Currency crisis
- First-Generation Crises: Fundamental Disequilibriums
- Second-Generation Crises: The Role of Expectations
- Financial Sector Weakness
- The Political Economy of Currency Crises
- Crisis Prevention and Management
Currency crises are among the most dramatic events in global financial markets. They generally involve large outflows of funds from currencies that investors fear may devalue or sharply depreciate. When a government lowers the level at which a currency is pegged to other currencies this is referred to as a devaluation, while a fall in the value of a market determined exchange rate is called depreciation. To keep their currencies fromdepreciating in the face of exchange market pressure, governments must run down their foreign currency reserves, borrow from abroad, hike interest rates, and/or impose capital controls to slowdown outflows. If they are successful, the speculation dies down, at least for awhile, and the currency value is maintained. More often, exchange rates are eventually forced to adjust. Although the vast majority of currency crises involve downward pressure on currencies, sometime there are strong market pressures for surplus countries to let their currencies appreciate; a vivid example is China beginning in the early 2000s. Crises can vary in their effects as well as in their causes, but they have in common a high degree of exchangemarket pressure.
International speculators are widely blamed for generating crises, especially by government officials eager to shift the blame from themselves. In reality, however, much of the pressure usually comes from ‘‘normal’’ domestic and international firms that are trying to protect themselves against losses rather than to generate speculative gains. Even the speculators are more often the messengers that there are problems than the independent cause of currency crises. Speculative attacks occur when international businesses and investors anticipate a change in the value of the currency, thus adding speculative outflows to the underlying balance of payments deficit.
Crises, though always uncomfortable, can sometimes have good effects, stimulating policy reforms and promoting economic recovery. In other cases, however, especially in developing countries with weak financial systems and considerable debt denominated in foreign currency, the effects of crises can include large recessions and widespread suffering. This is particularly likely when a currency crisis either leads to or was generated by a domestic banking crisis a so-called ‘‘twin crisis.’’
In general, the underlying causes of a currency crisis are inconsistencies between a country’s exchange rate policy and itsdomestic economic policies so that the currency is either over- or undervalued relative to the currencies of other countries. Economists have developed a variety of theoretical models to explain currency crises, and most of these models can be classified into three types. First-generation models consider caseswhere the government is either unable or unwilling to correct inconsistencies between its exchange rate and other domestic policy goals. As these become more serious, a crisis eventually becomes inevitable. Second-generation crisis models analyze cases where these inconsistencies place the economy in a ‘‘zone of vulnerability,’’ making a crisis possible but not inevitable. In these cases, the potential for a crisis outbreak depends critically onmarket perceptions of the government’s willingness and ability to take corrective action. Finally, newer crisis models have focused on financial sector weaknesses and the role of politics in the emergence of currency crises.
First-Generation Crises: Fundamental Disequilibriums
The classic theoretical explanations for currency crises the so-called first-generationmodels focus on fundamental economic disequilibriums such as large government budget deficits. In these models, governments are assumed to pursue fiscal and monetary policies that are inconsistent with maintaining their fixed or slowly adjusting pegged exchange rate regimes. The resulting balance-ofpayments deficits are financed by running down foreign reserves. When the level of reserves falls to a certain threshold, there is a sudden balance-of-payments crisis. Capital flight by domestic residents escalates, even if the country has borrowed little from abroad. This leads to a loss of reserves and forces the government to devalue its currency or float the exchange rate. In these models, expansionary policies and the resulting bad economic fundamentals push the economy into crisis.
The particular causes of the fundamental disequilibriums in past crises have been varied. They often result from a lack of fiscal discipline, but firstgeneration- type currency crises have also been caused by budget deficits in the center country of a pegged exchange rate system(the center country is the nation to whose currency another country’s currency is pegged) or external developments such as export slumps. These problems generally coincide with weak governments facing strong political pressures that prevent them from addressing the emerging disequilibriums by cutting spending or raising taxes to keep monetary expansion under control. Very often, this inability to balance the budget results in high rates of inflation. When economic problems emerge, markets initially often give governments the benefit of the doubt, especially when they have large stores of international reserves. But as imbalances continue, market participants frequently realize well before governments do that the problems are not temporary. The capital outflows that result from this realization involve not only currency speculators but also businesspeoplewanting to hedge against the risk of exchange rate adjustments. These flows in turn tend to increase payments imbalances further and eventually lead to a crisis and a forced adjustment.
