Chapter 4 The Effectiveness of Intervention
Author:
Mr. Marcos d Chamon
Search for other papers by Mr. Marcos d Chamon in
Current site
Google Scholar
PubMedClose
Language:
English
Keywords:
BOOK; central bank; intervention; exchange rate; foreign exchange intervention; spot market; intervention policy; dollar sale; foreign exchange; Exchange rates; Foreign exchange intervention; Currency markets; International reserves; Global; Central America; intervention data; intervention program; purchase intervention; derivative intervention; Foreign exchange purchases; Foreign exchange sales; Caribbean
Download PDF (354.5 KB)

Abstract

Foreign exchange intervention is widely used as a policy tool, particularly in emerging markets, but many facets of this tool remain limited, especially in the context of flexible exchange rate regimes. The Latin American experience can be informative because some of its largest countries adopted floating exchange rate regimes and inflation targeting while continuing to intervene in foreign exchange markets. This edited volume reviews detailed accounts from several Latin American countries’ central banks, and it provides insight into how and with what aim many interventions were decided and implemented. This book documents the effectiveness of intervention and pays special attention to the role of foreign exchange intervention policy within inflation-targeting monetary frameworks. The main lesson from Latin America’s foreign exchange interventions, in the context of inflation targeting, is that the region has had a considerable degree of success. Transparency and a clear communication policy have been key. For economies that are not highly dollarized, rules-based intervention helped contain financial instability and build international reserves while preserving inflation targets. The Latin American experience can help other countries in the design and implementation of their policies.

This chapter surveys the empirical evidence on the effectiveness of foreign exchange intervention. There is an inherent identification problem when estimating the effect of intervention on the exchange rate, and several econometric techniques have been used to tackle it. Results vary widely, however; some studies find little or no effect, and others find sizable effects. The results are particularly mixed for the impact on the volatility of the exchange rate. Overall, the literature indicates that intervention has more traction when it is preannounced, that foreign exchange sales tend to be more effective than purchases, and that intervention through derivatives has an impact similar to that of spot interventions.

Introduction

Interventions in the foreign exchange market may entail considerable costs, as Chapter 2 discusses. So why are central banks willing to incur these costs? In other words, what is known about the effect of foreign exchange interventions? To what extent are interventions indeed effective in moving the exchange rate and instrumental in achieving the motivating objectives discussed in Chapter 2?

In theory, it is reasonable to expect unsterilized interventions—which directly affect domestic liquidity conditions and the interest rate differential with foreign markets—to substantially and durably affect the exchange rate. However, the effectiveness of sterilized interventions, which are designed specifically not to affect monetary aggregates or interest rates, is much less clear. Sterilized interventions are the instrument of choice for the inflation-targeting central banks in Latin America.

A considerable and growing literature empirically assesses the effectiveness of sterilized foreign exchange interventions, including several studies on Latin American economies that have accessed the intervention data published by the main central banks in the region. This chapter reviews the literature but first briefly discusses the substantial methodological challenges associated with the empirical work on interventions.

Endogeneity

As already noted, assessing the effectiveness of foreign exchange interventions is empirically challenging. This is because of an inherent endogeneity problem when trying to estimate the effect of intervention on the exchange rate. If a central bank wants to smooth shocks to the exchange rate through intervention, it will buy (sell) foreign exchange when the currency faces appreciation (depreciation) pressures. Suppose that intervention does indeed have traction, and the central bank perfectly succeeds in that effort; in that case the data would show a relatively stable exchange rate, along with purchases and sales of foreign exchange, which would bias the econometric results toward zero. Or even worse, suppose that intervention has traction but can only partially smooth the shock; in that case, during periods of appreciation pressures, the data would show an appreciation of the exchange rate combined with foreign exchange purchases, and vice versa during periods of depreciation pressures. The result can point to intervention having the “wrong” sign, because of the observed correlation of intervention and the trend in the exchange rate. Simple ordinary least squares (OLS) cannot reliably identify the causation between these two variables.

One technique often used to overcome this endogeneity problem is instrumental variables (IVs), which can flip the “wrong” sign obtained from OLS estimates. A good IV would explain the decision to intervene but would not directly affect the exchange rate through other channels. The IV estimates would rely on the changes in intervention that can be explained by that instrument. But a major challenge is to find good instruments. The literature often uses volatility as an instrument, but that is far from ideal. Some instruments can provide stronger identification, such as the use in Blanchard, Adler, and de Carvalho Filho (2015) of capital flows to other countries. However, they are not available at high frequencies. Although IV techniques help sharpen identification (typically yielding stronger results than OLS), it is not clear how far they go to address the inherent endogeneity problem.

