Although easy external financial conditions are overall good news for emerging markets, they come with risks that need to be managed. Past episodes of easy conditions often have led to sustained accelerations of domestic demand and significant real appreciation in emerging market countries, sometimes also accompanied by fast credit growth. Importantly, responses to easy conditions have varied in degree, reflecting differences in policies as well as structural features. With interest rates of major advanced economies and global risk aversion both likely to remain low for a sustained period, the challenge for policymakers is to conduct macroeconomic and prudential policies that make the most out of the enhanced financing possibilities while reducing the likelihood of boom-bust cycles. Policies on many fronts will be relevant, with exchange rate flexibility, fiscal policy, and macroeconomic prudential regulation being among the most important in insulating against unwanted demand and credit booms.

Although easy external financial conditions are overall good news for emerging markets, they come with risks that need to be managed. Past episodes of easy conditions often have led to sustained accelerations of domestic demand and significant real appreciation in emerging market countries, sometimes also accompanied by fast credit growth. Importantly, responses to easy conditions have varied in degree, reflecting differences in policies as well as structural features. With interest rates of major advanced economies and global risk aversion both likely to remain low for a sustained period, the challenge for policymakers is to conduct macroeconomic and prudential policies that make the most out of the enhanced financing possibilities while reducing the likelihood of boom-bust cycles. Policies on many fronts will be relevant, with exchange rate flexibility, fiscal policy, and macroeconomic prudential regulation being among the most important in insulating against unwanted demand and credit booms.

What Are the Concerns?

External financial conditions for emerging market economies with strong fundamentals are likely to be “easy” for some time, with low interest rates, low spreads, and ready market access. As discussed in Chapter 1, the expected subpar recovery of advanced economies points to an extended period of loose monetary policy, and financial markets expect that G3 policy interest rates will increase only gradually. Long-term interest rates on G3 risk-free government securities also are likely to remain low, unless fiscal risks gain new importance in the eyes of investors.

At the same time, with effects of the global financial crisis subsiding, global risk aversion has fallen back sharply, and the spreads charged on “risky” assets already have returned to near their historic lows, at least for investment grade assets. As perceptions of risk decline, investors facing low returns in advanced economies will naturally look for yield in emerging markets.

This chapter examines the diverse set of macroeconomic and financial concerns that arise during a prolonged period of easy external financial conditions. The issues and risks are wide-ranging and interrelated, and critically contingent on policy responses.

These various concerns, analyzed over the course of this chapter, can be summarized as follows.

Easy external financial conditions have a strong potential to trigger accelerated growth of domestic demand. Low external interest rates and greater access to foreign finance can directly stimulate both consumption and investment. The boost to demand will be reinforced if foreign investors bid up asset prices in emerging markets, increasing their residents’ wealth. In turn, growth of domestic demand—in excess of output growth—will drive a widening current account deficit and an appreciation of the real exchange rate (other things constant). These forces result in large inflows of foreign capital—readily available at low cost, but not forever—that create their own set of macroeconomic issues and prudential risks, especially if accompanied by a parallel boom in domestic credit. Booming demand and credit could lead to dynamics that end in recession or even crisis. This fate is not inevitable, of course, as policies can be decisive as to whether such risks build or stay contained.

Among the issues raised by easy external financial conditions, implications for economic growth are often on the minds of policymakers. The potential concern is that real appreciation will reduce the profitability of the traded goods sector and eventually also the volume of net exports—in turn with negative implications for growth. Of course, in the particular context of booming domestic demand, contractions in net exports need not necessarily affect the overall level of activity.29 A very different concern relates to the supply side: real appreciation, though temporary, could damage a country’s economic growth trend in the longer term, if the sectors that lose some of their profitability play a special role in productivity growth. This concern is akin to “Dutch disease.”

This chapter sheds light on the relevance of these risks by analyzing the experiences of a group of Latin American countries—compared with other countries—and by surveying the relevant literature. In light of the lessons from these past experiences, the chapter concludes with a review of policy options to manage the risks that would emerge in the expected new episode of sustained easy external financial conditions.

A central message of this chapter is that policies can act either to mitigate or to amplify the risks arising from episodes of easy external financial conditions. For example, flexible exchange rates can serve to choke off foreign inflows, whereas systematic efforts to limit (nominal) appreciation can stimulate capital inflows and the expansion of credit. Fiscal policy is also relevant, as governments can choose to partially offset surging private demand—or instead choose to accelerate their own expenditure amid easy financing opportunities and booming tax bases. Macroprudential and other policies can also play important roles.

The empirical investigations here, using several methods, focus especially on the experiences of two groups of countries. One is the “FCE” group of financially integrated commodity exporting countries identified in Chapter 2: Brazil, Chile, Colombia, Mexico, and Peru. The second is a comparator group of relatively small advanced economies that are also commodity exporters: Australia, Canada, New Zealand, and Norway, or “ACE4,” all of which permit a very high degree of exchange rate flexibility—and which have been relatively successful in managing episodes of easy external financial conditions. Reference is also made to experiences of countries in Asia, and in Eastern and Central Europe.

Booming Domestic Demand …

Looking over the last twenty-five years, the two episodes of easy external financial conditions (Box 3.1) coincided with accelerated demand growth. FCE domestic demand grew on average 4 percentage points per year faster than in other periods (Figure 3.1). This pickup was somewhat greater in the 2000s episode (when FCE domestic demand also got a boost from rising commodity export income—see below).

Although this boom in FCE domestic demand was also associated with a marked pickup in GDP growth, even beyond trend, demand outstripped production in both episodes, and therefore real imports surged. The biggest acceleration of demand was in fixed capital formation (by 7 percentage points), while private consumption picked up by 3 points and government consumption by 2 points.

On average across the two episodes of easy conditions, domestic demand in FCE picked up more than in emerging market economies of other regions, and also more than in the ACE4 group of advanced commodity exporters. A closer look at each period, however, shows also marked pickups of demand in Asian countries in the 1990s episode, and in both European and ACE4 countries in the 2000s episode (Table 3.1).

Figure 3.1.
Figure 3.1.

