Chapter

3. Foreign Exchange Market Intervention: How Good a Defense Against Appreciation Winds?

Author(s):
International Monetary Fund. Western Hemisphere Dept.
Published Date:
May 2011
Share
  • ShareShare
Show Summary Details

Emerging markets have increasingly relied on foreign exchange intervention to confront currency appreciation pressures arising from “easy external financing conditions.”Focusing on the last seven years, this chapter discusses the experience of selected Latin American economies with these policies, and compares them with those of other regions. It also examines the role of different modalities of intervention, and offers a new cross-country approach to assess its effectiveness and costs.

Central banks intervene for various (nonexclusive) reasons, but the nature and time profile of their interventions often suggests an effort to mitigate currency appreciation pressures. When practicing intervention, a number of countries in Latin America rely on rules but these often leave significant room for discretion; it is unclear whether rule-based settings are more effective (in mitigating appreciation). Evidence suggests that interventions can slow the pace of appreciation, but such effects decrease rapidly with the degree of capital account openness; and interventions appear to be more effective when there are signs that the currency may be becoming overvalued. Costs associated with intervention often have been sizable, reflecting not only high interest rate differentials but also large valuation losses.

3.1. Introduction

Abundant liquidity in global markets and a high exposure to international capital movements have put foreign exchange intervention (FXI) at center stage of the policy debate in Latin America. Despite the widespread use of FXI policies to confront the spillover effects of easy external financing conditions—including on exchange rates—there is no guarantee of their success. The empirical literature (focused mostly on advanced economies) has failed to reach a conclusion about the effects of FXIs on exchange rates, frequently suggesting their absence. Under current global conditions favoring capital flows to emerging markets, and with added currency appreciation pressures arising from marked changes in fundamentals—for example, terms-of-trade gains for commodity exporters—the effects of FXI have become even more difficult to grasp. Still, many central banks appear to believe in the effectiveness of FXI and continue to pursue such policies (see Neely, 2008; Bank for International Settlements, 2005).

At the same time, whatever their own assessment of the effects and benefits of FXI, central banks are aware that FXIs are “no free lunch.” If aimed at preventing a necessary adjustment of the exchange rate toward equilibrium, they are likely to incentivize one-sided bets, attract further capital inflows, and induce currency mismatches. They can also carry nonnegligible quasi-fiscal costs. And in inflation-targeting frameworks, they may lead to inconsistencies with main monetary policy objectives. Against this backdrop of uncertain benefits and more apparent costs, the desirability of such policies is an open question.

This chapter takes a fresh look at the issue of FXI with a focus on Latin America. In particular, how have Latin American countries intervened, and how has this differed from other emerging market economies (EMEs)? What motives have driven such polices? How effective has FXI been in affecting the exchange rate—assuming that this has been one of the objectives—and how costly has it been?

In focusing on the possible effects of FXI on the exchange rate, this chapter puts aside the issue of whether affecting the exchange rate is desirable. It also leaves aside the issue of how FXI compares with other policy tools available to manage currency movements (and, in general, the effects of strong capital inflows). Those broader questions are discussed extensively in a number of recent papers, including recent editions of the Regional Economic Outlook—Western Hemisphere.1

The main object of study is sterilized FX purchases. This has been, by far, the more prevalent direction of intervention among the countries studied (except during the 2008–09 financial crisis, which is not analyzed in this work),2 and it is of considerable policy interest to know whether such operations could mitigate current appreciation pressures. The emphasis is on sterilized rather than unsterilized interventions because only the former entails pure exchange rate policy—the latter involves also a decision to simultaneously relax monetary policy, for which an effect on the exchange rate would seem more obvious.3

Despite an abundant literature on intervention, there is often little clarity on the precise definition of FXI. Conceptually, we consider FXI to be any operation that affects the central bank’s net foreign exchange (FX) position.4 In practice, however, high frequency data on central banks’ FX position are often unavailable, requiring the use, instead, of observable FX market transactions or changes in international reserves as proxies (Box 3.1).

The analysis draws on the experience of Latin American economies as well as relevant comparator countries, during the period 2004–10,5 although data unavailability restricts parts of the analysis (see Annex 3.1 for details on the data employed). This time span is meant to capture a period of accentuated capital flows to EMEs and FXI. The sample excludes countries with pegged exchange rates (for which the decision to intervene and the scale of intervention are not matters of policy choice, given their commitment to a peg).

The chapter is structured as follows. Sections 3.2 and 3.3 present essential stylized facts on global trends and on the modalities of intervention, both quantitative and qualitative. Section 3.4 investigates empirically the effectiveness of intervention in influencing exchange rates. Section 3.5 briefly examines the costs of intervention. The chapter concludes with a discussion of key findings and their potential policy implications.

3.2. Key Trends

Nearly all of the 2004–10 period—with the exception of the global crisis episode—was characterized by highly favorable external financing conditions for EMEs, in terms of ample liquidity and low risk aversion (as reflected, for example, in levels of the real federal funds rate, the VIX, and emerging economies sovereign spreads). During this period also, large capital inflows tended to bring heavy or accelerated FXI, particularly in the run-up to the 2008 crisis and during the postcrisis period (Figure 3.1).

Figure 3.1.EM interventions co-move, responding to shifts in global financial conditions.

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

1 U.S. dollar trade-weighted exchange rate. A decline in the index corresponds to an appreciation.

2 12-week average flows to emerging market dedicated mutual funds (in percent of assets under management).

3 International reserves minus gold. Annualized 3-month moving average, in percent of 2006–07 average GDP.

4 Includes Brazil, Chile, Colombia, Czech Republic, India, Indonesia, Israel, Hungary, Korea, Malaysia, Mexico, Peru, Philippines, Poland, Romania, Russia, South Africa, Thailand, Turkey, and Uruguay. Simple average.

5 Includes Brazil, Chile, Colombia, Mexico, Peru, and Uruguay. Simple average.

6 Includes India, Indonesia, Korea, Malaysia, Philippines, and Thailand. Simple average.

7 Includes Czech Republic, Hungary, Israel, Poland, Romania, Russia, Turkey, and South Africa. Simple average.

A glance at changes in central banks’ international reserves puts in perspective these trends, and highlights the common (asymmetric) direction of intervention during the sample period—as well as the tendency for the intensity of “leaning against the wind” in different regions to show common fluctuations over time. Still, the magnitude of intervention has varied across countries. In general, Latin American countries have intervened less than the economies of emerging Asia.

Box 3.1.Actual Intervention and Changes in Gross International Reserves

Data availability limitations usually hamper the analysis of FX intervention. Because FX intervention data are often confidential and not available, the literature usually addresses this by using the change in gross international reserves as a proxy for intervention.

However, actual intervention data and the change in gross reserves frequently differ from each other. The reason is that reserves change not only because of FX intervention, but also because of valuation changes, income flows (for example, accrual of interest), debt operations on behalf of other agents, and so on.

How good is the change in reserves as a proxy for intervention? A regression of reserves on actual intervention data, for countries with both forms of data available (Colombia, Costa Rica, Guatemala, Peru, and Uruguay) suggest that, at a daily frequency, intervention data and the reserve proxy can differ markedly.1 The regression coefficient relating these variables tends to be quite low. This is evident for highly dollarized economies, where reserves can change on account of regular liquidity operations with the domestic banking system. However, the use of reserves as a proxy for intervention improves significantly at lower frequencies. On a weekly basis, the regression coefficient is higher for most countries. This feature supports the use of weekly reserve series as a proxy in the econometric exercises.

Actual Intervention Data vs. Gross Reserves1

Daily data (left) and weekly data (right), 2004–10 (Millions of U.S. dollars)

Source: IMF staff calculations.

Note: Daily chart blue line: predicted value. Black line: 45 degree line. Regression coefficient: 0.59 with standard error 0.03 and R2 = 0.03. Weekly chart blue line: predicted value. Black line: 45 degree line. Regression coefficient: 0.75 with standard error 0.04 and R2 = 0.19.

1 Includes Colombia, Costa Rica, Guatemala, Peru, and Uruguay.

More importantly, the measurement error is unlikely to have a significant influence on the econometric estimations assessing the impact of intervention on the exchange rate. This is confirmed by the low and two-side correlation between the measurement error and the exchange rate (see econometric section for further discussion).

Note: This box was prepared by Camilo E. Tovar.1 Regression results for individual countries are not reported owing to confidentiality reasons.

Within Latin America, however, there have been noticeable differences, with Chile, Colombia, and Mexico displaying limited amounts of intervention relative to Brazil, Peru, and Uruguay, where interventions have been very large at times. Chile is noteworthy for its long “spells” without any FXI.

Examination of (monthly) data on actual intervention and exchange rates for some countries of the region shows that the widespread use of FXI has been accompanied by marked currency appreciation (Figure 3.2). This highlights the inherent difficulties in assessing the effects of intervention in the absence of an observable counterfactual, as simple correlations would misleadingly suggest that positive interventions (purchase) tend to appreciate the currency.6

Figure 3.2.Intervention has nearly always been in the same direction, but varies greatly in both frequency and intensity across countries. Intervention has been accompanied by exchange rate appreciation in many cases.

Source: IMF staff calculations on the basis of central bank data.

Note: Latin America includes Costa Rica, Guatemala, and Uruguay. Positive values of intervention refer to purchases, whereas negative values refer to sales. For the sake of completeness, both purchases and sales are depicted. Upward movements of the exchange rate correspond to depreciations. Arrows on the axis denote that the scale has been changed relative to previous and subsequent panels.