Other studies have used vector autoregressive, vector error correction, and structural vector autoregressive models, which can, in principle, capture how the exchange rate and intervention affect each other. However, their success will depend on whether the model’s structure correctly captures the relationship between these variables, which can be a challenge. For example, the decision to intervene may be driven by sustained appreciation, by perceptions of misalignment, or by concerns about the future rather than by recent movements in the exchange rate. Similarly, the response of the exchange rate to intervention could stem from expectations about future intervention through a signaling channel, which may not be fully captured by recent intervention.

Some episodes of intervention lend themselves to event-study analysis, which could also help mitigate endogeneity issues. Most of the largest economies in the region have, at different points, announced changes to foreign exchange intervention rules. One can gauge the effect of that policy change by comparing the behavior of the exchange rate before and after the announcement. The event study can control for other variables affecting the behavior of the exchange rate around the announcement. Some authors have also used a synthetic control approach for major announcements, where a counterfactual exchange rate is constructed based on the experience of peer countries.

In principle, these strategies can help overcome the identification problem, provided that the announcements were indeed the major news affecting the exchange rate. The drawback is that this works only when there is a large pre-announced shift in intervention policy. Another concern is that the effect is gauged on the basis of the aftermath of the announcement before most of the intervention takes place (and portfolio balance effects materialize). Although the exchange rate is a forward-looking variable, and markets would price that future intervention, the transmission may not be complete.

An important empirical issue is whether an estimated effect of interventions is permanent or temporary. This is inherently difficult to measure given the very nature of the exchange rate’s stochastic process. Error bands will increase over time (they increase with the square root of time, in the case of a random walk). As a result, any estimated impact will eventually be swamped by that uncertainty. Attempts to estimate the persistence of an effect will likely be driven by the assumptions imposed on the estimation model (for example, treat all changes as permanent). The question of persistence is very important from a policy perspective. If a central bank incurs costs to intervene, any cost-benefit analysis will depend on how durable those benefits are. Even if intervention is effective on impact, it may not be worthwhile if the costs are persistent while the benefits are transitory. From a theoretical perspective, if intervention has traction through a signaling channel, that effect should eventually fade as agents receive more and more information, and the original signal loses its relevance. However, the portfolio balance channel could have a permanent effect—for example, if the central bank makes a sterilized purchase of foreign exchange and holds on to it, it will have permanently affected the relative supply of local and foreign currency assets. However, it will be very difficult, if not impossible, to empirically ascertain the persistence of that effect.

The inherent endogeneity problem complicates the research into the effects of foreign exchange interventions. What can be ascertained with some confidence, however, is that the bias involved tends to attenuate the effect of intervention, as discussed earlier. Studies that find an effect do so despite that attenuation bias.

The Effectiveness of Foreign Exchange Intervention

Despite the methodological challenges, the literature on the effectiveness of foreign exchange interventions is substantial. This section discusses this literature, with a focus on studies covering Latin America. A relatively large number of studies have looked at Latin American countries over the past 10–15 years (the period in which many countries transitioned to inflation targeting), facilitated by relatively transparent intervention policies and the public availability of intervention data for several of the region’s key economies.

The findings on the effectiveness of intervention are discussed along two main dimensions: (1) the effect on the exchange rate level, and (2) the effectiveness in reducing exchange rate volatility. In addition, this chapter distills lessons from the literature on the relative effectiveness of specific intervention modalities. It also discusses the available evidence on the impact of capital controls on the effectiveness of foreign exchange interventions as well as the impact of interventions on real variables.

Effect on the Exchange Rate Level

Recent studies have found a wide range of results for the estimated impact of foreign exchange interventions on the level of the exchange rate, depending on the countries and periods considered and methods used (Table 4.1). Overall, the evidence is mixed. For all the countries surveyed, at least one study finds no statistically significant effect. For example, using a sample of 15 countries (many in Latin America), Adler and Tovar (2014) find no effect of foreign exchange intervention on the level of the exchange rate. Similarly, for Colombia, Villamizar-Villegas (2016) and Rincón and Toro (2010) also find no effects. Other papers report regressions with “wrong” signs (such as the domestic currency depreciating after a sale of foreign exchange reserves), possibly because of the endogeneity issues discussed in the previous section. Several other studies do find positive effects, however, and sometimes the estimated impact is quite large.

  • Brazil: Barroso (2014) finds that a $1 billion intervention in Brazil has an effect on the exchange rate of 0.51–1.18 percent. Other studies that find a significant effect tend to point to smaller estimates (0.25–0.50 percent).

  • Chile: Pincheira (2013) estimates the effect of a $1 billion intervention on the Chilean peso to be as large as 3 percent in 2008. However, other intervention episodes point to smaller effects.