Domestic demand in the FCE countries grew faster during earlier periods of easy external financial conditions.

FCE: Domestic Demand Growth under Easy External Financial Conditions

(Annual percent change, average over the period)

Source: IMF staff calculations.

When Are Financial Conditions “Easy”?

Easy external financial conditions are here defined as years of simultaneously low real short-term interest rates and global risk aversion. Real interest rates are measured as the U.S. federal funds rate less expected core U.S. inflation. The indicator of global risk aversion is proxied by the Chicago Board of Options Exchange Volatility Index (VIX), with low values of the VIX usually associated with lower spreads. These two indicators are considered low when below their median annual values computed over the last twenty-four years (1986–2009).

According to these joint criteria, external financial conditions were “easy” during two multiyear episodes: 1991–96 (excluding 1995) and 2004–07. Real interest rates and global risk aversion were both broadly equal across the two episodes, on average. In the second episode, however, commodity prices increased considerably, and world growth was much faster than in the 1990s episode. At mid-April 2010 values of real interest rates and the VIX, external financial conditions were easy again.


Episodes of Easy Ext ern al Financial Co nditions 1/

(In percent and index)

Source: Haver Analytics.1/ Shaded areas show periods when low real interest rates coincide with low global risk aversion. The black line represents medians over 1986-–2009. Data for 2010 are averages through April.
Note: This box was prepared by Jorge Iván Canales-Kriljenko.

A multivariate VAR analysis confirms that FCE’s domestic demand has systematically responded, positively and considerably, to lower external real interest rates and greater risk tolerance in global financial markets. (Importantly, this result holds after controlling for the positive effects of external demand and commodity prices.)

In this VAR analysis, these two indicators of external financial conditions together explain about 35 percent of FCE domestic demand fluctuations over a year. For the ACE4 countries, this sensitivity also exists, but the corresponding share is only about 20 percent. Figure 3.2 shows that FCE’s demand response to shocks of a given size is greater than for the ACE4, for both of these aspects of financial conditions.

This greater sensitivity of FCE to external financial conditions could reflect a variety of reasons. These countries may have more prevalent financing constraints, both at the household and firm level, that loosen at times of easy financial conditions. Declines in global risk aversion may expand financial possibilities to a greater degree in Latin American countries, as these are considered more risky than those in the ACE4.

Finally, policy frameworks and policy responses are relevant. For example, government spending in the ACE4 countries has been much less responsive to easy conditions than in FCE, where a pickup of public spending contributed to the expansion of overall domestic demand. In the same vein, exchange rates have been more flexible in the ACE4, as will be discussed later.

Table 3.1.

Emerging Markets: Selected GDP Components under Easy External Financial Conditions 1/2/

(Percent a year, period average)

article image
Source: IMF staff calculations.

Significant under a standardized normal statistic at 5 percent probability;

Significant at 10 percent probability.

Based on annual data 1986–2008. Easy liquidity conditions are defined as those with low VIX and federal funds rates. The years are 1991–94, 1996, and and 2004–07.

Countries include Brazil, Chile, Colombia, Mexico, and Peru in Latin America; India, Indonesia, Korea, Malaysia, Philippines, and Thailand in Asia; Czech Republic, Hungary, Poland, and Turkey in Europe (flexible); Bulgaria, Estonia, Latvia, and Lithuania in Europe (fixed); Egypt, Israel, Lebanon, Pakistan, South Africa, and Tunisia in the Middle East and Africa; and Australia, Canada, New Zealand, and Norway, among advanced commodity exporters.

Only the 2000s epidose, owing to data availability.

… And Currency Appreciation …

Significant real appreciation took place in Latin America during earlier episodes of easy external financial conditions (Figure 3.3). The appreciation for the FCE was on average 4 percentage points a year higher than in other periods (more in the second episode). Real appreciation was also prevalent in the ACE4 and in the other emerging market economies, particularly in the 2000s episode.

Econometric analysis confirms that the real exchange rate responds to external financial conditions, even after controlling for the effect of commodity prices (Figure 3.4). Both lower VIX and lower international interest rates tend to appreciate the REER. Moreover, the reaction appears to be larger in the FCE countries than in the ACE4 countries. Episode analysis is consistent with this observation. For example, during the 2000s episode, currency appreciation in the FCE averaged 4.8 percent, whereas in the ACE4 it averaged 3.4 percent. This difference in behavior of real exchange rates likely reflects the greater sensitivity of FCE to external financial conditions rather than exchange rate policy (on the whole, the ACE4 have practiced less foreign exchange market intervention than the FCE have).

The joint occurrence of domestic demand booms and real appreciation during episodes of easy external conditions is not surprising. Both respond to a similar set of factors. Moreover, an expansion of domestic demand (in excess of potential output growth) will bid up the relative price of nontradable goods, leading to a real exchange rate appreciation. In the other direction, currency appreciation may boost domestic demand through financial wealth effects. Currency appreciation could further boost domestic demand if it shifted income from firms’ profits to the wages of workers with a higher propensity to spend.

Previous literature has confirmed the importance of domestic demand as a determinant of the REER by identifying the effects of government consumption in particular. A typical finding is that an increase in such expenditure by 1 percentage point of GDP is associated with appreciation of the real exchange rate of about 2½ percent (Isard and Faruqee, 1998; and Lee and others, 2008).

Figure 3.2.
Figure 3.2.

The systematic response of domestic demand to external conditions has been stronger in the FCE countries than in the ACE4 countries. 1/

Sources: Haver Analytics; and IMF staff calculations.1/ Each chart shows the dynamic response of domestic private demand (expressed in percentage change, year over year) to a 1-standard deviation shock to the VIX and real interest rates (expressed in percent per year) and 1-standard deviation shock to external demand and commodity prices expressed in percentage change, year over year). The impulse response function is based on a vector autoregression model, estimated over the period 1994Q2 and 2009Q2, which includes external domestic demand, VIX, interest rates, commodity prices, and private domestic demand.2/ Simple average of Brazil, Chile, Colombia, Mexico, and Peru.3/ Simple average of Australia, Canada, New Zealand, and Norway.
Figure 3.3.
Figure 3.3.