Examples of first-generation-style crises include the currency crises of LatinAmerican countries in the post Bretton Woods period, such asMexico in 1976 and Brazil on several occasions. In these cases, the government financed large deficits by creating more money, resulting in inflation. Combined with pegged or slowly adjusting exchange rates, such high levels of inflation led to progressive overvaluation of exchange rates and generated numerous currency crises. Inflation does not need to be in triple or quadruple digits in order to generate currency crises, nor are inflation-driven crises a threat only to developing countries, as demonstrated by Italy’s crisis in 1992. As a member of the European Monetary System (EMS), Italy had pegged its exchange rate to the deutschmark. Although its inflation rate was far below triple digits, it was still well above the inflation rates of most of its partners in the EMS, and in particular higher than that of the center country, Germany. The resulting disparity eventually led to a currency crisis in 1992. Large budget deficits can at times cause currency problems even if they do not lead to monetary accommodation and rising inflation; one example of this is the 2001 crisis in Argentina. As part of a ‘‘shock therapy’’ program designed to break out of a cycle of high and rising inflation, Argentina had adopted a currency board in 1991 that took money creation out of the hands of the government by fixing the domestic currency to a foreign anchor currency, and maintaining full convertibility between domestic currency and the foreign anchor currency. This proved to be highly successful in bringing inflation under control but fiscal profligacy continued. As external developments contributed to aworsening of the situation, the result was a combined debt and currency crisis that ended in abandonment of the currency board and default on the government’s foreign debt.
The collapse of the BrettonWoods system in the 1970s provides an example of an exchange rate system failing due to budget deficits in the center country (in this case theUnited States). For domestic political reasons, the U.S. government delayed tax increases required to pay for the large increases in expenditures associated with the Vietnam War. Seeking to prevent interest rates from escalating, the Federal Reserve Board financed a sizable portion of the resulting budget deficits with monetary accommodation. The consequent overheating of the U.S. economy led to increasing balance-of-payments deficits and ultimately to the currency crisis that was the final straw leading to the widespread abandonment of the Bretton Woods regime of adjustably pegged exchange rates.
Another type of first-generation crisis that need not be directly caused by bad domestic policies involves export slumps. Generally, monetary and fiscal policies that are sustainable under good circumstances become unsustainable when circumstances take a turn for the worse. In a country whose government’s budget depends on export revenues, a sharp decrease in those revenues can cause a similarly sharp change in investor expectations of that country’s ability to meet its debt-servicing obligations. A clear example of this is the 1998 Russian ruble crisis. By the mid-1990s Russia had tamed the runaway inflation that followed the breakup of the Soviet Union and its economic prospects had substantially improved. The domestic coalition favoring economic reform and stability was fragile, however. In the aftermath of the 1997Asian crisis, oil prices fell to half of their precrisis levels due to decreased demand for oil in Asia. As a result, the Yeltsin government, dependent as it was on oil revenues to fund its fiscal programs, found its treasury rapidly being emptied. Initially investors’ belief that Russia was ‘‘too big to fail’’ and their perception that the Russian government was attempting to alleviate the situation by negotiating a loan package from the International Monetary Fund (IMF) and seeking to implement dramatic fiscal reforms caused most investors to give Russia the benefit of the doubt. After the IMF negotiations failed and the Russian legislature, the Duma, removed the most critical components of Yeltsin’s proposed fiscal reform plan, however, it became clear that Russia was not going to be able to adjust its policy enough to avoid a fiscal crisis. In August 1998, there was a severe speculative attack on the ruble, and the Russian government announced that it would no longer support the crawling peg and would default on its foreign-held debt. The remaining investors immediately andfrantically sought to divest themselves of the Russian liabilities and the ruble plummeted.
Second-Generation Crises: The Role of Expectations
A surprising feature of a number of the currency crises of the 1990s was that they hit countries whose macroeconomic fundamentals were not particularly bad. First-generation models were unable to explain, much less predict, many of these crises. In response, a ‘‘second generation’’ of crisis models was developed that focused on investors’ expectations and governments’ conflicting policy objectives and predicted that speculative attacks could occur when a country’s fundamentals were merely in an intermediate or vulnerable zone. Although these models retained the assumption that speculators would not attack countries with good fundamentals, they showed that once a country finds itself in a vulnerable zone, a change in the private sector’s expectations about the future course of government policy can trigger a second-generation crisis.