  • Colombia: Kuersteiner, Phillips, and Villamizar-Villegas (2016) estimate that a put option (foreign exchange purchase) of about $100 million depreciates the exchange rate by as much as 2 percent. However, other studies fail to find an effect. Echavarría, Melo Velandia, and Villamizar (2014) find that preannounced interventions move the exchange rate by about 0.50 percent per $100 million of intervention.

  • Mexico: Tobal and Yslas (2016) report impulse response functions when the Mexican peso depreciates 2 percent after a one standard deviation intervention shock.

  • Peru: Studies typically struggle to find an effect, given the relative stability of the exchange rate. Lahura and Vega (2013) estimate a $25 million intervention moves the exchange rate by 0.1 percent.

Table 4.1.

Summary of the Literature, by Country

article image
article image
article image
article image
article image
Source: Authors’ compilation. Note: 2SLS = two-stage least squares; CUE = continuously updated estimator; EGARCH = exponential generalized autoregressive conditional heteroskedasticity; GARCH = generalized auto regressive conditional heteroskedasticity; IV = instrumental variable; OLS = ordinary least squares; SVAR = structural vector autoregression; and VAR = vector autoregression.

Although these studies find substantial intervention effects, they are on the upper end of the estimates for these countries.1 If differences in estimates across studies were normally distributed, the largest estimates would be discounted as outliers. However, given the endogeneity and attenuation bias involved, large point estimates could be at least partly driven by the success of these studies to tackle the identification problem.

The average net foreign exchange purchases were substantial in these five economies. For example, over 2010–12, average yearly net purchases were about $35 billion in Brazil, $4 billion in Chile and Colombia, $20 billion in Mexico, and $6.5 billion in Peru. Multiply these average yearly intervention figures by the point estimates above, and the implied effect would range from 12 percent to 26 percent depreciation of the exchange rate on that year. This suggests foreign exchange interventions may have had a substantial effect on the evolution of the exchange rates. Two important caveats are in order. First, this back-of-the-envelope calculation is based on the upper range of estimates from the studies reviewed (and on the lower end of estimates, there are other studies that point to no effect). Second, and perhaps more important, it is not clear whether the effect on the exchange rate is permanent or transitory, which is subsequently discussed in more detail.

The Effect on Exchange Rate Volatility

A second, frequently cited objective of foreign exchange intervention is to limit the excessive volatility of the exchange rate. Empirical examination of the effect of foreign exchange intervention on exchange rate volatility is, however, subject to two challenges. First, in the empirical literature, volatility is often associated with the variance of the shocks or with the option-implied parameter for the standard deviation of the exchange rate stochastic process. In practice, authorities are more likely to be worried about sharp movements in the exchange rate—a risk that may be better captured by measuring sizable deviations from the equilibrium level, rather than by these traditional measures of volatility. Second, most papers tend to focus on the immediate, short-term effect of intervention on volatility. In practice, however, there could be separate short- and long-term effects. Although intervention may increase volatility on impact (including if it succeeds in facilitating a correction in the level), it may lower volatility in the medium to long term, reflecting central banks’ commitment to use foreign exchange intervention in response to large movements or the perception of disorderly market conditions. A few recent studies that use deviations from equilibrium as a volatility measure have found interventions to be effective. For example, Adler and Tovar (2014), in a study covering 15 countries, including several in Latin America, estimate the effect of intervention on the pace (or acceleration) of appreciation of exchange rates after global shocks, and they find that interventions are effective in reducing this pace—and thus in limiting deviation from the equilibrium exchange rate. Fratzscher and others (2019) use exchange rate variation as a measure of volatility and also find intervention effective for smoothing exchange rates (comparing exchange rate changes over five trading days before and after intervention) in a larger sample of countries.

Most studies, however, define volatility more traditionally as the standard deviation of daily exchange rate returns, with mixed results. Berganza and Broto (2012) find a positive effect of foreign exchange intervention on exchange rate variance; they estimate interventions in inflation-targeting countries (mainly in Latin America) to be more effective in lowering volatility than in non–inflation-targeting countries.

Similarly, Nedeljkovic and Saborowski (2017) also find that foreign exchange intervention (both in spot and nondeliverable futures) was effective in reducing implied volatility in Brazil (with a 2.5 percent reduction in volatility per $1 billion spent). For Peru, Tashu (2014) estimates that foreign exchange interventions in the a.m. sessions reduce volatility by up to 0.14 percent between a.m. and p.m. trading sessions.2 In Colombia, Kamil (2008) shows that intervention reduced volatility when purchases were made during a period of monetary easing. However, interventions were ineffective in a period of monetary tightening, when markets may have viewed large interventions to curb appreciation to be incompatible with meeting the inflation target in an overheating economy. Furthermore, Villamizar-Villegas (2016) finds that a $100 million sterilized purchase reduces realized volatility, measured by a squared log change in the daily exchange rate, by up to 0.5 percent in Colombia. Domaç and Mendoza (2004) and Chamon (2015) find that interventions also reduce statistical volatility in Mexico.

