Chapter 2 Why Intervene? (2024)

This chapter discusses some of the motives for why emerging market countries may want to intervene in the foreign exchange market, even under a flexible exchange rate regime. The motives include reserve accumulation for precautionary reasons, intervention to attenuate financial stability risks from sharp movements in the exchange rate, and efforts to manage the exchange rate due to pass-through or competitiveness concerns. The discussion focuses on the potential benefits of those channels, but the overall desirability of intervention will also depend on its cost.

Introduction

Foreign exchange intervention plays a central role in fixed exchange rate regimes. If a central bank is committed to maintaining an exchange rate, it must stand ready to buy or sell foreign exchange at that price. However, it is much less clear what role intervention should play in a flexible exchange rate regime. Standard macroeconomic models provide no guidance on the role of foreign exchange intervention. If anything, they suggest that intervention should not impact the exchange rate—that is, intervention would have no traction. Open economy models—dating back to Fleming (1962), Mundell (1963), and Dornbusch (1976)—typically assume perfect capital mobility. The uncovered interest parity (UIP) condition has become a cornerstone for such models. Under UIP, the expected change in the exchange rate is given by the interest rate differential, which in log form yields the familiar expression:

$\begin{array}{cc}{i}_t-i_t^{*}=e_t-E_t\left[e_{t+1}\right],& \left(2.1\right)\end{array}$

where i and i* denote the home and foreign interest rate, respectively, and e is the exchange rate (with an increase denoting an appreciation of the home currency). This condition implies that the exchange rate will respond only to changes in the interest rate differential or in the expected change in the exchange rate. No amount of sterilized foreign exchange purchases or sales by the central bank would affect the exchange rate. In practice, however, UIP does not hold. Even covered interest parity (in which futures are used in place of the expected exchange rate) has started to break down in practice (see Chapter 5). The literature emphasizes two main channels through which sterilized intervention (purchases and sales of foreign exchange that leave the central bank’s interest rate unchanged) can affect the exchange rate: the portfolio balance and the signaling channels.1

The portfolio balance channel works through the change in the relative supply of domestic and foreign currency assets (Kouri 1976). If both types of assets were perfect substitutes (that is, if UIP held), then that change in relative supply would not matter. To the extent that assets are imperfect substitutes, however, investors will demand a premium for holding more of the asset whose supply has increased, thus depreciating the currency of that asset. This portfolio balance channel may indeed have played a small quantitative role in advanced economies, where the magnitude of interventions was very small compared with their large bond markets. For example, the average coordinated intervention operation in support of the US dollar from January 1985 to December 1988 involved $278.5 million, while the average coordinated sale involved $373.2 million (Frankel and Dominguez 1993).2 However, in many emerging markets, the stock of foreign exchange reserves is of a similar order of magnitude to the stock of domestic currency assets (Figure 2.1). The magnitudes involved suggest that the cumulative quantitative effects on asset prices through this portfolio channel could be significant, even if the channel has limited traction (see Chapters 4 and 5 for evidence on the effectiveness of foreign exchange intervention). Traction could also be higher if the emerging markets are not as well integrated into the global financial markets as their advanced counterparts (so that local and foreign currency assets become less perfect substitutes).

A few recent theoretical papers introduce frictions in otherwise standard models, which allow intervention to have traction.3 For example, Benes and others (2015) present a model in which sterilized interventions lead to deviations from UIP through portfolio effects in a standard New Keynesian framework. Portfolio effects drive deviations from UIP in Gabaix and Maggiori (2015). Cavallino (2015) builds on that model to show how foreign exchange intervention can have sizable effects.4 Ghosh, Ostry, and Chamon (2016) present a reduced-form model that illustrates the type of frictions that can allow foreign exchange intervention to play a role. UIP implicitly assumes that capital flows would immediately move to arbitrage away any expected return differential. Suppose instead that capital flows respond to return differentials, but at a finite pace:

$\begin{array}{cc}\mathrm{\Delta }{k}_t={\gamma }_i(i_t-i_t^{*}+E_t\mathrm{\Delta }{e}_{t+1})-{\gamma }_k{k}_{t-1},& \left(2.2\right)\end{array}$

where k stands for capital flows and γ < 1 . Thus, the standard balance of payments equilibrium condition:

$\begin{array}{cc}\mathrm{\Delta }{k}_t+\mathrm{\Delta }CurrentAccou{t}_t(Output,e_t)=\mathrm{\Delta }\mathrm{Re}serve{s}_t& \left(2.3\right)\end{array}$

implies that both quantities (capital flows and reserves) and prices (interest rates and exchange rate) matter. That is, foreign exchange intervention impacts the exchange rate, even if interest rates remain unchanged.5 Despite imperfect capital mobility, the foreign exchange market can always clear, provided that a sufficiently large adjustment in asset prices brings demand and supply of foreign exchange in line with each other. However, this adjustment may require very large swings in asset prices, including the exchange rate, which may be undesirable for several reasons. Central bank purchases or sales of foreign exchange assets can help narrow the magnitude of this adjustment by reducing the amount of excess supply or demand that needs to be accommodated by the private market.