Earlier episodes of easy money also coincided with significant real appreciation in the FCE countries.

FCE: Currency Appreciation Under Easy External Financial Conditions 1/

(Annual percent change, average over the period)

Source: IMF staff calculations.1/ Both increases are statistically significant.
Figure 3.4.
Figure 3.4.

Real exchange rates respond more to external financial shocks in the FCE than in the ACE4 countries.

Response of Real Exchange Rate to Global Financial Conditions 1/

Sources: IMF, International Financial Statistics; Bloomberg, L.P.; Haver Analytics; and IMF staff calculations.1/ The figure shows the dynamic response of the real effective exchange rate to a 1 standard deviation shock to the VIX and real interest rates (expressed in percent per year). The impulse response function is based on a vector autoregression model, estimated over the period 1994Q2 and 2009Q2, which includes external domestic demand, VIX, interest rates, commodity prices, private domestic demand, and the real exchange rate.2/ Simple average of Brazil, Chile, Colombia, Mexico, and Peru.3/ Simple average of Australia, Canada, New Zealand, and Norway.

Sensitivity to global financial conditions is also apparent in how nominal exchange rates respond to short-run movements of the VIX. Currencies of many emerging markets—and also of some smaller advanced economies—tend to appreciate when the VIX declines. Many currencies also tend to appreciate when certain commodity prices rise; for example, this is clearly the case for most currencies of the FCE and ACE4 countries, reflecting the importance of commodity exports to their economies (Figure 3.5). Studies of the longer-run determinants of real exchange rates have confirmed the importance of the commodity terms of trade for many countries.

Figure 3.5.
Figure 3.5.

Nominal exchange rates systematically react to information contained in the VIX index (as a proxy for global risk aversion) and commodity prices, after controlling for the value of the U.S. dollar.

FCE and Other Regions: Sensitivity of Exchange Rates to Global Shocks 1/

Sources: Bloomberg, L.P.; and IMF staff calculations.1/ Regressions of weekly changes (Friday over Friday), from January 2000 to January 2010, excluding the period September 1, 2008 to May 31, 2009. Exchange rate increases (decreases) denote depreciation (appreciation).2/ Country-specific coefficient times 1 standard deviation of either the VIX, U.S. dollar index, or commodity prices. Standard deviations expressed as log changes times 100.3/ Change in exchange rates regressed on changes in the VIX, U.S. dollar index, and each of an overall commodity price index, a metals price index, food price index, and the WTI oil price index. Commodity price index for each country selected as the specification with the highest R-squared: composite index for Brazil, Colombia, India, Indonesia, Israel, Korea, Malaysia, Mexico, New Zealand, and the Philippines; metals index for Australia, Peru, South Africa, Thailand, and Turkey; copper price for Chile; oil price index for Candad, Czech Republic, Hungary, Norway, and Poland.4/ Standard deviation of weekly log change in local currency per U.S. dollar exchange rate, times 100.5/ Dependent variable defined as local currency per euro.

… May Lead to Vulnerabilities …

Large increases in domestic demand could end badly if the economy is forced to adjust abruptly. The increases in demand are often partly financed by capital inflows and lead to current account deficits. The associated dangers include financial crises triggered by sudden stops and current account reversals, whose effects may be aggravated by excessive financial risks that may be taken during the demand upswing.

An extensive literature on “early warning indicators” and determinants of crises highlights current account deteriorations, large credit growth, and exchange rate misalignment, among the most important developments that should trigger alarm. Balance-sheet exposures of the corporate and financial sectors, such as currency or maturity mismatches, could aggravate the impact (Table 3.2).

Some well-known financial crises have occurred after periods of easy external financial conditions, as was the case, for example, in the Latin American crisis that followed the easy episode that ended in the 1982 financial crisis, the Asian crisis that followed the 1990s episode, and those in many European countries that followed the 2000s episode. In contrast, the FCE countries analyzed here escaped a financial crisis following this last episode—an unsurprising outcome in light of their stronger vulnerability indicators.

How did vulnerability indicators evolve during the earlier two episodes of easy financial conditions?

  • In both episodes, the region experienced rising capital inflows. In particular, the increase in crossborder liabilities in FCE was faster than at other times, including through foreign direct investment (FDI) and portfolio flows. The latter, however, were notably smaller on average in the 2000s episode. At the same time, FCE also tended to accumulate crossborder assets during these episodes, particularly in the 2000s episode. Thus, on net, the increase in capital inflows was relatively smaller in the second episode than in the first episode (Box 3.2).

  • Somewhat surprisingly, current account/GDP ratios in FCE did not worsen significantly despite the noted pickups of domestic demand and import volumes (Table 3.3). Although the current account deficit worsened slightly during the first episode, the current account balance actually improved during the episode of the 2000s, which is explained by the good fortune of large terms-of-trade gains. The real exchange ratio appreciation during both periods also raised these ratios’ U.S. dollar GDP denominator.30

  • The credit-to-GDP ratio increased in Latin America by about 2 percentage points of GDP every year, in both episodes of easy conditions. In other years, this ratio declined slightly. In other regions, credit-to-GDP ratios sometimes were affected strongly as well, notably in Asia in the 1990s episode and in Europe in the 2000s episode (Figure 3.6). Especially striking was the surge in credit growth in the European countries with fixed exchange rate regimes, suggesting that exchange rate regimes matter, an issue explored later in the chapter. Banking indicators in Latin America did not change much, with nonperforming loans and capital adequacy virtually unchanged.

Table 3.2.

A Glimpse at the Literature on Early Warning Indicators, Sudden Stops, and Current Account Reversals

article image
Table 3.3.

Selected Balance of Payments Accounts under Easy External Financial Conditions 1/2/

(Percent of previous year GDP, period average)

article image
Sources: IMF, International Financial Statistics; and IMF staff calculations.

significant at the 10 percent probability

Significant under a standardized normal statistic at 5 percent probability.

Based on annual data 1986–2008. Easy external financial conditions years are identified as 1991–94, 1996, and 2004–07.