On the other hand, several other papers find that interventions have no significant effect and sometimes even an adverse effect on traditional measures of volatility. For example, Moura, Pereira, and de Moraes Attuy (2013) and Stone, Walker, and Yasui (2009) find higher volatility in response to futures market intervention in Brazil. Rincón and Toro (2010) find that intervention increases volatility significantly in Colombia. Furthermore, foreign exchange sales are found to increase both short- and long-term volatility in Mexico (Guimaraes and Karacadag 2004).

One reason for the wide range of estimates on the effect of intervention on volatility could be the use of different estimation methods and identification strategies. However, Broto (2013), who uses a homogeneous model to assess effectiveness of intervention in Chile, Colombia, Mexico, and Peru, also finds a wide range of results for the countries in her sample. Specifically, Broto finds that intervention leads to higher volatility in Chile, and that although both purchases and sales lower volatility in Mexico and Peru, only purchases do so in Colombia. Thus, results are mixed even when using a consistent empirical method across countries.

Several papers suggest that the effectiveness of intervention in dampening volatility operates through the signaling channel. For example, Mundaca (2011) shows that publicizing information about past interventions has strengthened the effectiveness of interventions and lowered exchange rate volatility in Peru (albeit for very short time periods). Furthermore, Broto (2013) finds for the four countries in her sample that first interventions, whether isolated or under a rule, reduce volatility and that the size of intervention plays only a minor role, which also suggests that interventions work mainly through the signaling channel.

Duration of Effects

Most studies do not attempt to measure the persistence or duration of the effect of intervention. This is likely because of the methodological constraints discussed earlier. Among the studies that try to address duration, some find effects to be (very) short lived. For example, Mundaca (2011) uses high-frequency data and finds that interventions in Peru have no effect past the two-hour window during which the central bank intervenes. Similarly, Kohlscheen and Andrade (2014) find no significant effect of intervention in Brazil on the exchange rate levels after 90 minutes. However, other studies find more persistent effects. For example, Tobal and Yslas (2016) use a structural vector autoregressive model framework and estimate the effect of interventions on the exchange rate level to last about two months in Mexico and one month in Brazil. Echavarría, Melo Velandia, and Villamizar (2014) estimate the effect of interventions through volatility options to be effective for periods of up to 25 days in Colombia—a result that is corroborated by Villamizar-Villegas (2016). For Brazil, Chamon, Garcia, and Souza (2017) also find effects to last several weeks.

Modalities of Intervention and Effectiveness

Since the modalities of intervention differ between countries and over time, a comparison of the various studies for Latin American countries also sheds some light on the relative effectiveness of different types of intervention on the exchange rate level.

Purchases versus Sales

In principle, it is not clear that the effects of foreign exchange sales and purchases should be asymmetric. There are theoretical arguments supporting the effectiveness for both sales and purchases. As discussed in Chapter 2, intervention has traction in a setting where capital flows respond to return differentials, but at a finite pace (unlike a setting where uncovered interest rate parity (UIP) holds and flows would come until the return differential is arbitraged away). An asymmetry in the effect of intervention sales and purchases could come from flows being more responsive to returns or conditions at times of appreciation pressures, or vice versa. For example, it could be that during “risk-on” periods, capital flows are more sensitive to the effect of intervention than during “risk-off” episodes. Also, interventions may have more traction under disorderly market conditions, which are more likely to be associated with depreciation pressures. During such episodes—associated with foreign exchange sales by the central bank—the authorities typically intervene in large amounts, aiming explicitly at influencing market prices. Such a pattern of intervention could be expected to have a stronger effect on the market, including through signaling effects, than that of a central bank that is gradually accumulating foreign reserves (that is, purchases) in an environment of appreciation pressures.

Foreign exchange sales might also send a stronger signal because selling foreign reserves is costlier for the central bank than buying reserves, since foreign currency is a limited resource. It could be argued that foreign exchange sales reveal a significant disagreement of authorities with the current exchange rate and thus send a strong signal to markets (provided there are sufficient reserves to follow through). However, a central bank seeking to stem appreciation pressures by buying foreign exchange could, in principle, do so in unlimited amounts, which is highly credible and could therefore also imply a strong signal.