1 Intervention measured as a percentage of the average annual GDP between 2004 and 2010.

2 Some FX operations conducted by Banco de Mexico may not be considered as intervention and show how difficult it is to have a proper definition. In particular, prior to the crisis, the central bank was selling, according to an announced rule, exactly half of the increase in net reserves, which reflected Pemex and the federal government’s law-mandated transfers of their FX receipts to the central bank. The policy adopted by the foreign exchange commission was to reduce the pace of accumulation of international reserves. Actual purchases (through options) have taken place only since March 2010. Option auction data reported.

3Simple averages.

3.3. Modalities of Intervention

In general, knowledge of the manner in which central banks intervene in FX markets is limited. This is partly because many central banks withhold such information, but also because the country information that is available is dispersed, and the literature on intervention tends to focus on one country at a time. Some studies have examined intervention practices through surveys, aiming at drawing lessons on best practices (Neely, 2008, 2001; Bank for International Settlements, 2005; Shogo and others, 2006; and Canales-Kriljenko, 2003).7 Still, systematic and up-to-date cross-country information on modalities of intervention is scarce.

This section discusses key findings from our research project to characterize intervention practices across EMEs. In contrast to the previous discussion, this section relies on daily data.8 A novelty is that the quantitative data are augmented with qualitative information describing the manner in which central banks conduct intervention. The database was constructed from official central bank statements, as found in web sites, communiqués, press releases, and annual or other periodic reports.

How frequent are foreign exchange interventions? Most, but not all, countries in the region have had a fairly regular presence in the FX market (Table 3.1 and Figure 3.2). On average about one-third of the countries in the region intervened in any given day, a relatively high number considering that most of them declare themselves to be floaters. Although FXI in the region tends to come in waves—frequently corresponding with shifts in global financial conditions—there are important cross-country differences. The central banks of Brazil and Uruguay have had a frequent presence in the market—about two-thirds of the time (not reported). At the other extreme are central banks with fairly rare market presence over the 2004–10 period—Chile, Mexico, and Guatemala more recently. Even so, two central banks traditionally viewed as “noninterveners” have entered the FX market in the postcrisis period: Mexico in February 2010 and Chile in January 2011, with the announcement of reserve accumulation programs.

Table 3.1.Stylized Facts of Foreign Exchange Purchases, 2004–10
Frequency

(Percent of working

days)
Intensity
Cumulative intervention as percent of GDP1,2Daily average (Millions of U.S. dollars)1Daily maximum (Millions of U.S. dollars)1Has there been active FX intervention in 2011?
Chile63.85050yes
Colombia3210.334733yes
Guatemala191.69332yes
Mexico310.6600600yes
Peru3936.155494yes
Latin America41910.5150442
Others
Australia5622.515377n.a.
Israel2422.384300no6
Turkey6612.5614966yes
Source: IMF staff calculations on the basis of central bank and its information.Note: Some countries do not maintain an active permanent presence in the market during the full period (e.g. Chile, Israel, or Mexico).

How large have foreign exchange purchases been? A rough comparison of the relative size of interventions— scaled by GDP—shows that Chile, Guatemala, Mexico, and Colombia (in that order) are low or moderate interveners. Uruguay and Peru—highly dollarized economies—are heavy interveners (Table 3.1). Daily reserves data (as well as monthly intervention data) suggest that Brazil’s intervention has also been large at times (Figure 3.2).

What are the stated motives for intervention? The two reasons most often stated for intervening have been: i) to build international reserve buffers; and ii) to contain exchange rate volatility (in some sense, as will be discussed below—see also Box 3.2). Slowing the speed of appreciation is a motive stated only at one point in our survey, by Colombia’s central bank. “Other” reasons sometimes stated, as summarized in Figure 3.3, have included correcting misalignments, addressing disorderly market conditions, and managing liquidity in FX markets. These statements were often vague.

Figure 3.3.Reducing exchange rate volatility and building up reserves are the most often stated motives for intervention.

Source: IMF staff calculations.

1 Based on declared ex post motives for intervening as made publicly available in official central bank statements (e.g., press releases, annual reports, web site, etc.), otherwise ex ante statements of objectives are employed. Averages for the period.

2 Includes Latin America, Australia, India, Indonesia, Israel, Russia, Thailand, and Turkey.

3 Includes Brazil, Chile, Colombia, Costa Rica, Guatemala, Mexico, Peru, and Uruguay.

At some point in the sample period, most central banks declared that their intervention was to strengthen their reserves buffers, often simultaneously stating that they had no intention to influence the exchange rate (for example, Chile and Mexico).9 Other central banks (Peru, Colombia, and Guatemala) have explicitly stated to have intervened to contain excessive exchange rate volatility, but—unless there was a rule in place—thresholds to determine what excessive means were not always stated.

Not one central bank in our sample declared targeting an exchange rate level as a motive for intervention. It is noteworthy that a 2005 BIS survey reported that a significant share of central banks in emerging markets intervene to influence the exchange rate level or to “lean-against-the-appreciation-wind” (BIS, 2005). Although this could suggest a tension between declared and actual motives, it may also reflect that stated objectives are often not precisely defined. For example, “influencing” the exchange rate is somewhat ambiguous, as it could refer to its level, its appreciation rate, or its high or low frequency volatility. Importantly, “leaning against the wind” need not mean targeting a particular level of the exchange rate, and could be interpreted as seeking to reduce (low frequency) exchange rate volatility, in the sense of dampening a perceived cycle of temporary excessive appreciation.

Do central banks’ intervention frameworks favor rules or discretion? On average about one-third of the central banks had in place some form of rule-based intervention framework at any moment within our sample period (Figure 3.4). In Latin America, the share of countries with such a framework was somewhat higher (almost half). However, rules are not uniform in nature. FXI rules can be categorized into two main groups: “exchange rate-based” and “quantity-based.”

Figure 3.4.Some central banks rely on rules, particularly in Latin America

Source: IMF staff calculations.

1 Declared intervention rules according to official central bank statements (e.g., press releases, annual reports, web site, etc.). Exchange rate-based rules are triggered by some exchange rate-related measure (e.g., change or volatility). If the amount of intervention is specified then it is considered to be “with amount limits”; otherwise it is considered “with no amounts limits.” Quantity-based rules specify an amount to be exercised over a horizon along with the specific daily or weekly quantities. Averages for the period.

2 Rules using options are categorized as exchange rate-based (with amount limits) because it is the exchange rate that triggers the actual purchase of FX (that is, the option is exercised).

3 Includes Latin America, Australia, India, Indonesia, Israel, Russia, Thailand, and Turkey.

4 Includes Brazil, Chile, Colombia, Costa Rica, Guatemala, Mexico, Peru, and Uruguay.

Box 3.2.Foreign Exchange Intervention Rules in Practice: Selected Latin American Experiences

When applying rules for conducting foreign exchange purchases, Latin American central banks have relied on two main types: i) quantity-based rules; and ii) exchange rate-based rules.

• Quantity-based rules

These rules usually announce a window of time over which the central bank will be accumulating reserves along with a quantity of FX purchases. A recent example is that of the Central Bank of Chile, which announced on January 3, 2011, a program to accumulate US$ 12 billion over the course of 2011 through the daily purchase of US$ 50 million reserves in competitive auctions. The central bank explained that the purpose of the program was to accumulate reserves.

• Exchange rate-based rules

These rules announce conditions under which the behavior of the exchange rate can trigger central bank purchases of foreign exchange.

The central banks of Colombia and Guatemala have employed exchange rate-based rules. These rules specify a threshold determined by a moving average of the exchange rate, and the amount of intervention either in cash or instruments. In Colombia, the rule (operational from 1999 through October 2009) authorized the central bank to auction US$ 180 million in put options (granting holders the right to sell dollars to the central bank) whenever the exchange rate fell 5 percent below the average exchange rate of the previous 20 working days.1 In Guatemala, the central bank started to rely on a similar rule starting in 2005. In 2010, the rule established FX purchase (or sale) auctions of up to US$ 32 million per day whenever the exchange rate fell below its average of the previous 5 days plus a tolerance band of 0.6 percent. Both central banks stated that the purpose of these rules was to control the volatility of the exchange rate.

Mexico is a current example of an exchange rate-based rule adopted for the purpose of reserve accumulation. On February 22, 2010, the authorities announced the use of the put option mechanism as a means to accumulate reserves.2 The mechanism auctions US$600 million on a monthly basis in put options granting holders the right to sell dollars to the central bank, with a strike price equal to the previous day’s interbank reference rate (FIX), as long as that rate is below its previous 20-day average.

Note: This box was prepared by Camilo E. Tovar.1 This rule was replaced by a direct intervention mechanism through auctions. See the central bank’s web site for current regulations. See also Rincón and Toro (2010), Echavarría and others (2009), and Uribe and Toro (2005) for a detailed account of these rules in Colombia.2 This mechanism was also used by Banco de México between 1996 and 2001. See Sidaoui (2005).

Exchange rate-based rules are those announcing that intervention will be triggered by some exchange rate-related measure (for example, change, or volatility). The volatility-triggered rules in Colombia and Guatemala are examples of this (see Box 3.2). Some trigger rules have specified the amount of FX purchases, but others have not.

On the other hand, quantity-based rules do not specify any trigger for intervention, but do specify an intervention amount to be exercised over an announced time horizon (along with the daily or weekly intervention quantities). During the period studied, these rules were associated mainly with reserve accumulation as their stated motive, such as those recently employed by Chile.

This taxonomy, based on declared intervention frameworks, irrespective of whether FX operations were actually undertaken under those frameworks, reveals that about half of the rule-based frameworks in the sample were quantity based. In Latin America, there has been a preference for exchange rate-based rules—in particular those with amount limits (Figure 3.4).