In contrast to the passive role of policymakers in first-generation models, governments in secondgeneration models are able to take corrective measures, though they may prefer not to. For many governments, exchange rate stability is only one objective among many. Governments often make a trade-off between maintaining a pegged exchange rate and achieving other policy goals (such as low unemployment) that gain political salience when there is an economic downturn. In such situations speculators can lose confidence in the government’s commitment to the exchange rate peg and can decide suddenly to attack the currency. The cyclical state of the economycan thus become an important factor in these models, since the tight monetary policies necessary to defend a currency are politicallymore costly when the economy is in recession than when it is booming.
Speculators are also collectively in a position to influence the future course of policy in second-generation models. The greater the proportion of speculatorswho expect the government to defend the peg at the expense of a worsening recession, the smaller the capital outflows and themore feasible a defense of the exchange rate peg. On the other hand, if most speculators bet against a defense, capital outflowswill be larger and the costs of a defense will be higher. Since theseoutflows themselves can be detrimental to a government’s ability to defend the pegged exchange rate, market expectations can thus prove to be selffulfilling. Such speculation is not irrational, nor is it in itself destabilizing in the sense of going contrary to the fundamentals. It is, however, sometimes thought of as destabilizing in the sense that it can be the proximate cause of a crisis that may not necessarily have been inevitable.
In formal second-generation models, shifts in investor expectations from optimistic to pessimistic are treated as arbitrary. The resulting outcomes are therefore often referred to as ‘‘sunspot equilibria.’’ This can give a misleading impression, however, since in the typical cases of major speculative attacks on currencies, certain events trigger the shift from a ‘‘good’’ to a ‘‘bad’’ equilibrium. In the past, such triggers have included the assassination of a presidential candidate (Mexico in 1994) and crises elsewhere that lead to a reevaluation of underemphasized fundamentals (the wake-up call concerning financial sector weakness that was a major aspect of the contagion in Asia in 1997).
An example ofwhatmany experts consider to be a second-generation crisis is the 1992 93 crisis in the EMS, particularly the speculative attacks on the French franc. The proximate cause of this crisis was the reunification of Germany, which led to a large increase in German government expenditures. As German policymakers were unwilling to offset this higher spending with higher taxes, a large budget deficit emerged. In response, the Bundesbank (Germany’s central bank) raised interest rates. Since Germany was the EMS center country, the logic of the currency systemrequired all other EMS members to tighten monetary policy as well, even though this ran counter to the requirements of their macroeconomic situation at the time. As we have already discussed, this generated crises that led to the breakdown of the EMS system. The crises in Italy and, arguably, theUnited Kingdomwere quite consistent with first-generation models since both the lira and the pound appeared to be overvalued. The 1993 attack on the French franc was not as easy to explain, however, since some of France’s fundamentals were even stronger than Germany’s.
What proved to be more important was the lack of willingness of the individualEMS countries to adjust to the mutual payments imbalances that had emerged. Here Germany held the cards. Even so the EMSmember states put considerable pressure on the Bundesbank to lower interest rates, but the independent German Bundesbank refused to back off its tightmonetary policies.When it became known that Germany would also not continue providing shortterm financing to maintain the pegged rates of the EMS, the only alternatives left for the other EMS members were depreciation or a substantial tightening of their own macroeconomic policies. Given the state of the economy in many of these countries, international financial market participants therefore increasingly questioned policymakers’ willingness to implement policies that would further slow growth and increase unemployment. Eventually, financial markets launched speculative attacks, forcing governments to choose between unpalatable options: abandoning the peg, which they had worked hard to maintain, or facing a deepening recession. In the case of France, for example, the authorities withstood several bouts of speculative pressure before finally devaluing the franc in 1993.
The fact that France’s fundamentals had not been particularly bad led many commentators, especially French officials, to conclude that France had been the innocent victim of destabilizing speculation. France was indeed an innocent victim in this case, but of geopolitical developments in Germany and its own commitment to the European pegged-rate system, not of capricious speculators. This illustrates that crises do not always fit neatlywithin one type of crisis model or another. Some economists consider France’s crisis to be a classic second-generation example, but others argue that it wasmore of a first-generation crisis because of a fundamental disequilibrium generated by Germany’s huge budget deficit. The crisis certainly had elements emphasized in both types of models, as did the 1997Asian crisis, towhich we now turn.