Several studies explore asymmetries in the effects of interventions that buy and sell foreign exchange. Most of these (such as Kohlscheen and Andrade (2014) for Brazil, Broto (2013) for Chile, Domaç and Mendoza (2004) for Mexico, and Lahura and Vega (2013) and Tashu (2014) for Peru) find that foreign exchange sales have a larger effect than foreign exchange purchases. However, Echavarría, Melo Velandia, and Villamizar (2014) and Kuersteiner, Phillips, and Villamizar-Villegas (2016) find stronger effects of purchase interventions (put options to buy foreign exchange) in Colombia.

Spot versus Derivative Interventions

While most studies focus on spot interventions, a few cover episodes of derivative interventions. These studies mostly find derivative interventions to be either similar or less effective than spot interventions. For example, Nedeljkovic and Saborowski (2017) show, for Brazil, that $1 billion of interventions on the spot market impacts the exchange rate by 1 percent, a statistically indistinguishable effect from the change of 0.7 percent resulting from interventions through derivatives. Similarly, Barroso (2014) finds that $1 billion of interventions on the spot and derivatives market affect the exchange rate by 0.51 and 0.31 percent, respectively. In an event study, Echavarría, Melo Velandia, and Villamizar (2014) find that intervention through options is more effective than other types of intervention, in the sense that it delivers the intended outcomes more often (they focus on appreciation/depreciation events, not on the magnitudes of them).

Preannounced versus Discretionary

The literature also suggests that preannounced interventions are generally more effective than discretionary ones. For example, Echavarría, Melo Velandia, and Villamizar (2014) find that preannounced interventions in Colombia have three times more traction. In a related vein, several studies examine the direct effect of announcements themselves. Stone, Walker, and Yasui (2009), who compare the effect of announcements versus actual interventions in Brazil, conclude that announcements are significantly more effective than interventions themselves. For Chile, Fuentes and others (2014) find that although the effect of actual interventions is transitory, the effects of the announcements of intervention programs are both significant and persistent. This result is consistent with Tapia and Tokman (2004), who found that intervention announcements—not necessarily the actual interventions—were the main policy instrument affecting the exchange rate in Chile in 2001–02. Chamon (2015) also estimates a sizable effect of announced changes in foreign exchange intervention rules in Mexico. And Chamon, Garcia, and Souza (2017), who use a synthetic control approach to estimate the effect of the announcement of a large intervention program in Brazil, show that the exchange rate responded strongly after the announcement of the program, even though the pace of intervention actually declined (the central bank was intervening even more heavily prior to the announcement, but in a discretionary fashion and failing to revert the depreciation pressures).3 While preannounced interventions should have an effect similar to that of discretionary interventions through a portfolio-balanced channel, the aforementioned evidence suggests that preannounced intervention may have a stronger effect through the signaling channel.

Interventions and Capital Controls

Several studies explore the effectiveness of interventions in relation to capital account openness. More capital mobility can help make domestic and foreign assets closer substitutes, which could reduce the effect of foreign exchange intervention through the portfolio balance channel. In contrast, it may be more difficult for capital flows to offset the effect of intervention in a closed economy (where frictions can make “quantities”—that is, intervention—play a relatively larger role in shaping the foreign exchange market for a given interest rate differential). In line with this theory, Adler and Tovar (2014) find that financial openness (measured by the Ito-Chinn Index) reduces the effect of intervention. Similarly, for Colombia, Rincón and Toro (2010), conclude that foreign exchange interventions are effective only when used in conjunction with capital controls.

Countries in Latin America, however, are relatively open compared to other regions (and remain so, even during episodes where additional capital controls are imposed). Thus, any additional gain in traction from imposing new capital controls is likely to be small relative to the gains identified in cross-country studies, which may be based on comparisons with economies that are much more closed.

Exchange rate restrictions can also shape the structure and development of the foreign exchange market itself. For example, restrictions on denominating contracts in foreign currency contributed to the development of derivatives markets in Brazil (Garcia and Volpon 2014). The Brazilian market for nondeliverable forwards has become much more liquid than the spot market, and the former is where much of the price discovery process for the exchange rate takes place. This can help explain why derivative interventions are more commonly used in Brazil.

The Effect on Other Variables and the Role of Market Conditions

Most of the literature on foreign exchange intervention focuses on the effect on the level and volatility of the exchange rate, and little attention is generally given to effects on other variables, such as interest rates, inflation, and inflation expectations. Since the foreign exchange interventions examined are typically sterilized interventions, such an effect on interest rates or real variables would likewise not be expected. A few studies confirm this; for example, Tobal and Yslas (2016) find that foreign exchange interventions are not associated with an immediate expansion in the monetary conditions (that is, an increase in the monetary base and a fall in interest rates) in Brazil and Mexico. Similarly, Villamizar-Villegas (2016) concludes that neither the 1-year Treasury bond yield nor the interbank rate respond to intervention in Colombia. These studies seem to indicate that interventions are indeed sterilized.