Brazil, Chile, Colombia, Mexico, and Peru.

Capital Inflows when External Financial Conditions Are Easy

Private capital inflows (gross and net) increased substantially in past episodes of easy external financial conditions. In both episodes, Latin American crossborder liabilities rose at roughly the same rate (about 6½ percent of previous year GDP on average). Nevertheless, crossborder assets of Latin American residents rose significantly faster during the second episode. As a result, net capital inflows were larger in the first episode (5.6 percent of GDP) than in the second (2.6 percent of GDP).


FCE: Gross Capital Inflows: Increase in Crossborder Liabilities with Nonresidents 1/

(Percent of previous year GDP)

The composition of the crossborder assets and liabilities changed from one episode to the other. On the crossborder liability side, the share of foreign direct investment (FDI) in Latin America rose, and that of nonresident portfolio investments in the region fell, during the 2000s compared with the 1990s episode. On the crossborder asset side, portfolio investments rose much faster than FDI abroad and even exceeded the increase in crossborder portfolio liabilities. Given this, the region arguably faced lower risks of sudden stops during the second episode than during the first episode because FDI in the region is typically more stable than portfolio flows, and portfolio assets are more liquid than FDI.1


FCE: Net Capital Inflows: Crossborder Liabilities Minus Crossborder Assets

(Percent of previous year GDP)

How did these flows compare with other regions? All regions experienced a large increase in inflows during the episodes of easy external financial conditions.2 The net increase, however, was significantly higher during the 1990s in Latin America and Asia and during the 2000s in Emerging Europe. The ACE4 countries actually experienced net outflows. Gross inflows, however, increased markedly during the second episode across the board, which is consistent with the increased financial globalization observed during the last decade. The ACE4 experienced the largest increase in gross inflows during the second episode, but even higher gross outflows. As in the FCE region, gross FDI inflows increased across the board. In contrast to the experience of FCE, however, gross portfolio inflows increased markedly elsewhere during the second episode.


Emerging Markets: Net Capital Inflows Under Easy External Financial Conditions

(Percent of previous year GDP, increase relative to other periods)

Note: This box was prepared by Jorge Iván Canales-Kriljenko.1 For a review of Latin America’s experience with FDI, see the October 2009 Regional Economic Outlook: Western Hemisphere (Box 2.4).2 Over the period 1990–2009, Vera-Martin (forthcoming) identifies 31 episodes in which individual countries experienced large capital inflows. The frequency of these episodes peaks during the periods identified by this chapter as easy external financial conditions.

In sum, Latin America experienced a sharp pickup in gross inflows during periods of easy external financial conditions, although net of outflows, they were notably smaller during the most recent episode. Yet, the current account ratios remained well contained (helped by rising commodity export prices), although credit growth notably increased.

… And Hinder Competitiveness and Possibly Growth

Currency appreciation can trigger two distinct concerns about economic activity. First, appreciation may damage activity in the short term, through a contractionary effect on aggregate demand. Second, it is often suggested that appreciation, if prolonged, could hurt growth through effects on the economy’s supply side. Concerns of the latter type are variations on a “Dutch disease” issue. At the center of most of these fears is the belief that certain sectors—traded goods, or exports, or industrial sectors that are associated with exporting—are in some way special in terms of being conductive to economy-wide growth.

Figure 3.6.
Figure 3.6.

Credit growth responses have varied by region, growing fast in Asia in the 1990s and in Europe in the 2000s.

Emerging Markets: Credit Growth Under Easy External Financial Conditions 1/

(Percent increase over other periods)

Sources: IMF, International Financial Statistics; and IMF staff calculations.1/ Increase relative to periods in which external financial conditions were not easy.
Figure 3.7.
Figure 3.7.

Despite the real appreciation, export volumes actually increased during the earlier episodes of easy money.

FCE: Real Appreciation and Export Growth During Episodes of Easy External Financial Conditions

(Annual percent change, increase in means during episodes)

Source: IMF, Information Notice System.

Therefore, the first empirical question is to what extent—and when—has appreciation been associated with reduced exports?

This is not an easy question to answer, as so many factors are involved in export performance, and in determining the level of the exchange rate. No tight association exists simply between real appreciation and weaker export volumes. Looking at episodes of easy external financial conditions illustrates the problem: FCE export volume growth actually picked up, despite fairly strong real appreciation at the same time (Figure 3.7).

This joint occurrence of appreciation and faster export growth could reflect developments that can induce both real appreciation and growth of exports, such as the rise in commodity prices (if it leads to expansion of commodity production), higher demand volume, or productivity growth in the export sector. Adverse effects of appreciation may also take time to materialize.

Econometric exercises on the dynamic relationship between appreciation and export growth—taking into account other factors—help detect the relationships involved, but are often inconclusive when applied to emerging market countries. System estimation with annual data suggests that export volumes in Latin America respond somewhat to growth in advanced economies and China, as well as to changes in the real exchange rate and export prices.31 Estimation of standard types of “trade equations” also suggests some short-term negative effect of real appreciation on FCE export volumes, when controlling for external demand and commodity prices.32

Regarding the sensitivity of imports, an association between appreciation and import volumes can be detected much more clearly. In addition to the direct, expenditure-switching effect of appreciation on import volume, this association likely reflects also an indirect channel: to the extent that appreciation stimulates overall domestic demand, this additional demand will fall partly on imported goods.

Taken together, these points do not provide much support for the concern over contractionary behavior of aggregate demand at times of easy external financial conditions. Although appreciation itself will have some negative impact on net exports, this impact has been more than offset by the strong growth of domestic demand. Expenditure-switching effects on the side of exports seem limited as the price elasticity of exports in Latin America may be smaller than in other countries given the larger weights of commodities in their total exports. The effects involved on the side of import substitution appear larger, but the surge in domestic demand seems strong enough to offset them.