Financial Sector Weakness
Most of the financial world and many economists were shocked in 1997 by the occurrence and the severity of the Asian financial crises. A crisis in the Thai exchange market had been at least partially anticipated, as there had been concerns about the possible overvaluation of the baht for some time. But the spread of the crisis to other countries was almost entirely unanticipated. Most if not all of the affected countries had enjoyed low inflation and robust economic growth for quite some time and appeared by and large to have strong enough economic fundamentals to preclude the type of crises predicted by first- and secondgeneration crisis models. In order to explain these events, economists developed what is sometimes referred to as the ‘‘third generation’’ of crisis models, which place a greater emphasis on the microfoundations of currency crises. Among these refinements are the considerations of financial sector weakness,moral hazard, contagion, and stock as well as flow disequilibriums.
Traditional macroeconomic models paid little attention to the financial sector, and the crises of the 1990s demonstrated that this was a major mistake. The discovery of serious weaknesses in financial sectors can generate major changes in international capital flows, especially where countries have weak international liquidity positions (i.e., high shortterm foreign debt relative to international reserves). Such a situation can quickly turn into a run on the currency without requiring any outright speculation; the scrabble to cover open positions could be sufficient. Indeed, many of the flows of funds during the Asian crisis were of this risk-covering nature.
One of the factors that can contribute to financial sector weakness is a high level of moral hazard, referring in the currency crisis context to the propensity to lend, borrow, or invest in enterprises under circumstances that would usually be considered excessively risky. In some cases, lending institutions or borrowers may have had either explicit or implicit guarantees that the government would cover any losses should the loan or investment fail to return a profit. In this way the contingent liability burden is shifted from the lender, investor, or borrower to the government, creating the situation described by economist Paul Krugman (1996) as a bet of ‘‘heads, I win; tails, the taxpayer loses.’’ If this practice is sufficiently prevalent, even a minor shock to the economy could result in a dramatic increase in government financing to prop up banks whose nonperforming loans have suddenly rendered them insolvent.
Many of the other factors described earlier, most notably financial sector weakness, also played key roles in precipitating the Asian crises of 1997. Although public finances in all of the crisis countries were relatively strong by conventional measures, domestic financial sectors were in bad shape. Both the high levels of nonperforming loans and heavy unhedged foreign borrowing were generated at least in part by problems of moral hazard. The combination of moral hazard and weak risk management and regulatory systems had led to many ill-advised loans, so that in a number of countries the financial sectors were suffering from serious solvency and liquidity problems. While good statistics were not generally available to reveal these problems, international funds continued to pour into these emerging markets because of their apparently strong macroeconomic statistics and as a result of capital account liberalization.
In July 1997, the already overvalued Thai baht came under speculative attack when the severity of Thailand’s problems with nonperforming loans in the financial sector became apparent. Highlighting Thailand’s financial sector problems, this crisis served as a wake-up call to international financial markets, alerting them that Thailand’s unhedged foreign borrowing was much riskier than many had believed. This in turn prompted foreign investors to reassess the vulnerability of their investments to financial sector risks. Thailand was not the only country found to have serious problems in the financial sector, and investors’ assessments of some countries previously perceived to have well-aligned exchange rates, such as Korea and Taiwan, changed (those exchange rates were now seen as overvalued), prompting a run for the exits. Although financial sector problems were not the only contributing factors (asmentioned, theThai bahtwas overvalued and political instability in Indonesia contributed importantly to the length and depth of the crisis there), the Asian crisis did highlight that crises could occur even when data on traditional economic fundamentals indicated a strong economic situation.
The Political Economy of Currency Crises
Recent research has also recognized that political considerations can have an important influence on crises through a number of channels. Since economic policies are generally determined through the political process (one exception being when monetary policy is administered by an effectively independent central bank), political developments can have a strong influence on expectations about future economic policies. Thus political instability and expectations about the election of newadministrations can have an important effect on capital flows even before there is any actual change in policy. Political considerations are also one of the major causes of the development of inconsistencies among policies that lead to fundamental disequilibriums. Because the costs of corrective action often show up faster than the benefits, short-run political pressures can lead to delays in undertaking needed policy adjustments, especially if elections are approaching (see Willett 2007).