A different question is what approach has been chosen at times when interventions have actually been conducted? To answer this question, we examine the use of rules or discretion, conditional on being in the FX market (Figure 3.5). When they did intervene, Chile and Mexico always used rules. Colombia and Guatemala also relied on rules—with certain objectives in mind—but at the same time gave themselves room for discretionary purchases. Brazil, Peru, and Uruguay did not use rules.

Figure 3.5.Despite the popularity of rules, most central banks have left room for discretion.

Source: IMF staff calculations.

1 Declared intervention rules according to official central bank statements (e.g., press releases, annual reports, web site, etc.). Exchange rate-based rules are triggered by some exchange rate-related measure (e.g., change or volatility). If the amount of intervention is specified then it is considered “with amount limits”; otherwise it is considered “with no amount limits.” Quantity-based rules specify an amount to be exercised over a certain time horizon along with the daily or weekly quantities of intervention. Averages for the period.

2 1 = always and 0 = never. Intensity refers to the proportion of days with FX purchases in which a specific rule is declared to be in place by the central bank.

3 Rules using options are categorized as exchange rate-based because it is the exchange rate that triggers the actual purchase of FX (that is, the option is exercised).

In which market does intervention take place? The dominant market across regions is the spot market (Figure 3.6), possibly reflecting a higher degree of liquidity vis-à-vis other markets. As derivative markets have expanded over time, however, some central banks have increased the use of such instruments (Figure 3.7). In the region, Brazil is the main example, with operations in the forward and swap markets.10 Two other central banks in the region (Colombia and Mexico) have used options. The rest have intervened only in the spot market (see Box 3.3 for a discussion on considerations for the choice of different instruments).

Figure 3.6.Intervention mainly takes place in the spot market.

Source: IMF staff calculations.

1 Declared intervention instruments according to official central bank statements (e.g., press releases, annual reports, web site, etc.). More than one instrument may be used for intervention by a single central bank, thus totals do not add to 100. Averages for the period.

2 Includes Latin America, Australia, India, Indonesia, Israel, Russia, Thailand, and Turkey.

3 Includes Brazil, Chile, Colombia, Costa Rica, Guatemala, Mexico, Peru, and Uruguay.

Figure 3.7.The most liquid market segment is the spot market, but derivative markets are catching up

Source: Bank for International Settlements.

1 According to Bank for International Settlements’ definitions.

2 Includes Brazil, Chile, Colombia, Mexico, and Peru.

3 Includes India, Indonesia, Israel, Russia, Thailand, and Turkey.

How transparent are central banks about intervention? Around the world, most EMEs refrain from publishing information about their FXI operations (or reserve stocks on a high frequency basis, from which FXI might be inferred). This is precisely what restricts the country sample in our analysis in several parts of this chapter, including here. Latin America is among the most transparent regions, with a level of transparency that has increased over the past seven years, particularly in comparison with other regions of the world. Latin American central banks tend to publish information earlier compared with other countries that publish (Figure 3.8).

Figure 3.8.Latin American central banks reveal their intervention, and more quickly than other central banks

Source: IMF staff calculations.

1 Disclosures according to official central bank statements (e.g., press releases, annual reports, web site, etc.). In certain cases, it was unclear when information was disclosed. Thus totals may not add to 100. Averages for the period.

2 Includes Latin America, India, Indonesia, Israel, Russia, Thailand, and Turkey.

3 Includes Brazil, Chile, Colombia, Costa Rica, Guatemala, Mexico, Peru, and Uruguay.

Box 3.3.Instruments for Foreign Exchange Purchases

Central banks have a range of instruments with which they might directly influence the exchange rate, including FX spot purchases, forwards, swaps, and options.1

  • FX spot purchases are transactions made by the central bank for “immediate” delivery.

  • Forward FX purchases entail a future purchase of FX at a preagreed exchange rate. These can be deliverable or nondeliverable.

  • Cross-currency swaps involve the simultaneous purchase and sale of one currency for another at two different dates. Interventions with this instrument are composed of two legs: (i) a spot FX purchase, reversed by (ii) a future FX sale at the spot exchange rate at that time.2

  • FX put options are contracts that give the holder the right to sell foreign exchange to the central bank under certain contingent conditions (see also Box 3.2).

The spot market is the most developed market in the region, and central banks have traditionally considered it as the natural market for interventions (see Figures 3.6 and 3.7).

Although forwards have been used only occasionally in the region, there is a long history of use of options (by Colombia and Mexico). Cross-currency swaps have been used only by Brazil (cupom cambial).3

A number of considerations can influence the choice of instruments.4 For instance, (i) the use of derivatives reduces the degree of transparency of central bank operations vis-à-vis spot transactions, thus weakening the signaling channel (although this can be partially addressed by a clear communication policy); (ii) they obscure the central bank’s balance sheet FX position; (iii) although normally they do not require immediate sterilization (except for some cross-currency swaps) thus helping mitigate ex ante the quasi-fiscal costs of interventions, their use exposes the central bank to the risk of a sudden capital loss, if interventions fail to contain appreciation pressures; and (iv) derivatives carry counterparty and liquidity risk, which can be particularly pronounced in thin markets. On the other hand, (i) put options offer the additional benefit of working as automatic stabilizers of the exchange rate, as they are exercised only under conditions of appreciation pressures; and (ii) derivatives can be settled in local currency, and do not necessarily entail the use of reserves at any point in the contract. This can be a desirable feature for central banks that prefer to avoid the potentially negative signaling associated with fluctuations in the level of reserves. Relatedly, the unwinding of derivative positions, once appreciation pressures have receded, seems easier than the unwinding of the reserve accumulation that would result from spot transactions.

Note: This box was prepared by Gustavo Adler and Camilo E. Tovar.1 Other policy instruments, not discussed here (for example, reserve requirements, interest rates), may also influence the exchange rate, but in a less direct manner, and are normally not used with this objective in mind.2 Cross-currency swaps are different from regular currency (FX) swaps. The latter—often issued for liquidity management, rather than FX intervention—entails a forward leg that is settled at a preagreed exchange rate, thus eliminating exchange rate risk. A cross-currency swap, on the other hand, carries exchange rate risk, as the forward leg is settled at the spot rate prevailing at the end of the contract, thus changing the FX position of the central bank and its counterparty.3 The cupom cambial is a derivative equivalent to a cross-currency swap that pays the difference between the local interest rate and changes in the real/U.S. dollar exchange rate. Although originally the central bank took the long real-open interest rate, it has recently switched to take the short real-interest rate position to dampen appreciation pressures.4 See also Canales-Kriljenko and others, 2003; Shogo and others, 2006; and Blejer and Schumacher, 2000.

3.4. Searching for Effects of Intervention: New Evidence

We investigate the effect of FX intervention on the exchange rate using two complementary approaches: (i) a panel regression analysis; and (ii) an event study. As a by-product, the first approach also offers empirical insights into the motivations or triggers for FX intervention.

An Econometric Panel Approach

A critical problem in assessing the effectiveness of intervention is overcoming the endogeneity of changes in exchange rates and intervention, as the latter tends to react to the former (see Boxes 3.4 and 3.5 for a discussion on transmission channels and a brief literature review). The econometric approach presented in this section addresses this issue by focusing on episodes of significant global shocks leading to appreciation pressures in emerging markets and estimating the effect of intervention 11 in a panel setting that exploits the heterogeneous reaction of different central banks to such shocks. As in other studies, a two-stage estimation procedure is used, with the first stage estimating a country-specific reaction function that allows for different behavior across countries. Predicted values of the reaction function are used as instruments for the second stage, which entails estimating a behavioral equation linking the exchange rate to intervention, in the panel setting. The approach allows identification of the effect of intervention, mitigating the endogeneity problem by focusing on short time spans during which unobservable country-specific shocks are less likely to be large (at least in relation to the global shock, on the basis of which the episode is identified).

Box 3.4.Channels of Transmission of Foreign Exchange Intervention to the Exchange Rate

The extent to which FX intervention can affect the exchange rate is not, a priori, obvious. Any incipient effect that might begin to move the currency away from its equilibrium value (that is, implied by fundamentals or market perceptions of these) should be arbitraged away by private agents. Thus, some form of market friction is necessary for sterilized1 intervention to have an impact on the exchange rate. Three main forms of market friction (channels of transmissions) have been identified by the literature:2

A portfolio balance channel operates when domestic and foreign assets are not perfect substitutes, and the risk premium increases with the supply of domestic assets. FX interventions precisely expand the amount of domestic assets (either high-powered money or sterilization instruments) potentially raising the risk premium and, by arbitrage, depreciating the currency.

An informational/signaling channel. Through its FX interventions, the central bank can potentially signal future policy intentions. For example, it could indicate its willingness to adjust its monetary stance (that is, reduce policy rates) to prevent further appreciation of its currency. Prospects of a lower interest rate would normally lead to a spot-market depreciation. Sterilization with interest-bearing instruments can reinforce this channel by increasing the financial gains of reducing interest rates. Interventions (or even simple “open mouth” operations) can also help to coordinate market expectations about the appropriate level of the exchange rate, if market participants believe the central bank has an informational advantage in this regard.

A microstructure channel. Some studies have argued that the structure of the FX market can play a role in determining the effectiveness of interventions, as frictions at a microeconomic level can affect the extent to which information embedded in central bank operations (assuming an informational advantage exists) reaches market participants and shapes their expectations.

Whether, or to what extent, these channels operate in practice remains an open question in the literature, as the empirical evidence on the effectiveness of intervention, let alone its channels, remains inconclusive.

