvestment in apparently safe projects.28
Corsetti, Pesenti, and Roubini distinguish three types of moral hazard. At the corporate level, control was exercised by the government to pursue investments, even if costs exceeded risk adjusted returns. But it also made clear, that it would stand ready to bail out firms in trouble. Financing further domestic investment required increasing foreign funding, disregarding growing trade deficits that had to financed as well. The weak banking system in most emerging markets constituted the financial moral hazard. A quick capital account liberalization and a subsequently lax supervision, but with heavy public intervention, were again backed by bailout promises. The explicit or implicit peg to the US dollar aimed at lowering the spread in dollar- denominated debt, adding to financial vulnerability. The international dimension has been equaled with a huge fraction of foreign short-term and unhedged lending, mostly provided by international banks. A proportion of over fifty percent to total debt was rather normal.29
The advocates of those third generation models argue that the given guarantees, if exercised, added to the public deficit to an extent, that even an originally balanced budget will be shifted immediately to an enormous burden that justifies the speculative attack forecasted by earlier models. In the Asian case, the speculative attack began when it became obvious that it was impossible both to bail out all firms and banks in trouble and maintaining the fixed exchange rate at the same time without causing a severe recession. The first depositors, originally lured into the country by bailout guarantees, who wished to convert their domestic currency back into foreign currency, were served by the country’s central bank with the respective amount out of its reserves. But by assumption, the governments were either unable or unwilling to further guarantee investments, making domestic assets less attractive because of their now increased riskiness. Deteriorating prices and massive capital outflows were the consequence. In an attempt to save the banking system, additional liquidity was provided by the central banks, disregarding the constraints of a pegged currency and, thus, making the devaluation inevitable. A banking crisis and a currency crisis are just two sides of the same coin.30
The moral hazard approach has been challenged in the subsequent time. It may be comprehensible that creditors who hold deposits at a bank, owned by the governing family, expect their exposure to be protected by the state. It becomes less comprehensible when investing into medium-sized private firm with obvious small links to the government, as it was the case in Indonesia.31 During the Asian crisis many international investors had to carry massive losses, undermining the bailout hypothesis. Conservative estimates put crisis related losses since 1997 at 240bn US$ for equity investors, 60bn US$ for international banks, and 50bn US$ for other private foreign creditors.32 However, the Mexican crisis of 1994 often serves as an example for a complete bailout because the IMF and the US Treasury provided enough funds to pay off holders of the nearly defaulting tesobonos. But in the antecedents of the rescue, bond prices fell and those investors, lacking the faith in a bailout carried losses when selling into a falling market.
Moral Hazard Revisited
The realization of a bailout is not crucial for the validity of the moral hazard argument. Investors might have wrongly expected to be rescued by the authorities, however, a history of private bail-ins, at least partly, should have influenced expectations against moral hazard, although the importance is arguably.
According to the moral hazard approach international capital flows can be expected to be excessive. In reality international capital flows are considerably small compared to capital flows within nations. An efficient allocation of capital should lead to similar capital/ labor ratios in rich and poorer countries, at least in the long run. Latin America has actually a capital/ labor ratio three times smaller than that of the United States. With the present capital flows to Latin America of five percent of GDP, it would take centuries to become level with the capital/ labor ratio of the United States.33 Thus it is hard to find evidence for excessive international capital flows which would support the moral hazard hypothesis, at first glance. But it may be that the moral hazard effect is offset by other forces, for example, the original sin hypothesis or institutional deficits which suggests rather small capital flows, especially the predicted reliance on short-term debt has been strong in the Asian and Mexican crises. Short-term liabilities exceeded liquid reserves at the dawn of the crisis.34 The ratio of short-term debt to overall debt rose between 1990 and 1996 in the Philippines by 33 percent (reaching 19.34 percent of total debt in 1996); Thailand by 39 percent (reaching 41.41 percent); Indonesia by 57 percent (reaching 24.98 percent); Korea by 61 percent (reaching 50.20 percent); and Malaysia by 123 percent (reaching 27.83 percent of total debt in 1996).35 But the figures can be misleading. Taiwan and Hong Kong had in 1996 a ratio of 68.44 percent and 43.57 percent, respectively, but were able to defend their currencies due to their huge foreign reserves. It is hard to distinguish whether the excess of short- term debt over liquid reserves is condition or consequence of the crisis.