One key example of how political considerations can play a role in the outbreak of crisis is the 1999 crisis in Brazil. Past fiscal excesses had left Brazil with a large public debt burden,whichwas substantial but manageable so long as interest rates remained moderately low. In 1999, polls increasingly indicated that leftist presidential candidate Lula da Silva and another leftist politician were likely to prevail in the impending elections. Lula’s past as a trade unionist raised doubts among market participants over whether Brazil’s sound economic policies would be continued after the election. Although a return to hyperinflation was not likely, the high debt levels meant that even amoderate loosening of fiscal policy might be sufficient to force default. As a consequence, a crisis of confidence and a speculative attack on the real ensued. It is important to note that this crisis was precipitated not by actually implemented policies, but rather on the evaluation of expected future policy.
The 1994 Mexican crisis provides another example of the importance of political events in understanding crises. The country had successfully brought down inflation from triple- to single-digit levels, but the combination of strong capital inflows and a slow rate of depreciation designed to limit domestic wage increases had resulted in a large current account deficit and a substantially appreciated real exchange rate.Thus althoughMexico’s domestic economic fundamentals were strong, the economy was vulnerable to a drop in capital inflows, and as with the Brazilian crisis to severe shocks to political stability. In this case, the shock was the assassination of the leading opposition (and proreform) presidential candidate, which prompted rapid reassessment on the part of investors as to whether necessary reforms in Mexico would be implemented, along with fears of a more volatile political climate. The combination of monetary tightening in the United States, temporary preelection loosening of fiscal policy in Mexico, and the emergence of domestic political instability brought the dreaded sharp fall in inflows and the crisis was on.
Political scientist Andrew MacIntyre (1999) has argued thatThailand and Indonesia illustrateways in which different types of governments can exacerbate their vulnerability to crises. Thailand’s political system with its large coalition governments tended to produce great policy stability (or paralysis) due to the difficulties in making policy changes. Intracoalitional politics caused budget and financial regulation reforms to be delayed, further weakening confidence in the Thai financial sector and generating uncertainty about the government’s capacity to act once the crisis occurred. In contrast, Indonesia’s political system was highly centralized and imposed little constraint on executive action. This led to its own set of problems by opening the way for erratic policy behavior. When crisis spread from Thailand to Indonesia in late July 1997, the Suharto government responded decisively and preemptively by widening the band within which the rupiah was allowed to fluctuate from 8 percent to 12 percent. As the rupiah continued to slide, Suharto appeared to be taking appropriate action by agreeing to a series of dramatic reforms as part of an IMF assistance program. Appearances can be deceiving, however, and in this case they certainly were, as Suharto quietly reneged on many of these reforms and as members of his government and family openly undermined others. It is not surprising that investor confidence in the Indonesian government plummeted, not because investors doubted whether Suharto could make the necessary reforms but instead because they doubted that he would.Where Thailand’s case can be characterized as inaction in the face of a looming crisis, Indonesia’s can be characterized as action but of the wrong kind.
Crisis Prevention and Management
There is widespread agreement that it is usually far easier to prevent crises than to fix them once they occur. Unfortunately, prevention is not easy to accomplish. Although analysis has often been able to identify sound reasons for crises after the fact, these reasons are not always easy to identify beforehand. The perception of crises as arising from the unjustified behavior of fickle financialmarkets doubtlessly stems at least in part from this dilemma. Significant advances have been made in the development of economic ‘‘early warning systems’’ for crises, though these tend to detect occasions when a country is vulnerable to a crisis rather than predict precisely when a crisis will occur.
Even when the threat of a crisis is unmistakable, however, countries often fail to take preventive actions in time. Sadly, it often takes the actual outbreak of crises to prompt major policy adjustments, as governments often find it politically difficult to take the actions needed to reduce vulnerability. Consequently, governments are often observed to initially respond to potential crises by addressing capital flows rather than macroeconomic fundamentals by running down international currency reserves, increasing domestic interest rates tomake the country a more attractive capital destination, or implementing restrictions on the flow of capital into or out of the country. When the country is facing a temporary liquidity crunch, such measures may be all that is required (see the 1997 case of Hong Kong, where some of these measures were successfully used to avoid the most serious potential repercussions of the crisis). If, however, the country faces a fundamental disequilibrium (such as is described in first-generation crisismodels), suchmeasures can only delay, not prevent, crises; borrowing and running down reserves cannot indefinitely offset the effects of a fundamental disequilibrium. In such cases, reserve cushions can only give countries the option of adjusting more gradually. Second- and third-generationmodels, however, provide insights intoways that strong reserve positions can help avoid crises. When countries are in vulnerable zones, shocks that would generate a run on the currency of a country with high levels of short-term foreign debt relative to its international reservesmight not generate such a run if the country had a strong international liquidity position. In other words, although strong reserve positions can do little to deal with insolvency (fundamental disequilibrium) problems, they are a valuable instrument for countries with merely vulnerable fundamentals.