Thus, concern about growth effects of exchange rate appreciation probably should turn more to supply-side effects. However, the existing literature does not settle the issue of whether currency appreciation—either temporary or permanent—hurts growth in the long term, or of how likely or large such an effect might be. The classic theory on Dutch disease is ambiguous about its effects, and the evidence is inconclusive. A new literature exploring the long-run effects of alternative exchange rate levels on growth has provided some evidence, but its interpretation remains controversial and its relevance unclear. This new literature does not address effects of temporary real appreciation but rather the consequences of an exchange rate level that is sustained—by some means33—for many years, as part of an ongoing development strategy that favors the profitability of certain sectors (Box 3.3).

Real Appreciation and Trend Growth: An Overview on Dutch Disease and Related Concerns

Concerns about sustained adverse growth effects of real appreciation have been explored for many years, going back to the “Dutch disease” literature of the early 1980s. The debate continues today—including with a related new literature that proposes further links from the real exchange rate to growth—but is still far from resolved. While the logic of some of the theoretical arguments for this link is clearly established, these arguments lean heavily on special assumptions about the nature of economic growth. And the evidence seems insufficient: to date, while there are a few exceptions, empirical studies of Dutch disease have focused mainly on how shocks that cause real appreciation may affect the level of traded goods production—rather on whether this sector has a special role in economic growth, or on whether it is permanently damaged by temporary episodes of real appreciation.


In its original form, Dutch disease results from an increase in the price of (or discovery of) an exportable natural resource, which lowers the relative price of tradable goods. This reallocates factors of production into the extraction of the natural resource and nontradable goods and away from the other tradable goods (Corden, 1981). From this, at least two distinct concerns are raised. On the one hand, the change in relative prices may hamper the development of an infant manufacturing export industry, which may have been getting stronger by learning by doing. This may decrease overall growth in the long term (Van Wijnbergen, 1984). However, this effect will not be permanent if learning by doing takes place in both the tradable and nontradable goods sectors (Torvik, 2001; Adam and Bevan, 2004). Second, with economies of scale, deindustrialization would occur if the appreciation persists for a long time or is too large (Krugman, 1987). If increasing returns to scale exist in the nontradable goods sector, the appreciation would actually increase growth (Sachs and Warner, 1999). However, if the real appreciation is an equilibrium response to the shock, there might be no disease after all (Edwards and Aoki, 1983; Harberger, 1983).

The new literature addressing the linkages between the level of real exchange rate and growth, still in its infancy, identifies several channels through which they could operate. First, the exchange rate may affect growth through externalities associated with export-linked activities, particularly from the manufacturing tradable sector (Prasad and others, 2007; Rodrik, 2008). Second, an “overvalued” exchange rate may hinder growth through a reduction in domestic saving. Dooley, Folkerts-Landau, and Garber (2004) argue that an appreciated real exchange rate tends to shift demand away from nontradable goods into tradable goods, which requires a reduction in real interest rates to maintain internal equilibrium, reducing domestic saving rates. Finally, an overvalued real exchange rate may be associated with higher real wages, leading firms to lower investment as well as saving rates required to finance that investment (Levy-Yeyati and Sturzenegger, 2007).


The evidence from the empirical literature on Dutch disease is mixed. This literature mostly focuses on the effect on the real exchange rate arising from a series of shock events (linked to natural resources, remittances, foreign aid, or capital inflows) and the reallocation of resources between the tradable and nontradable sectors. Whether or not such reallocation has a permanent effect on growth is usually not examined empirically. Regarding remittances, most studies find that there are Dutch disease effects—real appreciation and a decrease in the tradable to nontradable output ratio (Lartey and others, 2008; Acosta and others, 2009; Amuedo-Dorantes and Pozo, 2004). Rajan and Subramanian (2005), in contrast, find evidence that remittances do not create such effects. Studies about the effects of an increase in foreign aid are inconclusive. Some find that aid inflows lead to real exchange rate appreciation and deindustrialization (Rajan and Subramanian, 2005 and 2009; Adenauer and Vagassky, 1998), but others show that Dutch disease is not an inexorable consequence (Amuedo-Dorantes and Pozo, 2004; IMF, 2005; McKinley, 2005). Furthermore, it is unlikely that a sustained increase in aid would, through Dutch disease effects, hurt long-term growth (Barder, 2006).

While there is some evidence suggesting that sustained overvaluation of the exchange rate level hinders growth and that undervaluation stimulates it, the evidence is still insufficient (Hausmann, Pritchett and Rodrik, 2004; Prasad and others, 2007, Levy-Yeyati and Sturzenegger, 2007; Rodrik, 2008; Eichengreen, 2008). There is little systematic evidence on externalities that may foster growth. Similarly, while some studies try to document positive spillover effects from manufacturing, the evidence is inconclusive (Eichengreen, 2008).

Finally, on the hypothesis linking exchange rate levels and saving rates, Montiel and Servén (2008) argue that the hypothesized causal link from the real exchange rate to saving is empirically weak.

Note: This box was prepared by Nicolás Magud and Sebastián Sosa. It draws on a more extensive literature review and discussion by the same authors (Magud and Sosa, forthcoming).

In short, fears about the adverse impact of currency appreciation on growth can be valid in principle, but it is difficult to gauge their practical relevance. In the absence of clear evidence on the size or permanence of any such effects, policymakers still must rely on judgment and perhaps focus relatively more on the booming consequences of easy conditions (which by itself would exacerbate the competiveness issue).

Policy Matters

Policy frameworks and policy responses can influence the degree of risks associated with easy financial conditions—either mitigating or compounding those risks.

To inform future policymaking, it is useful to first review the typical responses of policies to such conditions.

… In the Past …

Macroeconomic and financial policies of FCE countries have improved markedly over the years, and thus differed significantly across the two earlier episodes of easy external financial conditions. One major achievement was the reduction in inflation in the context of a significant revamping of monetary policy frameworks that allowed increased exchange rate flexibility. Critical for this increased flexibility has been the establishment of a credible nominal anchor—the inflation target—which has lessened the effects of exchange rates on inflation and thus helped reduce the “fear of floating.” Even in the context of much more flexible exchange rates, however, some FCE central banks at times have purchased large amounts of foreign exchange, particularly in 2006–07, amid large capital inflows and a spike in revenues from commodity exports.34

Fiscal policy and public debt management also have improved significantly, and systems of financial regulation and supervision saw major improvements (see the May 2009 Regional Economic Outlook: Western Hemisphere).