Note: This box was prepared by Gustavo Adler.1 This discussion refers to sterilized interventions only, as the effect of unsterilized operations is arguably more straightforward, because the expansion of the money supply (beyond monetary growth consistent with inflation targets) would normally lead to a loss of value of the domestic currency, including through a depreciation. In other words, unsterilized interventions can be thought of as two distinct policies applied at the same time: a loosening of the monetary stance together with (or by means of) FX intervention.2 See BIS (2005); Shogo and others (2006); and Disyatat and Galati (2007) for a more detailed discussion.

The analysis focuses on 15 countries for which weekly data on intervention or international reserves are available. Countries include Australia, Brazil, Chile, Colombia, Costa Rica, Guatemala, India, Indonesia, Israel, Mexico, Peru, Russia, Thailand, Turkey, and Uruguay.

Triggers for Intervention

The first step to analyze the effectiveness of intervention—and as a by-product provide some insights into the motives for intervention—entails estimating individual central bank reaction functions for 11 countries in the sample.12 The model seeks to capture the role of various possible behavioral triggers that may lead central banks to intervene, or intervene more intensely, in the FX market.13 (Details on the specification and the methodology are discussed in Annex 3.2.) In particular, the model looks at the sensitivity of intervention (scaled by GDP) to:

(i) short-term (1-week) exchange rate movements;

(ii) the 30-day speed of appreciation, meant to capture more persistent pressures;14

(iii) the level of the real exchange rate (relative to an estimated equilibrium value);

(iv) the volatility of the exchange rate in the last week;15 and

(v) two measures of reserve adequacy (ratios of reserves to short-term external debt and to M2), meant to capture possible precautionary motives for FXI.16

Box 3.5.Overview of Recent Empirical Studies on Foreign Exchange Intervention in Latin America

There is a fairly large literature on foreign exchange (FX) intervention covering two questions: (i) the motives for intervention, and (ii) the effectiveness of intervention.1 Much of the work has focused on advanced economies.1 This box briefly reviews these main areas of research while highlighting relevant work for Latin America.

Studies Examining the Motives behind FX Intervention

This strand of literature builds upon surveys, drawing conclusions from declared practices (Neeley, 2008; BIS, 2005; and Canales-Kriljenko, 2003) or alternatively they infer them by estimating a FX intervention reaction function (see Sarno and Taylor, 2001). Few studies have estimated FX intervention reaction functions for Latin America. Evidence for Colombia suggests that the central bank’s main objective is to slow down the appreciation of the currency, and in a specific episode the central bank was responding to higher frequency exchange rate volatility (Kamil, 2008). In the case of Peru, it has been found that FX policies actually respond to exchange rate volatility as claimed by the central bank (Humala and Rodriguez, 2009). Specifically, the finding is that average FX purchases are eight times smaller when the exchange rate volatility is low. Possibly the most comprehensive study is that of Gonzalez (2009), who uses actual daily intervention data to examine the motives of intervention in Brazil, Colombia, Peru, and Uruguay. She finds that Brazil aims at preventing sharp appreciations; Colombia intervenes to prevent deviations of the exchange rate from a trend; Peru reacts to a number of measures of exchange rate pressures; and Uruguay aims at dampening volatility.

Studies on the Effectiveness of Intervention

Another strand of research studies the effectiveness of intervention by estimating a behavioral exchange rate equation that controls for some measure of intervention. The main challenge of this approach is to address the endogeneity bias (for example, see Kearns and Rigobon, 2005). Three main solutions have been proposed: (i) to lag the intervention policy measure. Unfortunately, under efficient markets, this can seriously distort the results; (ii) an alternative is to use simultaneous equations or a two-stage estimation approach modeling simultaneously the exchange rate and the intervention reaction function; finally, (iii) a more consistent method is an instrumental variable approach. The common practice is to use the predicted value of intervention obtained from the estimation of an intervention reaction function as an instrument (as done in this chapter).

A number of studies for the region focus on the effectiveness of FX intervention (for example, Tapia and Tokman, 2004, for Chile; or Rincón and Toro, 2010; and Kamil, 2008, for Colombia). They use actual daily intervention data and find some impact of intervention. However, these effects are usually not direct. For instance, in analyzing the Chilean experience with FX intervention early in the decade, Tapia and Tokman (2004) find that intervention has no effect on the level or change of the exchange rate. However, policy announcements do play a statistically significant role, inducing a currency appreciation ranging between 1.5 percent and 3 percent. A recent study for Colombia finds that interventions have no effect on the level of the exchange rate but can increase the volatility of the exchange rate (Rincón and Toro, 2010). However, when complemented with capital controls, FX interventions are found not only to have a level effect on the exchange rate but also to have no adverse effect on the volatility of the exchange rate. A key aspect is that these effects are short lived and their economic significance varies over time, becoming more important after 2008. A different study for Colombia has also found that the effectiveness of FX intervention may also depend on its consistency with the monetary policy stance (Kamil, 2008). In particular, it is shown that FX purchases can be effective in depreciating the exchange rate if monetary policy is in its easing cycle (a US$ 30 million FXI induces a 0.23 percent daily depreciation), but are ineffective when monetary policy is being tightened. The intuition is related to the impact of interest rate on capital flows: higher interest rates stimulate capital inflows, creating one-sided bets. Thus FX interventions must be consistent with market expectation about monetary policy.

Note: This box was prepared by Camilo E. Tovar.1 Few studies examine emerging market economies’ experience using actual intervention data. See Disyatat and Galati (2007) for an overview. See IMF (2007) for a cross-country analysis for Asia; and Dominguez, Fatum, and Vacek (2010) or Melvin, Menkhoff, and Schmeling (2009) for recent studies applied to transition economies. Finally, Jara and others (2009) present a discussion of intervention during the crisis period.

Results suggest that central banks have intervened for a number of reasons (Figure 3.9).17 In particular, while many of them appear to have intervened on concerns over exchange rate misalignments—the main exceptions being Costa Rica, Uruguay, and Russia—few countries have responded to the speed of appreciation (Colombia, Costa Rica, and Russia). Several, within and outside the region, appear to have been quite reactive to short-term (1-week) appreciation movements. Within the region, Peru has shown a high sensitivity to such short-term movements, followed at a considerable distance by Colombia. At the same time, there is scant evidence that within-week volatility has triggered intervention (with the one exception of Brazil).18 Evidence of precautionary motives is weak (with some coefficients taking opposite signs).19 In general—and possibly by construction—estimated reaction functions track trends relatively well, but do a poorer job of explaining high frequency spikes, often observed in the data.20

Figure 3.9.Intervention seems to respond to various triggers, including some linked to the exchange rate.

Source: IMF staff calculations.

1 Results of a Tobit model estimated for each country individually, on the basis of nonoverlapping weekly data, over the period for which either intervention or reserves data are available at least on a weekly frequency. Results should be interpreted as reflecting “average” preferences over the sample period 2004–10. As such, they may not reflect current preferences or objectives. See further details in Annex 3.2.

2 Lagged (U.S. dollar bilateral) exchange rate appreciation rate.

3 Deviation of the real effective exchange rate from the estimated equilibrium value, based on the history of the assessments of the Consultative Group on Exchange Rates (CGER). For Costa Rica, Guatemala, Peru, and Uruguay, a measure of deviation of the real effective exchange rate (REER) from its 5-year moving average is used, as CGER data are unavailable.

4 30-day appreciation rate.

5 1-week volatility.

6 Reserves in percent of external short-term debt on a residual maturity basis (relative to other EMEs in the sample).

7 Reserves in percent of M2 (relative to other EMEs in the sample).

Effects of Intervention

To assess the effects of intervention, we estimate an equation linking movements in the exchange rate to central bank intervention.21 To mitigate the endogeneity problem, the intervention variable is instrumented using the “shadow” intervention value obtained from the predicted values of the previous exercise.22 The exchange rate equation incorporates a number of controls (interest rate differential, sovereign spreads, commodity price shocks, and the U.S. trade-weighted exchange rate), allowing for country-specific effects in a number of them. It is estimated in first and second differences in order to study possible effects on the rate (speed) and “pace” (acceleration) of appreciation. The methodology is discussed in depth in Annex 3.2.

The model is estimated in a panel setting, with a sample of 15 countries over six common 12-week episodes of interest (simultaneous to all countries); and 12 weekly observations per episode and country. The episodes are identified by apparent shifts in global financial conditions: sharp declines in the U.S. dollar’s trade-weighted exchange rate (DXY) that take the index at least 1 standard deviation below its (Hodrick-Prescott filtered) trend (Figure 3.10). This measure is a good proxy for risk appetite (similar to the VIX) and consequently identifies episodes that coincide roughly with periods when flows into emerging market asset funds were fairly high, in relation to recent past values, or were rising strongly. As expected, these episodes have been associated with stronger appreciation of the emerging market currencies in our sample (Figure 3.11). There is also evidence that countries responded with more FXI in these episodes, but the pattern is somewhat mixed.23

Figure 3.10.We focus on six “shock” episodes of U.S. dollar weakening, coinciding with surges in flows to emerging market asset funds …

Sources: Bloomberg, L.P.; Haver Analytics; and IMF staff calculations.

1 U.S. trade-weighted exchange rate, index 2000 = 100.

2 Previous 12-week moving average, in percent of assets under management.

Figure 3.11.. . . . these global shocks were accompanied by appreciation and increased intervention

Source: IMF staff calculations.