The effectiveness of high interest rate policies and capital controls has been the subject of considerable controversy. Because so many factors are typically at work during a crisis it is hard to accurately measure the effects of particular policies. High interest rates may send an ambiguous signal to currency markets, as excessively high interest rates may indicate to the markets that the country is in desperate need of capital not an encouraging sign if the market is already concerned about the safety of investing in that country (Russia in 1998 provides an example of this dynamic). During the Asian crisis, high interest rates in Indonesia did not keep the rupiah from plummeting, because it was not clear at the time whether the high interest rates reflected tight monetary policy or premiums for inflation and political risk.
The effectiveness of capital controls alsodepends a great deal on how they are interpreted by themarket. Capital controls can be seen as prudential measures taken to protect an otherwise stable economy from overly volatile international capitalmarkets, inwhich case they can be effective in reassuring foreign investors. Capital controls can also be seen, however, as stopgap measures intended to prevent capital from flowing out of an already risky economy. In this case, the implementation of capital controls can itself be a signal to the market that the economy is in worse shape than previously thought. As with international reserves, the introduction of capital controls seems more promisingwhen the economy is in a vulnerable zone than when it exhibits strong fundamental disequilibriums. The introduction of capital controls by Malaysia in 1998 has been the subject of different evaluations along these lines. Probably the best we can say about theMalaysian case is that the controls proved to be neither the disaster predicted by some critics nor the panacea envisioned by some advocates.
Recent crises also remind us that international financialflows can befickle and that being the darling of international investors today is not a safe indicator of future stability. Although there remains considerable disagreement among economists about the conditions (if any) under which direct controls on international capital flows can be desirable, there is little disagreement that such flows and financial sectors more generally need to be subject to effective prudential regulation. Such policies need to be carefully crafted: in many countries, policies have contributed as much or more to the generation of crises (through moral hazard and other perverse incentives) as to good prudential oversight.
One of the most notable characteristics of recent currency crises has been the frequency with which they are associated with efforts to maintain various types of adjustably pegged exchange rate regimes. This phenomenon, which has been labeled ‘‘the unstablemiddle hypothesis,’’ was a basic cause of the breakdown of the regime of adjustably pegged exchange rates of the Bretton Woods system. Few analysts dispute that without effective capital controls such adjustably pegged exchange rate regimes are highly prone to crises in the face of substantial international capitalmobility. The debate is whether it is necessary tomove all the way to one extreme or the other hard fixes or relatively free floats to substantially reduce vulnerability to crisis, or whether moving to crawling bands or managed floats is sufficient.
The role of the IMF in crisis prevention and management has been the subject of considerable controversy. Although far from perfect, its ability to identify emerging crisis situations has been much better than its ability to get countries to adopt the necessary preventative actions; the ability of its ‘‘seal of approval’’ to calmcrisis situations has also suffered substantial erosion. The IMF has displayed considerable learning, but efforts at substantial reform of the international financial architecture have been largely unsuccessful, and the IMF’s ability to act as a crisis manager and international lender of last resort has, so far, only marginally improved. This has resulted in greater self-help efforts undertaken bymany countries, especially in Asia, to accumulate high levels of international reserves at both the country and the regional levels. Crisis liquidity is clearly most efficiently provided at the international level, but the failure of substantial reforms at that level makes these individual and regional efforts quite understandable. See also asymmetric information; balance of payments; banking crisis; Bretton Woods system; capital controls; capital flight; capital mobility; contagion; currency board arrangement (CBA); early warning systems; exchange market pressure; financial crisis; foreign exchange intervention; hot money and sudden stops; impossible trinity; International Monetary Fund (IMF); International Monetary Fund conditionality; International Monetary Fund surveillance; international reserves; lender of last resort; speculation; spillovers