Another critique of the moral hazard approach can be derived from the basic model. Overinvestment as a consequence of moral hazard occurs when firms take the profit from their projects but leave eventual losses to the taxpayer. Hence, they calculate in a Pangloss world, ignoring the expected returns but regard only the profit in the best of all possible outcomes.36 Using a classic example derived from Milgrom and Roberts, table 1 shows an investment of 100 and the good state and the bad state being equally probable.37
Table 2.1: Moral Hazard and Investment Decisions
Abbildung in dieser Leseprobe nicht enthalten
With a credible bailout promise, the rational investor will choose the risky project to maximize his return. Now consider the more realistic case of a financial intermediary that has not only one investment, but undertakes a number of projects at the same time. As well, assume that the government will guarantee the survival of the bank but not bail out each project individually. With one hundred equal projects similar to the investment described in table 1, the expected profit would be 700 for the safe investment, but zero for the risky undertakings because the bad results drive out the good. Under the presence of a survival guarantee, it would not be sensible for the bank’s CEO to approve the risky investments, as he would be better off on the safe path. The picture changes if the bank is in such a shaky state that it would face bankruptcy even with a profit of 700. Things cannot get worse because of the bailout guarantee of the government. But the bank may gamble on redemption. If not every second project goes bust as expected due to the probability distribution but by luck and craft only every third or less, then the return of the risky investments would exceed the safe one’s and might rescue the bank. From the bank manager’s point of view this may be worth a try as there is nothing to lose and nothing to gain otherwise. Hence, the presence of bailout guarantees does not justify a priori moral hazard on the corporate and financial level, but, in the case of economic distress, those guarantees will aggravate moral hazard as they give an incentive to gamble on redemption rather than an orderly workout of debt. On the international level, the basic moral hazard assumption holds. International investors lend money or buy assets of a number of different firms and banks in emerging markets. Due to government’s survival guarantee, all projects are covered individually, leading to an investment decision under moral hazard as presented in table 1.
Following the moral hazard explanation, the severity and virulence of recent currency crises must come at least partly to a surprise. The Asian crisis of 1997 can again serve as an example. In the aftermath, a number of conditions have been identified that may have caused the meltdown of the currencies. Growth rates remained high during the 90s, figures for 1996 range from five to over eight percent in the Asian-5. But appearances may be deceptive. Other data suggested a gloomier picture of the economic state. Dramatically raising office vacancy rates in central districts of Asian capitals exemplified the bubble in the real-estate sector. Figures for 1997 reached thirty percent in Shanghai and fifteen percent in Bangkok, albeit Malaysia’s Kuala Lumpur and the Philippines’ Manila remained comparatively sound with three and one percent vacancy, respectively, underlining the heterogeneity of Asian conditions. Stock markets declined in general, with the respective price indexes in 1996 dropping sometimes to less than half of their 1993 values, such as in Thailand, and, of course, there was further decline in 1997.38 As a consequence of the unpreceded boom in the United States during the 90s, the US dollar became the strongest performing major currency in the world of that period. The Asian-5 currencies were more or less effectively pegged to the greenback, leading to an appreciation of the real exchange rates prior to the crisis. With 100 being the average in 1990, the Philippines reached a real exchange rate of 116.4 in 1996, 105.4 in Indonesia, 112.1 in Malaysia, 107.6 in Thailand, and 87.2 in Korea, up from 84.7 in 1994. The high exchange rate rendered domestic production less competitive and was paralleled by growing trade deficits of three to eight percent of GDP in 1996 in the Asian-5.39 But capital inflow to the later affected countries remained strong, rising in 1996 again from 62.9 to 72.9bn US dollars.40
Pressure on a downward adjustment of the affected currencies could only be offset by luring foreign capital into the countries to finance the imbalances in the current account. In 1996, foreign debt exceeded GDP by forty percent in four of the Asian-5 countries, Korea being the exception with 28.4 percent. Most of these funds came in the form of short-term capital. The share of short-term debt to total debt was rising almost constantly during the 90s, reaching ratios of twenty to over fifty percent. Hence, short-term liabilities exceeded foreign reserves in all affected Asian countries, except Malaysia and the Philippines, making them extremely vulnerable to a speculative attack, as immediate withdrawals could not have been satisfied by the central bank at the fixed exchange rate.41
The financial level is characterized by a high degree of bank lending compared to mature financial markets like the United States. Banks in developing countries often have comparative advantages in gathering and assessing information about prospective borrowers. They can better assess expected returns than a market for corporate bonds. This asymmetric distribution of information can prove to be disastrous when a bank in trouble has to liquidate its assets. The market will only pay an average price based on incomplete public disclosure, which will be lower than the bank’s estimate of the value based on its private information. Thus the return by selling the assets might not suffice to pay off the liabilities, leading the bank into bankruptcy.