Have these fundamental changes of policy frameworks made a difference in how FCE economies respond to changes in external financial conditions? Further evidence from the VAR-based econometric approach mentioned earlier suggests that something has changed. In particular, sensitivities to the VIX and also to U.S. real interest rates seem to have been larger and faster in the past than more recently.35

Did exchange rate flexibility matter? In moving toward greater exchange rate flexibility, the region has lowered the likelihood of currency crises and the likely adverse impact on economic activities associated with interest rate increases to defend an exchange rate when external financial conditions tighten. The increased flexibility—including in the short term—also played a role in reducing the sustained attractiveness of ongoing portfolio investment inflows (as noted earlier, these were smaller during the 2000s episode). Notably, the shift to allowing higher nominal exchange rate flexibility was not associated with larger real appreciation during the second episode, as might have been expected given the terms-of-trade gains experienced then.

Looking at a wider range of country experiences suggests that monetary policy frameworks and exchange rate regimes are important. It is helpful to look at a broader sample because the FCE and ACE4 countries in general now follow broadly similar regimes and monetary frameworks. Although substantial cross-country variation exists across regimes, simple cross-country scatterplots of data during the 2000s episode of easy conditions suggest that countries with more flexible exchange rate regimes tended to have lower domestic demand growth in excess of their trend growth rate, which leads to smaller current account deficits. Because more flexible exchange rate regimes tended also to accumulate less international reserves, a positive association between reserves and domestic demand growth is also apparent (Figure 3.8).36

Figure 3.8.
Figure 3.8.

Countries with more flexible exchange rate regimes and less reserve accumulation tended to experience lower surges in domestic demand.

Emerging Markets: Exchange Rate Regimes and Domestic Demand Growth: 2004–07 1/

Source: IMF International Financial Statistics, Information Notice System, and World Economic Outlook databases.1/ Demand growth is percent change less average GDP growth over the last 10 years.

For example, a number of European countries with strong fixed exchange rate regimes provide a sharper contrast to those following inflation targeting regimes in the FCE and ACE4 country groups. Within Emerging Europe, those countries with more flexible exchange rates had a much less pronounced credit boom, smaller current account deficits, and lower inflation—despite allowing substantial nominal appreciation—than did other countries (Bakker and Gulde, forthcoming). The earlier Table 3.1 also shows that European countries with fixed pegs experienced much larger increases in domestic demand. Looking at a broader country sample, the Spring 2010 Global Financial Stability Report finds that the link from global liquidity to “receiving country” asset valuations is weaker for countries with greater degrees of exchange rate flexibility (IMF, 2010).

Accumulation of international reserves in FCE countries was also faster during the earlier two episodes of easy external conditions than during other periods. This suggests that policies of reserve accumulation have responded to low real interest rates or to global risk aversion or both—a pattern which is verified in an econometric analysis that controls for other variables. According to that estimated relationship, a decrease of 1 standard deviation in the VIX increased the growth rate of international reserves by about 4 percentage points within one year, whereas an analogous decrease in real interest rates increased reserves growth by an additional 2.5 percentage points.37

Such a policy pattern of building reserve holdings faster at times of easy global financial conditions—despite the lower returns earned at the time on reserve holdings—suggests that purchases of foreign exchange often were timed to lean against pressures for currency appreciation, in part owing to fear of eventual contractionary effects of a firmer currency. Central banks also may have been motivated to accumulate reserves more rapidly in those times for prudential reasons, seeking to self insure against the risk of a sudden reversal of easy external financial conditions. Having an ample stock of reserves available could limit the effects of a potentially large currency depreciation that could occur when those easy conditions end. Indeed, reserve levels are key vulnerability indicators and there is some evidence that they are factored into emerging market borrowing costs (see Baldacci and others, 2008; Rojas and Jaque, 2003; and Rowland, 2004). Reserve accumulation may also result from intervention that is motivated by a type of “industrial policy”; that is, as countries try to defend their export sectors from a deterioration in competitiveness.

Conversely, because reserve accumulation can easily reduce (actual or perceived) exchange rate flexibility, it may encourage domestic demand growth and capital inflows, increasing the risk faced by the country during a sudden stop. In containing currency appreciation, reserve accumulation may allow expectations of further appreciation to persist for some time while keeping asset prices attractive. In contrast, a fast, upfront appreciation would increase the expected financial cost of external financing by generating expectations of future depreciation, and, at the same time make domestic assets less attractive for foreigners, damping the effect of easy external financial conditions on domestic demand.

In sum, exchange rate flexibility can play a crucial role in containing the risks that may build during periods of easy external financial conditions. Yet, it requires having a strong alternative nominal anchor in the form of a credible objective for inflation as well as prudential regulations that discourage large currency exposures. In turn, reserve accumulation can help central banks play a supportive role later, in the event of a subsequent sudden stop.

Credit in FCE was allowed to grow significantly faster in both episodes of easy external financing, as noted earlier. Explicit policies for curtailing credit growth were in general isolated (other than increases in interest rates that were called for under the monetary framework to contain inflation). The pickup of credit, however, was not as much as in other regions, especially Emerging Europe. Some countries, however, took notable specific steps, particularly during the final boom years before the global crisis hit. In particular, measures adopted by countries in the FCE region, for example in Brazil, Colombia, and Peru, included tightening reserve requirements on some financial instruments, prudential regulations on exchange rate risks, and reinforcing dynamic provisioning requirements, among others.

Regarding fiscal policy responses, FCE governments expanded their spending at somewhat faster rates during both periods of easy external conditions than at other times (Medina, forthcoming). That pickup, however, was much smaller than that seen in private domestic demand, and also smaller than that of GDP. Amid such accelerated growth of the bases for income tax and VAT revenue, and with commodity-based revenues also surging in the second episode, the pickup of government expenditure could be seen as relatively modest, particularly in the second episode.38 Still, expenditure policy was by no means acyclical. Estimated impulse responses confirm that FCE spending tended to increase under easier external financial conditions and more favorable external demand. By way of comparison, estimated responses for ACE4 governments’ expenditure were much smaller, and may not be statistically significant (Figure 3.9).