However, one study finds that a positive intervention shock leads to a significant increase in credit in Brazil: Garcia-Verdú and Zerecero (2013) present a model in which foreign exchange intervention causes banks to hold more bonds that are issued as part of sterilization. As a result, the portfolio balance channel stimulates banks to increase the supply of loans, substituting away from holding additional bonds. The paper finds this channel to be empirically relevant in Brazil. Pincheira (2013) estimates that in Chile a preannounced intervention program in 2008 had a significant, albeit short-lived, effect on inflation expectations at long horizons, although interventions carried out in 2011 showed no such impact. The effect on inflation expectations might be because interventions lower the credibility of the central bank’s inflation-targeting policy, particularly in 2008, when inflation in Chile was already high.

A shortcoming of the literature is that few studies seem to pay much attention to the prevailing macroeconomic and market conditions at the time of intervention. However, as discussed earlier (in the context of the asymmetries between interventions involving foreign exchange sales and purchases), macroeconomic and market conditions should have a considerable impact on the effectiveness of intervention. For example, Adler and Tovar (2014) find that interventions to stem an appreciation are more effective when the real exchange rate is seen as overvalued. Also, Fratzscher and others (2019) and Humala and Rodríguez (2009) find evidence that interventions are more effective in times of market turbulence. Kamil (2008) finds interventions in Colombia contain appreciation pressures that are effective during periods of monetary easing but not during times when the economy is overheating. Most studies, however, do not systematically account for, or control for, the macro and financial environment in which interventions are conducted.

Conclusion

Overall, the literature is mixed on the effectiveness of interventions; with the effectiveness of sterilized interventions on the level of (or volatility of) exchange rates ranging from negligible, through counterproductive, to large and positive. The wide range is probably due, at least in part, to the considerable identification challenges for which different authors have used various approaches and instruments, arguably with variable success. A related shortcoming of the literature is that few studies provide meaningful evidence on the duration or persistence of the effect of interventions—which is an important dimension of effectiveness, but difficult to measure.

The fact that many studies do find significant intervention effects against the attenuation bias strongly suggests that interventions have at least some effect. This is consistent with the more informal evidence for effectiveness; for example, Figure 4.1 shows the path of the Peruvian sol relative to its regional peers. Among the Latin American countries considered in this chapter, Peru is particularly active in the foreign exchange market, with frequent daily interventions and a stock of reserves that amounts to more than 30 percent of GDP. And while econometric studies for Peru find variable and often only short-lived effects from these interventions, a glance at the trajectory of the exchange rate over the past decade suggests that the sol exchange rate has indeed been much less volatile than the rates of its regional peers.

Figure 4.1.
Figure 4.1.

Latin America: Nominal Exchange Rate Indices, 2005–18

(2008 = 100)

Source: Bloomberg Finance L.P.

It is particularly informative to compare Chile and Peru, given that both countries rely on metal commodity exports, have strong macroeconomic policy management, and are opposites on intervention policy (Peru intervenes frequently and Chile very rarely). The evolution of their respective exchange rates shows that the Chilean peso appreciated more strongly during good times (favorable commodity prices and capital inflows) but also depreciated more strongly under less favorable global conditions. For Peru, while there are many factors that may have contributed to the relative stability of the Peruvian sol, intervention likely played a significant role. Relative exchange rate stability combined with frequent intervention precludes finding an econometric effect of intervention, for the reasons discussed earlier in this chapter.

The literature sheds limited light on the determinants of intervention effectiveness. As one would expect, interventions seem to become less effective the more open the capital account is. Evidence also suggests that interventions to stem depreciations may be more effective than those countering appreciation pressures, and that announced interventions are more effective than unannounced ones. However, a systematic investigation of the circumstances, specific market conditions, institutional setups, and intervention modalities of each type of intervention, and when each is effective, would fill an important gap in the literature.

References

  • Adler, Gustavo, and Camilo E. Tovar. 2014. “Foreign Exchange Interventions and Their Impact on Exchange Rate Levels.” Monetaria, Centro de Estudios Monetarios Latino Americanos (1):1–48.

    • Search Google Scholar
    • Export Citation

  • Barroso, João R. B. 2014. “Realized Volatility As an Instrument to Official Intervention.” Working Papers Series 363, Central Bank of Brazil, Research Department.

    • Search Google Scholar
    • Export Citation

  • Berganza, J. Carlos, and Carmen Broto. 2012. “Flexible Inflation Targets, Forex Interventions and Exchange Rate Volatility in Emerging Countries.” Journal of International Money and Finance 31 (2): 428–44.