1 Episodes of global shocks indentified on the basis of movements in the U.S. trade-weighted exchange rate (DXY).

2 Local currency per U.S. dollar, Index t0 = 100.

3 In percent of GDP.

The econometric results do not detect an immediate impact of interventions on the rate of appreciation, but do find statistically significant effects on the “pace” (acceleration) of appreciation (Table 3.2). The coefficient point estimates suggest that an additional 0.1 percent of GDP in FXI (about the size of the average weekly intervention during the identified episodes) would deliver in that week a 0.3 percent slowdown in the pace of appreciation (relative to a country that is not intervening).24

Table 3.2.Effectiveness of Intervention1
Base Model (without controls)2Base Model (with controls)3
Dependent Variable
Appreciation4Pace of Appreciation5Appreciation4Pace of Appreciation5
Sample of countries:All countries
RegressorsIIIIIIIV
Interest rate differential6
First difference0.24 *0.35 *
(1.73)(1.77)
Country spread7
First difference-0.14 ***-0.14 ***
(6.41)(4.36)
Intervention
Amount80.16-2.78 ***0.08-2.86 ***
(0.30)(-3.83)(0.16)(4.05)
R2
Within0.000.010.200.15
Between0.100.020.240.04
Overall0.000.010.200.12
Number of observations10241024964964
Number of countries15151515
Probability > F0.76780.76190.00000.0000
Source: IMF staff calculations.

A look at the effects of various modalities of intervention (Table 3.3) offers a number of additional insights:

  • Amounts of intervention appear to matter more than the mere presence of the central bank in the FX market (column I). This result could suggest either that the signaling channel is weak or that small interventions may not be enough to signal policy intentions.

  • The regressions do not find evidence that effectiveness of interventions depends on whether they are conducted under rule-based (including with preannounced amounts) or discretionary settings (columns II and III).25

  • Transparency of FX operations (measured by whether intervention data are made publicly available within a week of the operations) seems to weaken the effect on the exchange rate (column IV); however, this result seems to reflect other country characteristics that are correlated with transparency, as discussed below.

  • The effectiveness of interventions greatly depends on the degree of the country’s financial integration with the rest of the world, as captured by the interaction with the Chinn-Ito index of capital account openness26 (column V): greater financial integration seems to reduce the effectiveness of intervention. Interestingly, when we control for financial integration (column VI), the dummy on transparency loses significance, suggesting that there is high correlation between the degree of openness and the transparency of intervention operations. Still, the point estimate for capital account openness remains large, while the estimate for transparency decreases markedly.

  • A breakdown by region points to significantly higher effects in Asia than in Latin America, which are consistent with a higher degree of financial integration in the latter (columns VII–IX).

  • Interventions are more effective when there are signs that the currency may be becoming overvalued (more precisely, when it already has appreciated significantly relative to its recent history). This result is particularly pronounced in Latin America (columns X–XII).

Table 3.3.Factors Affecting the Effectiveness of Intervention1
Modalities of InterventionFinancial IntegrationRegional ComparisonsExchange Rate Misalignment
Dependent Variable
Pace of Appreciation2
Sample of countries:AllEM LAEM AsiaOther EMEsAllEM LAEM Asia
RegressorsIIIIIIIVVVIVIIVIIIIXXXIXII
Interest rate differential3
First difference0.36 *0.35 *0.35 *0.37 *0.38 *0.37 *0.161.31 **0.690.330.261.35 **
(1.79)(1.74)(1.77)(1.86)(1.89)(1.85)(0.67)(2.48)(1.44)(1.63)(0.80)(2.56)
Country spread4
First difference-0.14 ***-0.14 ***-0.14 ***-0.13 ***-0.13 ***-0.13-0.27 ***-0.03-0.35 ***-0.14 ***-0.33 ***-0.03
(4.32)(4.36)(4.36)(4.31)(-4.24)(-4.23)(-5.06)(-1.04)(-2.84)(4.42)(5.84)(-1.09)
Intervention
Amount5-2.98 ***-4.13 **-2.86 ***-4.86 ***-9.00 ***-9.44-1.81 **-7.91 ***-2.82-2.13 ***-1.52 *-7.16 ***
(3.88)(-2.53)(4.02)(4.60)(-4.41)(3.50)(-2.14)(-5.44)(-1.60)(-3.18)(-1.87)(-4.54)
Dummy of intervention60.06
(0.37)
Interaction with dummies of modalities
Dummy of discretionary setting71.34
(0.86)
Rules with preannounced amounts8-0.02
(0.00)
Transparency93.05 **-0.89
(2.54)(-0.25)
Interaction with dummy of capital account openess107.74 ***9.07
(3.20)(1.57)
Interaction with dummy of REER misalignment11-1.58 *-6.53 ***-2.24
(-1.72)(2.54)(-1.23)
R20.150.150.150.150.160.160.130.310.160.150.170.31
Within0.040.030.040.030.050.050.000.870.100.090.000.86
Between0.120.120.120.110.110.120.110.210.160.110.150.21
Overall
Number of observations964964964964964964549175216964481175
Number of countries1515151515158331573
Probability > F0.00000.00000.00000.00000.00000.00000.00030.00000.00170.00000.00000.0000
Source: IMF staff calculations.

Effectiveness—Event Analysis of Regime Shifts

Announcements of FX policy regime changes provide another opportunity to detect the effect of intervention on exchange rates, by looking at the behavior of the exchange rate within short intervals following the policy announcement.27 The idea is to avoid the drawbacks of traditional econometric estimations, which can be blurred by endogeneity problems.28 Moreover, if markets are forward looking, and if most interventions do not come entirely as a surprise to markets, then traditional estimations may fail to detect the true effect of intervention. Focusing on instances of (unexpected) policy announcements avoids that problem. With this in mind, we focus on policy shifts that entailed a sizable increase in ongoing amounts of intervention, or a clear and farpreceded change in FX policies, because marginal or recurrent policy changes are unlikely to produce significant impacts.

A first comparison of the behavior of the exchange rate against those of a peer country group around recent events of significant policy shifts suggests that these can have a noticeable impact on the currency, either depreciating it or leading to a shift in its trend. Examples of this are (Figure 3.12):

  • The announcement of a reserve-building program of FX purchases by the central bank of Chile, in April 2008—after a prolonged period with no presence in the FX market.29

  • Colombia’s announcement of a new program of discretionary FX purchases, in June 2008.

  • Other identified cases of significant policy shifts, within and outside the region, show similar patterns. Mexico’s new program of FX option auction, launched in February 2010, seems to be an exception (possibly reflecting the choice of put options as main instrument—not shown).

  • On the other hand, marginal and frequent policy changes (not shown) do not have a discernible economic impact on the exchange rate.

Figure 3.12.Events of policy shifts suggest that intervention can have a noticeable impact on the currency.

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

1 Synthetic currency refers to the predicted value of the exchange rate from an OLS-estimated econometric model regressing the country’s own exchange rate against the exchange rates of the other emerging market/advanced market countries in the sample (based on daily data). See Annex 3.3 for a test of statistical significance.

2 In November, 2010, FX purchases were increased to US$75 million (and to US$100 million for the months of November-December 2010).

The analysis is refined by comparing the country’s own exchange rate to a counterfactual or “synthetic” currency—a weighted average of other currencies—with weights resulting from a regression of the country’s own exchange rate against a set of exchange rates of other emerging market/advanced economies.30 This approach helps to find a more informative “synthetic” comparator—than the average of the neighbors used above—thus providing an arguably better counterfactual. It also allows to capture the information embedded in other exchange rates regarding unobservable global shocks that could affect the country under study following the event, leading to incorrect conclusions about the effects of the policy shift. The synthetic currency model is estimated with ordinary least squares for each event over a time span of 180 days prior to it. The estimated coefficients as well as the observed exchange rates of other countries are used to predict (conditional forecast) the country’s own exchange rate following the event.31 The difference between the predicted and the actual exchange rate is interpreted as the effect of the policy change, although the horizon at which this should be assessed remains an open issue.

The results of this comparison (also shown in Figure 3.12) confirm that currencies tend to depreciate (that is, relative to their predicted values) following regime change announcements that introduce new FXI. Annex 3.3 describes a bootstrapping exercise that confirms the statistical significance of the results.

3.5. Quasi-Fiscal Costs of FXI

To ensure consistency with other objectives of monetary policy, central banks that undertook large FX interventions sterilized them by issuing significant amounts of domestic interest-bearing liabilities (as indicated by reductions in their net domestic asset (NDA) positions).32 This was particularly marked in emerging Asia, reflecting a higher degree of intervention during the sample period. Within Latin America, Brazil, Peru, and Uruguay stand out for the sizable expansion of their central banks’ domestic liabilities (Figure 3.13).33 In contrast, FX interventions were smaller in Chile, Colombia, and Mexico, and so the expansions of the money supply that arose from them were, for the most part, able to fit inside monetary expansion paths consistent with inflation targets.

Figure 3.13.Central banks’ balance sheets expanded rapidly along with heavy intervention.

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

1 In percent of 2006–07 average GDP. NFA figures are based on local currency reported items, adjusted for exchange rate effects. In the cases of Peru and Uruguay, NFA and NDA figures may overstate the degree of intervention and sterilization, respectively, to the extent that they include operations between domestic financial institutions and the central bank in foreign currency (banks increasing holdings of FX-denominated assets at the central bank). Arrows on the axis denote that the scale has been changed relative to previous panels.

2 Line denotes NDA adjusted for increases in reserve requirements during 2010 meant to partially replace repo operations as an instrument of sterilization.