How have these policies interacted with external developments in determining outcomes? In particular, how much of the approximate 4 percentage point pickup in domestic demand in the first episode and of the approximate 5 percent increase in the second episode could be attributed to external financial conditions? Panel system estimation of the key relationships discussed in this chapter may provide some hints. This system attempts to provide a first approximation at how domestic demand growth, real appreciation, export growth, and export prices relate to easy external financial conditions and world demand (Table 3.4).39

The estimated relationship for domestic demand suggests, taking as a given the policy responses over the period studied, that about 3 percentage points of its increase was associated with external financial conditions during both episodes, whereas about 1 additional percentage point in the second episode was associated with the increase in commodity prices. Government consumption and credit growth explained most of the rest, suggesting that they also contributed to the boom and real appreciation.

Figure 3.9.
Figure 3.9.

The systematic responses of credit and government consumption have been stronger in FCE countries.

Response to Credit and Government Consumption to External Financial Conditions 1/

Sources: Bloomberg, L.P.; Haver Analytics; and IMF staff calculations.1/ Each panel shows the dynamic response of financial sector credit and government spending (expressed in percentage change, year over year) from a 1 standard deviation shock to the VIX and real interest rates (expressed in percent per year). The impulse response function is based on a vector autoregression model, estimated over the period 1994Q2 and 2009Q2, which includes external domestic demand, VIX, interest rates, commodity prices, private domestic demand, and each of the two variables alternatively.2/ Simple average of Brazil, Chile, Colombia, Mexico, and Peru.3/ Simple average of Australia, Canada, New Zealand, and Norway.

… And in the Future, when an Array of Policy Options Need to Be on the Table

In a situation of sustained easy external financial conditions, concerns are broad, and policy objectives may conflict with each other. This chapter has argued that reducing the likelihood of undesirable booms in domestic demand and credit and their consequences is an overarching objective. The risks to stability and growth associated with boom-bust cycles in demand and credit are clear, and focusing exclusively on trying to avoid currency appreciation would miss them. Reduced growth from episodes of temporary real appreciation itself is also an important concern, although it is more difficult to confirm and evaluate. In any case, in situations where policymakers are concerned by appreciation itself, policies for controlling domestic demand growth again will be relevant, as these are also important for containing real appreciation.

Table 3.4.

FCE: System Estimation of Domestic Demand Growth, Real Appreciation, Export Price, and Real Export Growth 1/2/

article image
Sources: IMF, International Financial Statistics; Bloomberg, L.P.; Haver Analytics; and IMF staff estimates.

System estimation by seemingly unrelated regressions using annual data for 1986–2008. To present a representative number for the FCE, the parameters have been assumed to be the same for Brazil, Chile, Colombia, Mexico, and Peru. Individual country regressions reveal dif f erences in country parameters, and that not all coefficients are signif icant in all countries. The system also captures the interaction betw een the real Fed funds rate, the VIX, and demand growth in advanced economies in China. Results do not change qualitatively when the recent global f inancial crisis is excluded.

Constants vary by country. ** signif icant at the 5 percent level.

Against the wide range of concerns, an array of policies will need to be considered during times of sustained easy global financial conditions. No single policy can counter all risks. Moreover, in practice, all the relevant policy options will have their own limitations—including in terms of political and administrative feasibility—as well as in their effectiveness, and their potential for adverse side effects. No solution will be perfect, but acting with moderation on multiple fronts seems more prudent.

Recognizing that no easy fix will be available, the experience of past episodes and policy approaches suggests priorities for certain policies and some guidelines for the sequencing and conditions in which various policies should be considered.

To meet the challenges posed by episodes of easy external financial conditions, theory and experience both suggest that maintaining exchange rate flexibility is key to help contain demand and credit booms. Put another way, efforts to lean against the wind of nominal appreciation can actually exacerbate such booms, maximizing the transmission of easy external financial conditions. Still, it is difficult for even the purest exchange rate float to deliver complete insulation, and a possibly large currency appreciation can create its own concerns. For instance, sustained appreciation could lead to bankruptcies in the export and import-competing sectors, with potential long-lasting effects on economic activity.

Consideration should be given to shifting the macroeconomic policy mix—with a tighter fiscal and a relatively looser monetary stance. A firmer fiscal position can open space for a monetary stance that is looser than would otherwise be possible. Lower policy interest rates will reduce incentives for “carry trade” inflows especially. Consideration will need to be given to removing any domestic financial market distortions that may be encouraging carry trade. The case for early action on the fiscal policy front is especially strong when domestic demand is growing significantly faster than potential output and economies are approaching or already above full employment. A tighter fiscal stance is directly relevant for containing domestic demand growth and the widening of the current account deficit, and at the same time helping to reduce real appreciation. As noted, studies suggest this fiscal tool can be quite powerful in this context,40 with maximum effect when concentrated on cuts in spending on nontraded goods.41 Of course, fiscal tightening may encounter political resistance and take time to develop and to have its full impact; this argues for planning ahead and not delaying.

Prudential policies should play an important role in limiting the risks that may arise and grow during times of easy external financing. Prudential policies can help contain rapid credit growth that may facilitate a boom in domestic demand. In particular, they can provide incentives to smooth credit growth over the cycle, for example, through dynamic provisions and cyclically adjusted reserve requirements. Prudential regulations can also head off excessive risk-taking in private financial and corporate sectors’ balance sheets during good times, particularly those related to currency exposure, including through financial derivatives and off-balance sheet instruments.