    • Search Google Scholar
    • Export Citation

  • Blanchard, Olivier J., Gustavo Adler, and Irineu E. de Carvalho Filho. 2015. “Can Foreign Exchange Intervention Stem Exchange Rate Pressures from Global Capital Flow Shocks?” IMF Working Paper 15/159, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Broto, Carmen. 2013. “The Effectiveness of Forex Interventions in Four Latin American Countries.” Emerging Markets Review 17 (C): 224–40.

    • Search Google Scholar
    • Export Citation

  • Chamon, Marcos. 2015. “Mexico, Selected Issues.” IMF Country Report No. 15/314, International Monetary Fund, Washington, DC.

  • Chamon, Marcos, Márcio Garcia, and Laura Souza. 2017. “FX Interventions in Brazil: A Synthetic Control Approach.” Journal of International Economics 108 (C): 157–68.

    • Search Google Scholar
    • Export Citation

  • de Roure, Calebe, Steven Furnagiev, and Stefan Reitz. 2015. “The Microstructure of Exchange Rate Management: FX Intervention and Capital Controls in Brazil.” Applied Economics 47 (34–35): 3617–32.

    • Search Google Scholar
    • Export Citation

  • Domaç, Ilker, and Alfonso Mendoza. 2004. “Is There Room for Foreign Exchange Interventions under an Inflation Targeting Framework? Evidence from Mexico and Turkey.” Discussion Papers 0206, Research and Monetary Policy Department, Central Bank of the Republic of Turkey.

    • Search Google Scholar
    • Export Citation

  • Echavarría, Juan J., Luis F. Melo Velandia, and Mauricio Villamizar. 2014. “The Impact of Foreign Exchange Intervention in Colombia: An Event Study Approach.” Desarro, SOC, No. 73, Bogotá, Primer Semestre de 2014.

    • Search Google Scholar
    • Export Citation

  • Fratzscher, Marcel, Oliver Goede, Lukas Menkhoff, Lucio Sarno, and Tobias Stöhr. Forthcoming. “When Is Foreign Exchange Intervention Effective? Evidence from 33 Countries.” American Economic Journal: Macroeconomics.

    • Search Google Scholar
    • Export Citation

  • Fuentes, Miguel, Pablo Pincheira, Juan Manuel Julio, Hernán Rincón-Castro, Santiago García-Verdú, Miguel Zerecero, Marco Vega, Erick Lahura, and Ramon Moreno. 2014. “The Effects of Intraday Foreign Exchange Market Operations in Latin America: Results for Chile, Colombia, Mexico, and Peru.” BIS Working Paper No. 462, Bank for International Settlements, Basel.

    • Search Google Scholar
    • Export Citation

  • Garcia, Márcio, and Tony Volpon. 2014. “DNDFs: A More Efficient Way to Intervene in FX Markets?” Textos para discussão 621, Department of Economics PUC-Rio, Brazil.

    • Search Google Scholar
    • Export Citation

  • García-Verdú, Santiago, and Miguel Zerecero. 2013. “On Central Bank Interventions in the Mexican Peso/Dollar Foreign Exchange Market.” BIS Working Paper 429, Bank for International Settlements, Basel.

    • Search Google Scholar
    • Export Citation

  • Guimaraes, Roberto, and Cem Karacadag. 2004. “The Empirics of Foreign Exchange Intervention in Emerging Market Countries: The Cases of Mexico and Turkey.” IMF Working Paper No. 04/123, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Humala, Alberto, and Gabriel Rodríguez. 2009. “Foreign Exchange Intervention and Exchange Rate Volatility in Peru.” Working Paper 2009–008, Banco Central de Reserva del Perú.

    • Search Google Scholar
    • Export Citation

  • Kamil, Herman. 2008. “Is Central Bank Intervention Effective under Inflation Targeting Regimes? The Case of Colombia.” IMF Working Paper No. 08/88, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Kohlscheen, Emanuel. 2013. “Order Flow and the Real: Indirect Evidence of the Effectiveness of Sterilized Interventions.” BIS Working Paper 426, Bank for International Settlements, Basel.

    • Search Google Scholar
    • Export Citation

  • Kohlscheen, and Sandro C. Andrade. 2014. “Official Interventions through Derivatives.” Journal of International Money and Finance 47 (2014): 202–216.

    • Search Google Scholar
    • Export Citation

  • Kuersteiner, Guido M., David C. Phillips, and Mauricio Villamizar-Villegas. 2016. “Effective Sterilized Foreign Exchange Intervention? Evidence from a Rule-Based Policy.” Borradores de Economia 964, Banco de la Republica de Colombia.