3 Monetary base as defined by International Financial Statistics (includes reserve requirements).

In most cases, central bank debt and deposits from the banking system were the main instruments for sterilization, with Brazil and Colombia being exceptions (as these central banks undertake open market operations with the Treasury’s debt instruments). Rapid increases in such interest-bearing liabilities are one clear sign of the quasi-fiscal costs associated with FX intervention; looking at the interest paid on those liabilities would capture the “book” costs of intervention. We estimate a broader measure of the economic (opportunity) cost of intervention as the deviation from interest rate parity multiplied by the changes in net foreign assets arising from FXIs. This measure may exceed the book costs of FXI.34

Our measure reflects the drain from sterilizing FX operations with domestic currency denominated liabilities, net of the gains resulting from higher holdings of FX assets of similar maturity. Ex ante deviations from covered interest rate parity—which in normal circumstances mainly reflect country risk premiums—have been relatively low in most cases. Thus, our discussion focuses on ex post costs.35

Estimated ex post costs have been significant in the region and across EMEs, as deviations from interest rate parity have been substantial, not only because of high interest rate differentials but also because of sizable exchange rate appreciation over this period (only partially offset by the depreciation that followed the collapse of Lehman Brothers in 2008). Latin America stands out primarily for its higher extent of appreciation during this period, although interest rate differentials have also been particularly high in Brazil, indeed the highest in the sample (Figure 3.14).

Figure 3.14.Deviations from interest rate parity have been substantial.

Sources: Haver Analytics; IMF, International Financial Statistics; and IMF staff calculations.

1 Sizes of bubbles are proportional to the NFA accumulation during the period (in percent of 2006–07 average GDP).

2 Average annual nominal appreciation rate, against U. S. dollar.

3 Average interest rate differential (between domestic policy—or interbank—rate and U.S. federal funds rate).

The combination of early sizable interventions— leading to substantial net FX positions at central banks in the region—and the exchange rate appreciations that eventually followed such interventions, implied marked capital losses. This highlights the costs of intervening “too early” to fight appreciation (that is, when the likelihood that the currency will appreciate further, notwithstanding resistance from the central bank, is high). For the group of LA6 countries, the cost of interventions that took place during the period 2004–10 (that is, excluding operations in previous years) averaged about ½ percent of GDP. Within the region, however, there are marked differences, with estimates for Brazil, Peru, and Uruguay significantly higher than for Chile, Colombia, and Mexico (Figure 3.15).36

Figure 3.15.The costs of intervention since 2004 are sizable in Brazil, Peru, and Uruguay.

Sources: Haver Analytics; IMF, International Financial Statistics; and IMF staff calculations.

1 Estimated cumulative costs since 2004. In percent of 2006–07 average GDP. Simple averages per group. Bars denote economic costs (interest rate differential plus exchange rate appreciation) on the cumulative stock of NFA since 2004 (decomposed by vintage year of NFA accumulation). Dotted line denotes the average for the 2004–10 period.

3.6. Takeaways and Policy Considerations

Over the past seven years, many central banks in Latin America have had a regular, and at times large, presence in FX markets. In most instances, this intervention was in one direction only, and coincided with easing of global financial conditions that put appreciation pressures on many emerging market currencies. Although central banks have stated various (nonexclusive) motives for their interventions, their nature and timing often suggest an effort to mitigate currency appreciation pressures. Whether these efforts have been successful is an empirical question that is inherently difficult to answer—precisely because intervention often takes place at the same time that other forces are acting to strengthen the currency. Our panel regression analysis is able to detect an effect of intervention on the pace of exchange rate appreciation. This effect turns out to be smaller where there is a greater degree of capital account openness, and larger when the currency already has appreciated substantially (a situation in which the currency is less likely to be undervalued). Clear effects are also found in cases of significant FX policy “regime change” (for example, when a large program of FX purchases is first announced) although these appear to fade away when policy shifts become too frequent.

At the same time, estimates of quasi-fiscal costs point to sizable losses associated with intervention policies. In practice, much of these losses have come from valuation losses (as interventions often did not reverse appreciation). This pattern of losses and limited effectiveness of intervention, particularly when the currency is undervalued, highlights the perils of intervening “too early” against appreciation pressures.

Heavy interventions frequently impose additional nonfinancial costs, some of which may have arisen in Latin America. Increasing central bank liabilities can pose constraints on monetary policy, by raising the fiscal costs of hiking interest rates, particularly when automatic mechanisms to cover operating losses of the central bank are not in place. Most countries in the region do not have such arrangements (although Brazil is an exception). Heavy FXIs may also send conflicting signals to the markets, and may be perceived as inconsistent with other policy objectives—particularly at times when the cyclical position of the economy calls for monetary tightening. It would be especially problematic if intervention were to lead to a perception that there are competing nominal anchors in the economy.

Although it is unclear from our evidence that a rule-based approach to intervention (rather than a fully discretionary approach) makes intervention more effective in influencing the exchange rate, the use of rules may have other advantages. Rules (and clear communication) can help assure markets that interventions are conducted in a manner that is consistent with other policy objectives, including appropriate flotation of the exchange rate. Quantity-based intervention rules can be particularly useful to minimize side effects—especially when the main objective of FXI is to increase reserve buffers—because they retain exchange rate flexibility and make clearest that no specific level of the exchange rate is being targeted. This helps prevent incentivizing onesided bets, and attracts further capital inflows. Exchange rate-based rules, on the other hand, might be useful instruments to prevent excessive volatility, but their design needs to avoid excessive interference with necessary adjustments of the exchange rate toward its equilibrium—that otherwise could lead to “too early” interventions—and to preserve some healthy degree of volatility—that would encourage investors to internalize the costs of unhedged positions. Similarly, the choice of instrument would depend on the primary objective of interventions. Spot market operations would be preferred when the objective is to accumulate reserves—so as to avoid interfering with exchange rate flotation—while put options would be more suitable when the objective is to mitigate volatility or the speed of appreciation— although other considerations, including the impact on the central bank’s balance sheet and secondary objectives, may also need to be taken into account.

Overall, the desirability of FX interventions as one part of the broader toolkit to confront “easy external financial conditions” needs to be individually assessed, taking into consideration the specific conditions of each country, including whether the exchange rate is overvalued or not, the extent of financial integration with the rest of the world—which appears to constrain the effectiveness of interventions, as well as other structural features such as the degree of dollarization. Moreover, exchange rate policies should be designed within a comprehensive strategy considering the pros and cons of different instruments available in authorities’ policy toolkit. In particular, and given their uncertain effects, FX policies should not be deployed too early (that is, before other, arguably more effective, instruments). In some instances, intervention may also need to be complemented with macroprudential tools, to mitigate other effects of “easy external financing conditions” and, as well, dampen some of the fiscal costs associated with FXI.37

Annex 3.1. Foreign Exchange Intervention and International Reserves: Data Availability
Data AvailabilityData Used in the Chapter
Foreign Exchange

Intervention
Stock of International

Reserves
Section on

Modalities of

Intervention
Econometric

Section
DailyWeeklyMonthlyDailyWeeklyMonthlyDailyWeekly
Brazil
Chile
Colombia
Costa Rica
Czech Republic
Guatemala
Honduras
India
Indonesia1
Israel
Korea
Malaysia
Mexico
Peru
Philippines
Poland
Romania
Russia2
South Africa
Thailand
Turkey
Uruguay
Australia3
Canada
New Zeland
Norway4
Note: “●” indicates that data are only used to describe qualitative information (for example, motives, rules, instruments, transparency). “*” indicates that data are (or for some period were) confidential or were facilitated for the purpose of this study. “♦” indicates data have recently become publicly available.
Annex 3.2. Methodological Strategy for the Panel Approach

The econometric approach entails identifying episodes of global shocks leading to appreciation pressures in emerging market (and ‘small’ advanced market) economies, and estimating the effect of intervention in a panel setting by exploiting the heterogeneous reactions of different central banks to such shocks. This approach helps to mitigate well-known endogeneity problems described in the literature by focusing on short time spans during which unobservable country-specific shocks are unlikely to be large (in relation to the identified global shock). The chapter follows the same approach taken in other studies on the effectiveness of foreign exchange intervention by undertaking a two-step instrumental variable estimation aimed at overcoming endogeneity problems. At the same time, it offers a methodological innovation by estimating the effect of intervention in a panel approach (most other studies focus on individual countries).

The sample of countries include Australia, Brazil, Chile, Colombia, Costa Rica, Guatemala, India, Indonesia, Israel, Mexico, Peru, Russia, Thailand, Turkey, and Uruguay, which are chosen on the basis of availability of weekly data on intervention or international reserves. All forms of intervention are included in the analysis, provided data are available.

First Stage: Central Bank Reaction Function

The first stage entails estimating individual central bank reaction functions—for countries in the sample that display sufficient variability in their interventions38–to create an instrumented variable for the main exchange rate equation. Reaction functions are modeled as a censored variable and estimated with a Tobit model, on a country-by-country basis to allow for all coefficients to be country specific (as different central banks may have different preferences). The model is estimated with weekly data over the period 2004–10 (excluding the period September 2008–June 2009 of the global financial crisis). The reaction function takes the following form:

Iit denotes country i’s amount of intervention (scaled by GDP) during week t. When available, actual intervention data are used. Otherwise, this variable is proxied by the change in the stock of international reserves adjusted for the estimated effect of changes in the value of reserve currencies.39ei,t-1 denotes the lagged change in the nominal (U.S. bilateral) exchange rate; rei,t is an estimate of the real effective exchange rate; rei,teq is an estimate of the equilibrium real exchange rate (based on the history of assessments by the IMF’s Consultative Group on Exchange Rates—CGER); Δi,t denotes the four-week speed of exchange rate appreciation (measured either on the exchange rate level in itself, or a Hodrick-Prescott (HP) trend estimated recursively to capture the information available to the central bank at that point in time); σi,t is a measure of intra-week exchange rate volatility (simple variance of the exchange rate, or a similar measure that strips the volatility arising simply from moving along the trend—computed as the sum of square values of deviations of the exchange rate from its HP trend); rei,t−1M2 and rei,t−1STD denote the ratios of reserves to M2 and reserves to short-term debt relative to the average of emerging market countries in the sample. These two terms seek to capture possible precautionary motives.