Further regulatory and supervisory steps should be considered in the region to more thoroughly contain systemic risks that may develop during the upswing. This may involve widening the perimeter of prudential regulation and financial supervision, which should cover all systemically important financial institutions, whether banks or nonbanks. Most countries in the region need to advance further in strengthening consolidated supervision to better detect and correct in time systemic risks, to avoid “blind spots,” and to use supervisory methods that internalize macroeconomic dynamics. Enhancements of supervision may involve legal modifications that set guidelines for dealing with cases of systemic risk, perhaps clarifying the resolution framework that would apply to these cases, and separate capital requirements for systemically important institutions. In addition, further developments in prudential regulation and supervision in advanced economies will affect foreign banks operating in the region and set new standards that could be adapted to Latin American financial institutions, particularly as the region proceeds in the process of gradually implementing the Basle II recommendations (see Chapter 1, Box 1.1).

Enhanced policies on these fronts may take time to design and implement, as they often involve changes in legislation that must go through the political process. Laws and regulations to improve policy frameworks should be reviewed frequently, as the goal is to have policies in place before they are more obviously needed. In the meantime, there is always scope for further refinement in essential supervisory tasks, which include monitoring liquidity, market risk, and credit risks, as well as oversight over risk-taking through proprietary trading by institutions considered too big to fail.

As acknowledged earlier, appreciation of the nominal exchange rate could create problems of its own. In considering policies aimed at limiting currency appreciation directly, pragmatism calls for a combination of policies that take into account country characteristics. These include the current state of financial regulation and supervision, the effectiveness of bureaucratic procedures, the sophistication of the investor base, and the level of international reserves. The focus of policy responses in this direction should be greater if a country is already experiencing a domestic demand boom and has a fairly large current account deficit, and when other policy options have already been deployed to the extent possible.

In such a context, controls (taxes) on capital inflows could play a role, but it is important to keep in mind their limitations. Careful attention should be given to their design and the ability to enforce them. They need to be applied to the widest possible spectrum of inflows to lessen circumvention. Their implementation requires substantial administrative capabilities that cannot be developed overnight. To moderate distortions, price-based mechanisms are generally more appropriate. Because the effect of controls tends to wear off over time, and because controls may have undesired side effects, they cannot substitute for more fundamental adjustments, but can be a potentially useful transitory tool (Ostry and others, 2010).

Intervention in the foreign exchange market can be an appropriate tool as well. Because it is relatively easy to implement quickly with seemingly immediate effects on the nominal exchange rate, however, intervention is perhaps often used too quickly and too extensively. Sterilized intervention cannot substitute for action on more fundamental fronts, as it cannot by itself control growth of domestic demand or even credit. Before intervening, it should be verified first that the currency is not currently undervalued for fundamental reasons. This requires taking account of the latest developments in productivity and the terms of trade, as these may be affecting the equilibrium real exchange rate. Here, the outlook for prices of Latin America’s commodity exports is particularly relevant.

Before resorting to market intervention and implementing capital controls, it is generally advisable to allow a nonnegligible amount of currency appreciation. Deploying these tools too early risks being self-defeating, by continuing market expectations of future appreciation. Moreover, policymakers need to be wary that efforts to smooth or moderate the path of appreciation over an extended time could also backfire, by stimulating further capital inflows. Similarly, efforts to ensure adequacy of international reserves, which are clearly important, need to be carried out in ways that do not attract more inflows.

The issue of the appropriate level of international reserves to protect the country against sudden stops, external liquidity shortages, and current account reversals is in principle separate from exchange rate flexibility and foreign exchange intervention. For example, international reserves could be built by official long-term borrowing without any official intervention in the foreign exchange market (or by purchases of foreign exchange that are executed relatively smoothly, rather than timed to react to short-term market fluctuations). The appropriate level of self insurance by building reserves needs to balance two factors: the cost of holding reserves and the likelihood and severity of the possible adverse event (which may increase with the size of inflows). These factors are likely to differ across regions depending on country characteristics, which include, among others, the degree of financial dollarization in the economy (which can limit the desirability of allowing wide exchange rate flexibility).42 The need for self insurance through reserves diminishes with the reliability of other forms of insurance that allow countries to better pool these risks, including those designed through multilateral arrangements.

In conclusion, financially integrated countries in Latin America may face a sustained period of easy external financial conditions—with associated risks that are amply illustrated by past experience. In many respects, these countries are now better equipped than in the past to face these challenges, but none are immune. Economies that have already returned to full employment, amid strong demand growth, face the most immediate challenge. But policymakers in other countries also will need to look ahead and concern themselves with a broad range of issues and risks—going beyond the question of currency appreciation alone, focusing especially on demand and credit booms—even though these will not fully develop immediately. Policymakers will want to be ready to pursue multiple policies from within their toolkits.

Western Hemisphere Main Economic Indicator

article image
Source: IMF staff calculations.

End-of-period rates, i.e., December on December. These will generally differ from period average inflation rates reported in the IMF, World Economic Outlook, although both are based on identical underlying projections.

Simple averages.

Private analysts estimate that consumer price index inflation has been considerably higher. The authorities have created a board of academic advisors to assess these issues. Private analysts are also of the view that real GDP growth has been significantly lower than the official reports since the last quarter of 2008.

Fiscal year data.

The Eastern Caribbean Currency Union includes Anguilla, Antigua and Barbuda, Dominica, Grenada, Monseratt, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines.

Latin America and the Caribbean Main Fiscal Indicators 1/

article image
Source: IMF staff calculations.

Figures for overall public sector, including general government and public enterprises. Definitions of public sector accounts vary by country, depending on country-specific institutional differences, including on what constitutes the appropriate coverage from a fiscal policy perspective, as defined by IMF staff.

Simple averages.

Federal government and provinces. Includes interest payments on accrual basis.

Central government accounts only. Differences between the primary balance and revenues less primary expenditures are the result of statistical discrepancies.

General government accounts only.

Central government accounts only; reported on fiscal year basis.

Central government accounts only. Overall and primary balances include off-budget and public-private partnership activities. Revenue and expenditure components of these items are not available and not included in the revenue and primary expenditure estimates.

Central government accounts only.

Eastern Caribbean Currency Union includes Anguilla, Antigua and Barbuda, Dominica, Grenada, Montserrat, St. Kitts and Nevis, St. Lucia, and St. Vincent and the Grenadines. Central governent accounts only.