Throughout the 90s, the financial sector in the Asian-5 was poorly regulated and supervised. Together with the always present crony capitalism, the banking industry was in weak state. The rising number of non-performing loans, around ten percent of total lending in the year before the crisis, and the increasing dependency on foreign short-term borrowing, support this view.
Hence, enough fundamental flaws that explain the subsequent crisis of 1997. The question remains, however, why these factors were ignored prior to the Thai devaluation. Respective figures were published regularly but did not induce a more prudent lending behavior on behalf of international investors. The assumption that they were offset by moral hazard is not entirely convincing because the extent of a bailout surmounted government’s capabilities by far, which should have been anticipated by the creditors. Hence rational investors, aware of the rising riskiness of the region, should have reduced their Asian exposure or demanded higher interest rates to compensate, which they obviously did not.
In the period before the crisis international rating agencies such as Moody’s or Standard & Poor’s certified investment grade for all Asian-5 countries except the Philippines, but which was in fact upgraded to Ba1 in May 1997. Only Thailand which has been generally perceived to be the weakest country in the region, experienced a shift to the watch list with a downgrade to A3 in April 1997 (both Moody’s).42 The rating agency obviously failed to assess the economic condition of most Asian countries accurately.
Tremendous growth rates of the past fueled further confidence in the Asian development, covering economic flaws and rising asset price bubbles. Financial deregulation and a subsequent lax supervision made it easy for banks and other financial intermediaries to attract foreign loans by taking on substantial risks in the form of currency and maturity mismatch. International investors were too willing to provide extensive funds while feeling protected by governmental bailout guarantees. That most foreign creditors had to carry losses due to the Asian crisis tells a story. That the governments and the international community were not able or willing to bail out every foreign claim can either be seen as a moral hazard expectation that did not hold in the end or, perhaps, moral hazard might not be the final answer.
2.2 Multiple Equilibria and Financial Panic
Multiple equilibria characterize a situation in which a number of different outcomes can be stable. There exist no binding transfer mechanism for a particular result as in the canonical model of currency crises. Even a minor shock might be sufficient to switch from one equilibrium, for instance of financial stability, to another, for example financial crisis.
Consider a country that has its exchange rate pegged in order to have a nominal anchor to import price stability. High unemployment or a weakened financial sector might tempt policy makers to give up monetary austerity in order to stimulate economic growth and provide liquidity to troubled banks. Resulting capital outflows require the domestic central bank to change local against foreign currency out of its reserves. Following the 1979 basic model, the growth in money supply would endanger the currency peg, as soon as foreign reserves fall under the critical level.43 The government has to balance the benefits of a stimulating monetary policy against the costs of probably losing the currency peg and, hence, reduced long-term credibility in monetary issues. It may now decide that keeping the fixed exchange rate and pursuing a non-inflationary policy is worth more than a quick economic stimulus. Thus the government will declare that changing the exchange-rate regime is out of question and that it would stand ready to defend its currency against all odds.44
If a speculative attack is launched, the costs of maintaining the currency peg will increase because capital flight will tighten the money supply. The drain in foreign liquidity can threaten the banking system as mature loans might not be rolled over. Financial intermediaries react in the form of a credit crunch, thereby adding to the liquidity squeeze in the corporate sector. Economic growth might slow down, raising the unemployment rate and putting pressure on the political system. In reaction to the increased burden, the government might change its mind and let the currency devalue or float.