Second Stage: Exchange Rate Equation

The second step is the estimation of the exchange rate equation, using the predicted value of the first model as an instrument for the intervention variable. For countries with preannounced amount-based rules, actual intervention data are used.40 The model is estimated as a fixed-effects panel of 15 countries, with six episodes per country; and 12 weekly observations per episode and country. The episodes are selected on the basis of global (easy financing conditions) shocks that would normally lead to appreciation pressures on emerging market currencies. Such shocks are identified by looking at sharp movements in the U.S. dollar trade-weighted exchange rate (DXY) that take this variable at least 1 standard deviation below its long-term (HP) trend.

Focusing only on short episodes (t = 12 weeks) of global shocks allows us to mitigate endogeneity problems by ensuring that the main source of disturbances is the global shock and that unobservable country fundamentals do not change significantly over the episode window.

In the absence of guidance, either from theory or previous work, on how to model the short-term determinants of exchange rates, we choose a simple specification for the exchange rate equation, of the following form:

ei,t denotes the log of the nominal exchange rate (against the U.S. dollar) for country i at time t; ii,t is the domestic policy interest rate or interbank rate; it* is the U.S. Federal Funds interest rate; Si,t denotes the EMBI spread (or CDS spread when EMBI is not available); PtM, PtE, PtF are the logs of indices of metal, energy, and food commodity prices, which are introduced as a way to control for high frequency movements in terms of trade. DXY denotes the U.S. trade-weighted exchange rate and is introduced as a measure of market sentiment (similar to the VIX, this measure correlates closely with flows to EMEs); Îi,t denotes the predicted intervention amount estimated in the first step. Note that the effect of commodity prices and the DXY are allowed to be country specific, as different countries in the sample may have different trade structures and sensitivities to global financial shocks. Ideally, one would control also for other policy measures that could affect the exchange rate (for example, changes in reserve requirements, capital controls, and so on). Although their omission—owing to lack of data availability—could potentially introduce a bias in the estimation, we argue that such bias is likely to be small because these policy measures tend to be less frequent than—and so show low correlation with—FX interventions. Building a database of relevant policy measures is part of our research agenda.

Again, in the absence of theoretical guidance, the equation is estimated in first and second differences, in order to study possible effects on the rate and pace of appreciation.

The figure below highlights how the methodological approach helps to unveil the effect of intervention on exchange rates. In particular, it shows how the use of episodes rather than the full sample helps to eliminate the significance of the positive (wrong sign) coefficient in the equation in first difference (likely biased by endogeneity); and how the use of instruments rather than the actual intervention variable significantly increases the importance of the estimated effect. Finally, the introduction of controls in the regression does not appear to add much to the estimation, suggesting that the use of episode windows, rather than the full sample, usefully mitigates the effect of global and idiosyncratic shocks on the exchange rate.

This basic specification is subsequently modified by introducing dummy variables (related to modalities of intervention) that interact with the intervention variable. These alternative specifications allow us to answer some related questions such as: (i) is intervention more/less effective when the currency is undervalued? (ii) are rule-based interventions more or less effective than discretionary interventions? or (iii) does transparency enhance effectiveness?

Unveiling the Effect of FX Intervention—Results of Panel Approach under Different Specifications1

(Coefficient of intervention variable in exchange rate equation)

Source: IMF staff calculations.

1 Appreciation rate and pace of appreciation indicate first and second difference of the exchange rate.

2FXI: Without Controls—Full-time Span denotes model estimated with intervention variable (not instrument), without controls, and over the full period 2004–10 (excluding the 2008–09 financial crisis).

3FXI: Without Controls—Episodes denotes model estimated with intervention variable (not instrument), without controls, and over identified episodes only.

4IV- FXI: Without Controls—Episodes denotes model estimated with instrumented intervention variable, without controls, and over identified episodes only.

5IV- FXI:With Controls—Episodes denotes model estimated with instrumented intervention variable, with controls, and over identified episodes only.

Annex 3.3. Confidence Intervals for the Synthetic Currency

The synthetic currency approach discussed in the main text entails estimating the relationship between the value of a given country’s currency and those of the other emerging market/advanced market economies included in the sample—all of them measured relative to the U.S. dollar—before the policy shift takes place, as a way to construct a counterfactual. The model is estimated using ordinary least square over a 180-day window immediately preceding the event. The estimated model is then used to predict (out-of-sample conditional forecast) the value of the currency after the policy announcement. The difference (gap) between the predicted and the actual exchange rate is interpreted as a measure of the effect of the policy shift. By construction, however, the model provides a good fit up to the event, but a less accurate fit afterward, as the latter includes the out-of-sample forecast error. As a result, there could be a natural divergence of the actual and the predicted values, following but unrelated to the event. To ensure that the gap actually reflects the effect of the policy announcement, we check whether it is statistically significant by estimating the distribution of out-of-sample forecast errors at different horizons (1 day to 45 days after the end of the sample). This entails a bootstrapping exercise with 100 estimations identical to the one described above but shifting the estimation sample period so that the end point ranges from 150 days to 50 days before the policy shift announcement. Results confirm the significance of the gaps, at some point in the reported 45-day time windows (see figure below).

Bootstrap-Based Confidence Intervals for Synthetic Exchange Rates1

Source: IMF staff calculations.

1 Confidence intervals represent 1 standard-deviation bands based on bootstraped forecast errors.

Western Hemisphere Main Economic Indicators1
Output Growth

(Percent)
Inflation

(End-of-period, percent) 2
External Current Account Balance

(Percent of GDP)
1997–2006 Avg.20072008200920102011 Proj.2012 Proj.1997–2006 Avg.20072008200920102011 Proj.2012 Proj.1997–2006 Avg.20072008200920102011 Proj.2012 Proj.
North America
Canada3.52.20.5-2.53.12.82.62.02.51.80.82.22.02.01.10.80.4-2.8-3.1-2.8-2.6
Mexico3.53.21.5-6.15.54.64.08.23.86.53.64.43.53.0-1.7-0.9-1.5-0.7-0.5-0.9-1.1
United States3.21.90.0-2.62.82.82.92.54.10.71.91.42.11.4-4.2-5.1-4.7-2.7-3.2-3.2-2.8
South America
Argentina 32.88.66.80.89.26.04.67.08.57.27.710.911.011.00.52.31.31.80.90.1-0.5
Bolivia3.34.66.13.44.24.54.53.911.711.80.37.27.95.0-1.112.012.14.74.83.84.4
Brazil2.66.15.2-0.67.54.54.16.84.55.94.35.95.94.5-1.60.1-1.7-1.5-2.3-2.6-3.0
Chile4.04.63.7-1.75.35.94.93.37.87.1-1.43.04.53.2-0.64.5-1.91.61.90.5-1.3
Colombia2.86.93.51.54.34.64.58.85.77.72.03.23.23.1-1.5-2.9-3.0-2.2-3.1-2.1-2.2
Ecuador3.52.07.20.43.23.22.827.23.38.84.33.33.43.1-0.83.62.2-0.7-4.4-4.0-4.0
Guyana1.47.02.03.33.64.75.95.414.06.43.74.56.95.4-8.3-11.1-13.1-9.2-9.8-11.9-22.9
Paraguay1.66.85.8-3.815.35.64.59.25.97.51.97.210.77.5-1.01.5-1.80.6-3.2-4.1-3.7
Peru3.98.99.80.98.87.55.83.03.96.70.22.13.53.0-1.81.4-4.20.2-1.5-2.1-2.8
Suriname3.75.14.73.14.45.05.030.78.49.31.310.319.97.5-12.810.79.6-1.11.00.4-0.2
Uruguay1.27.38.62.68.55.04.29.28.59.25.96.96.86.5-1.0-0.9-4.70.60.5-1.0-1.6
Venezuela3.08.24.8-3.3-1.91.81.622.222.530.925.127.232.430.18.28.812.02.64.97.06.3
Central America
Belize6.11.23.80.02.02.32.51.54.14.4-0.41.34.42.5-12.8-4.1-9.8-8.4-2.7-8.1-6.7
Costa Rica5.37.92.7-1.34.24.34.411.110.813.94.05.86.05.5-4.3-6.3-9.3-2.0-3.6-4.5-4.7
El Salvador2.94.32.4-3.50.72.53.03.14.95.50.02.14.82.8-2.7-6.0-7.6-1.8-2.1-3.8-3.6
Guatemala3.66.33.30.52.63.03.27.38.79.4-0.35.46.35.5-5.4-5.2-4.30.0-2.1-3.3-4.0
Honduras4.26.24.1-2.12.83.54.09.58.910.83.06.58.06.5-4.7-9.0-15.4-3.7-6.2-7.3-7.1
Nicaragua3.93.12.8-1.54.53.53.78.016.913.80.99.28.67.3-18.4-16.6-23.3-11.9-14.1-17.6-16.5
Panama5.112.110.13.27.57.47.21.26.46.81.94.94.43.3-5.3-7.2-11.9-0.2-11.2-12.5-12.6
The Caribbean
Antigua and Barbuda5.06.51.8-8.9-4.13.12.51.55.20.72.42.93.03.1-13.5-34.3-30.5-22.5-13.9-18.6-18.2
The Bahamas3.71.9-1.7-4.30.51.32.31.82.94.51.31.72.01.7-10.9-17.8-15.9-11.7-12.4-15.0-14.3
Barbados2.33.8-0.2-4.7-0.52.02.53.14.87.24.35.17.02.3-5.6-4.5-9.6-5.5-7.4-6.7-6.0
Dominica1.02.53.2-0.31.01.62.51.46.02.13.22.43.51.3-18.5-25.0-31.8-28.1-28.0-29.2-27.5
Dominican Republic5.68.55.33.57.85.55.512.98.94.55.86.26.05.5-1.0-5.3-9.9-5.0-8.6-8.3-5.4
Grenada4.24.92.2-7.6-1.41.02.82.07.45.2-2.46.35.04.0-21.1-43.2-38.7-33.2-27.1-25.3-27.1
Haiti 40.83.30.82.9-5.18.68.816.07.919.8-4.74.79.16.5-0.7-1.5-4.4-3.4-2.3-4.0-4.6
Jamaica1.11.4-0.9-3.0-1.11.62.49.116.816.810.211.77.45.7-6.8-16.5-17.8-10.9-8.1-8.3-7.7
St. Kitts and Nevis3.44.24.6-9.6-1.51.51.54.02.17.61.02.23.82.9-24.8-24.3-33.2-34.0-27.5-30.5-28.9
St. Lucia2.11.50.7-3.60.84.23.92.76.83.81.0-0.65.21.9-15.6-31.3-27.8-14.4-16.7-29.1-20.8
St. Vincent and Grenadines3.78.0-0.6-1.1-2.32.52.52.08.38.7-1.62.05.91.9-19.7-34.6-35.2-35.0-33.6-37.5-34.3
Trinidad and Tobago8.54.82.4-3.50.02.22.45.07.614.51.313.49.55.57.524.831.39.017.618.719.2
Memorandum:
Latin America and the Caribbean3.15.74.3-1.76.14.74.28.06.28.14.86.66.85.8-1.20.4-0.7-0.6-1.2-1.4-1.8
(Simple average)3.45.43.6-1.43.03.93.97.87.98.92.85.97.25.3-6.5-7.5-9.4-7.1-6.9-8.4-8.2
LA-7 53.05.84.3-2.06.34.74.28.06.18.15.26.87.06.0-0.90.6-0.5-0.5-1.0-1.2-1.6
Eastern Caribbean Currency Union 63.35.61.9-6.8-1.72.62.82.25.84.20.62.44.42.7-19.1-35.6-36.9-26.5-26.4-31.2-28.5
Source: IMF staff calculations.
Latin America and the Caribbean Main Fiscal Indicators1
Public Sector Revenue