    • Search Google Scholar
    • Export Citation

  • Lahura, Erick, and Marco Vega. 2013. “Asymmetric Effects of FOREX Intervention Using Intraday Data: Evidence from Peru.” BIS Working Paper 430, Bank for International Settlements, Basel.

    • Search Google Scholar
    • Export Citation

  • Marins, Jaqueline, Gustavo Araujo, and José Vicnete. 2016. “Do Central Bank Foreign Exchange Interventions Affect Market Expectations?” Applied Economics 49 (31): 3017–31

    • Search Google Scholar
    • Export Citation

  • Moura, Marcelo L., Fatima R. Pereira, and Guilherme de Moraes Attuy. 2013. “Currency Wars in Action: How Foreign Exchange Interventions Work in an Emerging Economy.” Insper Working Paper wpe_304, Insper Instituto de Ensino e Pesquisa.

    • Search Google Scholar
    • Export Citation

  • Mundaca, Gabriela B. 2011. “How Does Public Information on Central Bank Intervention Strategies Affect Exchange Rate Volatility? The Case of Peru.” Policy Research Working Paper Series 5579, World Bank, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Nedeljkovic, Milan, and Christian Saborowski. 2017. “The Relative Effectiveness of Spot and Derivatives Based Intervention: The Case of Brazil.” IMF Working Paper 17/11, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Pincheira, Pablo. 2013. “Interventions and Inflation Expectations in an Inflation Targeting Economy.” Working Papers Central Bank of Chile 693, Central Bank of Chile.

    • Search Google Scholar
    • Export Citation

  • Rincón, Hernán, and Jorge Toro. 2010. “Are Capital Controls and Central Bank Interventions Effective?” Borradores de Economia 007622, Banco de la República.

    • Search Google Scholar
    • Export Citation

  • Stone, Mark R., W. Christopher Walker, and Yosuke Yasui. 2009. “From Lombard Street to Avenida Paulista: Foreign Exchange Liquidity Easing in Brazil in Response to the Global Shock of 2008–09.” IMF Working Paper 09/259, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Tapia, Matías, and Andrea Tokman. 2004. “Effects of Foreign Exchange Intervention under Public Information: The Chilean Case.” Working Papers Central Bank of Chile 255, Central Bank of Chile.

    • Search Google Scholar
    • Export Citation

  • Tashu, Melesse. 2014. “Motives and Effectiveness of Forex Interventions: Evidence from Peru.” IMF Working Paper 14/217, International Monetary Fund, Washington, DC.

    • Search Google Scholar
    • Export Citation

  • Tobal, Martín, and Renato Yslas. 2016. “Two Models of FX Market Interventions: The Cases of Brazil and Mexico.” Working Paper 2016–14, Banco de México.

    • Search Google Scholar
    • Export Citation

  • Villamizar-Villegas, Mauricio. 2016. “Identifying the Effects of Simultaneous Monetary Policy Shocks.” Contemporary Economic Policy 34 (2): 286–296.

    • Search Google Scholar
    • Export Citation
1

Few studies seem to exist on Costa Rica and Uruguay. Refer to Chapter 10 on Costa Rica and Chapter 13 on Uruguay for a detailed discussion of the effectiveness of intervention in these countries.

2

The foreign exchange market in Peru operates between 9:00 a.m. and 1:30 p.m. local time. Volatility is measured by the square root of the squared deviation of the exchange rate from the weekly average exchange rate. Te paper uses exchange rate movements during the a.m. session to estimate the Central Reserve Bank of Peru’s reaction function. Predicted values from these reaction functions are then used as instruments for interventions in the regressions for changes in exchange rate volatility between the a.m. and p.m. sessions.

3

This pattern could be the result of the program committing to a larger stream of interventions than the market was expecting under discretionary interventions.

Within Same Series

Other IMF Content

Other Publishers Content

Contributor Notes

The opinions expressed in this chapter are the sole responsibility of the authors.
  • Collapse
  • Expand
Cover Foreign Exchange Intervention in Inflation Targeters in Latin America
Foreign Exchange Intervention in Inflation Targeters in Latin America
Author:
Mr. Marcos d Chamon
,
Mr. David J Hofman
,
Mr. Nicolas E Magud
, and
Alejandro M. Werner
ISBN:
9781484375686
Publication Date:
28 Feb 2019
Publisher:
International Monetary Fund
DOI:
https://doi.org/10.5089/9781484375686.071
Pages:
320
  • <sup><xref ref-type="fig" rid="Rch04fig01">Figure 4.1</xref>.</sup>
    View in gallery
    Figure 4.1.

    Latin America: Nominal Exchange Rate Indices, 2005–18

    (2008 = 100)

Figure 4.1.

Latin America: Nominal Exchange Rate Indices, 2005–18

(2008 = 100)