[...]
1 See Krugman (1997, p. 16).
2 See Krugman (1997, pp. 17-20).
3 See Sachs (1998, table 1).
4 In Indonesia an additional 20 percent of the population were expected to have fallen into poverty in 1998 (World Bank, 1999, p. 3).
5 As exemplified for the case of Thailand (World Bank, 1999, figure 1, p. 17).
6 World Bank (1998, p. 2).
7 See IMF (1999b, p. 1).
8 See IMF (1999b, pp. 81-84).
9 Hong Kong’s currency board came under attack although its exchange-rate commitment was widely believed to be credible. But raising interest rates and a following decline in stock prices gave those investors a profit which were selling short prior to the attacked, therefore giving rise to the concern that the attack was launched only for this kind of speculation, then referred to as double play (IMF, 1999b, p. 173).
10 See Krugman in a Frontline Interview with regard to the IMF’s assistance packages to Brazil before its devaluation.
11 Although the counter-argument exists, that capital inflows led to overinvestment and hindered growth.
12 See Eichengreen (1999a, p. 1).
13 See Eichengreen, Rose, and Wyplosz (1995, pp. 279 - 284).
14 See Eichengreen (1999a, pp. 9 - 18).
15 See Dooley (2000, pp. 6 - 10).
16 See IMF (1998, p. 9).
17 See Edwards (1999b, p. 16).
18 See White (2000, pp. 22 - 23).
19 See Eichengreen and Hausmann (1999, p. 1 - 27).
20 See “East Asian Economies,” The Economist, March 7th 1998
21 See Krugman (1992, p. 62)
22 See Krugman (1992, pp. 61 - 77), and Flood and Garber (1984b, pp. 1 - 14).
23 See Obstfeld (1986, pp. 72 - 81).
24 See Eichengreen (1999a, p. 136).
25 See IMF (1999b, pp. 169 - 176).
26 To be explained in the next section.
27 World Bank (1993).
28 See Krugman (1998b, pp. 3 - 9).
29 See Corsetti, Pesenti, and Roubini (1998a, pp. 2 - 4).
30 See Eichengreen (1999a, p. 139).
31 See Radelet and Sachs (1998, p. 6).
32 See Haldane (1999, pp. 184 - 202)
33 See Eichengreen and Hausmann (1999, p. 18).
34 See Eichengreen and Hausmann (1999, p. 23).
35 Figures presented and calculated after Corsetti, Pesenti, and Roubini (1999, appendix table 24).
36 See Voltaire (1759).
37 See Milgrom and Roberts (1992, p. 166 - 202).
38 Figures taken from Corsetti, Pesenti, and Roubini (1998a, appendix, tables 4,9,10).
39 See Corsetti, Pesenti, and Roubini (1998a, appendix, table 17, 2).
40 See IMF (1998, Table 2.1).
41 See Corsetti, Pesenti, and Roubini (1998a, appendix, table 23, 24, 26).
42 See IMF (1999b, pp. 185 - 213). Although the failure to forecast the Asian meltdown accurately has been criticized by many market participants, the picture might be deceptive. A downgrade of major rating agencies below investment grade means, that many institutional investors, like US pension funds, are forbidden to hold respective bonds by law. An early, unjustified downgrade might lead to a crisis that otherwise would not have occurred. Hence, rating agencies act rather carefully.
43 See previous part.
44 See Eichengreen (1999a, p. 136).