(Percent of GDP)
Public Sector Primary Expenditure

(Percent of GDP)
Public Sector Overall Balance

(Percent of GDP)
Public Sector Primary Balance 2

(Percent of GDP)
Public Sector Gross Debt

(Percent of GDP)
2008200920102011

Proj.
2012

Proj.
2008200920102011

Proj.
2012

Proj.
2008200920102011

Proj.
2012

Proj.
2008200920102011

Proj.
2012

Proj.
2008200920102011

Proj.
2012

Proj.
North America
Canada39.838.338.037.438.635.740.039.938.437.90.1-5.5-5.5-4.6-2.84.0-1.7-1.8-1.00.771.383.484.084.283.1
Mexico22.722.022.123.023.221.424.323.922.323.1-1.3-4.8-4.1-1.8-2.41.4-2.3-1.80.80.243.044.642.742.342.1
United States32.630.830.530.532.336.240.938.538.637.1-6.5-12.7-10.6-10.8-7.5-3.7-10.0-8.0-8.1-4.871.284.691.699.5102.9
South America
Argentina 333.436.138.938.738.630.736.037.438.438.5-0.8-3.8-1.7-3.1-3.12.70.11.50.30.158.157.647.840.736.7
Boliva38.936.133.934.835.032.633.830.432.832.84.30.62.00.71.06.32.33.62.02.337.540.537.434.332.8
Brazil36.335.637.436.336.232.333.535.033.333.2-1.4-3.1-2.9-2.4-2.64.02.12.43.03.070.767.966.165.765.0
Chile27.222.025.124.425.522.325.925.024.224.84.3-4.4-0.4-0.40.44.8-3.80.10.20.75.26.28.810.910.2
Colombia 426.526.824.725.125.822.926.024.425.323.70.0-2.6-2.8-3.5-1.13.50.80.3-0.22.031.036.236.536.335.4
Ecuador33.729.834.937.636.933.233.834.736.736.7-0.8-4.7-0.6-0.7-1.10.5-4.00.20.90.226.324.720.420.419.8
Guyana 527.729.328.330.629.929.631.229.732.131.6-3.6-3.5-3.2-3.1-3.3-1.9-1.9-1.4-1.5-1.661.661.561.362.361.8
Paraguay20.923.222.523.122.817.322.121.223.024.13.00.50.9-0.4-1.73.61.21.40.1-1.319.118.015.012.812.3
Peru21.118.920.120.620.317.419.819.519.919.32.2-2.1-0.6-0.5-0.13.8-0.80.60.71.025.026.624.322.520.7
Suriname 627.529.926.225.824.425.031.628.826.324.71.8-3.0-3.6-1.7-1.52.5-1.6-2.6-0.5-0.318.018.521.620.018.6
Uruguay 728.930.531.531.331.427.629.529.729.329.3-1.5-1.7-1.2-1.1-1.01.31.11.72.02.161.761.055.352.850.2
Venezuela31.624.931.234.933.532.831.533.935.034.5-2.7-8.2-6.0-1.5-2.0-1.2-6.7-2.7-0.1-1.024.632.738.740.242.1
Central America
Belize 828.727.026.627.127.024.424.624.925.425.30.4-1.2-2.4-2.1-2.64.22.41.71.71.778.280.381.779.778.2
Costa Rica 523.622.522.222.424.620.924.225.525.425.50.2-4.0-5.6-5.6-3.92.6-1.7-3.3-3.0-0.936.038.439.442.042.8
El Salvador 716.015.416.617.417.916.318.518.718.918.2-2.6-5.6-4.4-4.0-2.4-0.3-3.1-2.1-1.5-0.339.749.050.850.650.4
Guatemala 812.011.111.211.811.712.312.813.013.112.5-1.6-3.1-3.3-3.1-2.7-0.3-1.7-1.8-1.3-0.820.023.024.025.126.1
Honduras26.425.124.823.123.727.429.026.825.024.3-1.7-4.7-2.9-3.1-2.0-1.0-4.0-2.0-1.8-0.619.924.126.326.526.4
Nicaragua 728.728.530.130.329.627.629.228.529.128.00.0-2.10.2-0.40.01.1-0.71.61.21.677.482.982.379.675.1
Panama 1025.024.524.626.826.421.422.523.926.225.20.5-0.9-2.0-1.8-0.93.62.00.70.61.241.540.440.938.335.5
The Caribbean
Antigua and Barbuda 922.920.223.122.924.026.831.921.120.019.4-7.0-19.5-0.3-0.41.7-3.9-11.72.02.94.686.5110.593.390.387.7
The Bahamas 819.418.017.517.618.519.520.919.620.320.5-2.1-4.9-4.6-5.3-4.8-0.1-2.8-2.2-2.7-2.036.841.447.150.753.4
Barbados 1140.638.038.038.339.643.040.139.237.337.3-7.2-7.3-6.6-4.4-3.3-2.3-2.1-1.21.02.397.3110.7113.7114.2114.2
Dominica 946.247.645.744.343.043.045.945.542.539.90.9-0.1-1.6-0.11.23.21.70.31.83.185.785.985.583.479.6
Dominican Republic15.913.713.613.715.117.215.314.012.913.5-3.0-3.5-2.3-1.6-0.6-1.4-1.6-0.40.81.625.328.529.025.525.4
Grenada 929.427.228.225.825.832.430.829.325.523.1-5.1-6.3-3.6-2.4-0.7-3.0-3.6-1.10.42.7102.2115.0114.6110.6104.7
Haiti 815.017.729.427.623.917.421.226.932.929.6-3.1-4.42.2-5.7-6.2-2.4-3.52.5-5.3-5.837.624.715.712.316.8
Jamaica 926.927.026.525.726.022.020.921.920.119.0-7.4-10.9-5.8-3.5-2.34.96.14.65.67.0126.1141.1139.7137.1129.5
St. Kitts and Nevis 937.242.739.241.539.633.937.939.338.638.3-5.0-3.8-8.9-6.3-7.63.34.9-0.12.91.2170.0195.6200.4191.9190.0
St. Lucia 930.631.032.032.130.328.331.934.234.831.0-1.0-4.3-5.8-6.7-4.82.3-0.8-2.3-2.7-0.766.473.977.279.679.9
St. Vincent and Grenadines 934.535.233.735.035.533.135.736.836.936.5-1.7-3.8-6.6-5.3-4.01.4-0.5-3.1-1.9-0.967.674.982.287.185.9
Trinidad and Tobago39.630.636.134.735.229.637.037.838.638.38.2-9.0-4.1-6.5-5.810.0-6.4-1.7-3.9-3.125.434.439.849.349.8
ECCU 1230.830.831.831.931.030.934.232.030.929.0-3.4-8.0-3.6-2.7-1.6-0.1-3.4-0.21.02.085.299.697.896.493.8
Latin America and the Caribbean
PPP GDP-weighted average23.823.023.423.223.921.624.324.323.823.6-0.1-3.5-2.9-2.6-1.82.1-1.3-0.8-0.60.240.545.945.445.144.3
Simple average28.527.728.528.728.727.029.328.928.828.2-1.3-4.7-3.2-3.0-2.31.5-1.6-0.4-0.20.555.761.160.960.459.4
Source: IMF staff calculations.

    Other Resources Citing This Publication