The signaling or expectation channel affects the exchange rate through a change in market expectations about fundamentals (Mussa 1981). If the central bank has more information about fundamentals (including its future monetary policy stance) than the market has, it can use intervention to signal that information. And to the extent that it signals information about the future monetary policy stance, such an intervention would have traction on the exchange rate when announced, even if UIP holds (since future interest rates would impact today’s exchange rate via their effect on the expected future exchange rate).6

The rest of this chapter discusses four main reasons that the central bank may want to intervene. As noted earlier, these include precautionary reserve accumulation, intervention to attenuate financial stability risks, intervention on concerns of pass-through to inflation, and intervention for managing more persistent shocks.

International Reserve Accumulation for Precautionary Reasons

Perhaps the least controversial motive for intervening is the need to accumulate reserves for precautionary reasons. This has gained prominence especially since the 1997 Asian financial crisis. With this motive, the central bank intervenes to build up international reserves for use if adverse conditions materialize in the future, and not to affect current developments in the foreign exchange market.

Several episodes of sudden stops or reversals in private capital flows have occurred in the past, some associated with full-fledged currency and financial crises, especially in economies with fixed exchange rate regimes. Intervention during such times of distress can help attenuate overshooting and other disorderly conditions that may arise in foreign exchange markets.

Furthermore, by amassing an adequate stock of reserves, the central bank can reduce the likelihood of adverse conditions materializing in the first place. For example, investors may be less likely to flee if they have confidence that the central bank can step in and help stabilize conditions in the foreign exchange market if a sudden stop takes place. Chapter 4 discusses these and other considerations when reviewing the effectiveness of intervention.

Foreign exchange reserves are among the main indicators of vulnerability that emerge from the early-warning-model literature. Most of that literature, particularly if published before the global financial crisis, focused on assessing the vulnerability to currency crises.7 Those currency crises were typically defined either based on sufficiently large nominal and real movements in the exchange rate or on indices of currency market pressure, which typically included reserves. The early warning models were inspired by the emerging market crises of the 1990s, such as the Mexican peso crisis of 1994. Extensive reviews of that literature are provided in Kaminsky, Lizondo, and Reinhart (1998), Hawkins and Klau (2000), Abiad (2003), and Frankel and Saravelos (2012). The latter performs a meta-analysis based on those reviews and other recent studies. Foreign exchange reserves are among the most frequent statistically significant indicators in the 83 papers they reviewed.8

From a theoretical perspective, reserves play a central role in currency crises models. Their depletion (because of unsustainable macroeconomic policies) is at the heart of “first-generation” currency crises models (such as Krugman 1979). The level of reserves is also a key determinant of whether a bad equilibrium can exist in “second-generation” models (such as Obstfeld 1996) and its global-game variants (such as Morris and Shin 1998). And sufficiently large reserves can in principle address the vulnerabilities created by the balance sheet effects in “third-generation” models (Aghion, Bacchetta, and Banerjee 2014).

Much of this literature focuses on how reserves can prevent a currency crisis, especially when the starting point was a fixed or tightly managed exchange rate regime. However, reserves can still bring prudential benefits, even in the context of a floating exchange rate. In principle, sufficiently large movements in the exchange rate can bring supply and demand for foreign exchange in line with each other following a shock. But large movements may involve economic costs and dislocations, which make them undesirable for several reasons. By intervening in the foreign exchange market, the central bank can reduce the excess demand or supply that needs to be satisfied by the market.

More generally, by accumulating foreign assets, the central bank can smooth the contraction in consumption following a sudden stop in capital flows. Jeanne and Rancière (2011) present a model in which policymakers choose a level of reserves to insure against a sudden stop. The optimal level of reserves depends on their cost, the probability of a sudden stop, its impact on output and consumption, and the degree of risk aversion. Their calibration found the stock of reserves to be adequate, on average, in Latin America, although they struggled to explain the continued rise in reserves over the past 10 years. Obstfeld, Shambaugh, and Taylor (2010) argue that higher reserves can be justified if they insure against domestic financial risks, including capital flight.

Low international reserves are typically a sign of vulnerability, in particular in economies with strongly managed exchange rates. However, the sharp depreciations observed in several Latin American countries after the global financial crisis led to neither high inflation nor to disruptive contractions, as they had in the past. Strong macroeconomic frameworks on the back of ample reserves likely contributed to this resilience.9

Even though the central bank may not necessarily be focusing on current market conditions when accumulating reserves for precautionary motives, it can still time its foreign exchange purchases in an opportunistic way. It can purchase its foreign exchange during periods when the domestic currency is appreciating. By leaning against the wind, it can moderate the pace of appreciation and make purchases when foreign exchange is perceived to be relatively “cheap.”

Central banks in the region have pursued this strategy, including through explicit rules. For example, when Colombia and Mexico were building up their reserves, they used options with a strike price based on the 20-day moving average of the exchange rate. That helped time their foreign exchange purchases to take place when appreciation pressures were stronger. This was also the case in Chile in 2011, when authorities announced a year-long program of international reserve accumulation to match reserve-to-GDP ratios of similar countries.

The pace of prudential reserve accumulation could pick up during episodes of capital inflows for other reasons as well. If these episodes are associated with an increase in short-term foreign exchange debt, then the pace of accumulation should increase to keep up with adequacy metrics that include that type of flow.10 The flip side of that argument is that episodes in which short-term foreign exchange debt decrease would imply a reduction in the desired prudential level of reserves. In practice, though, countries are reluctant to deploy reserves. When prudential motives abate, they tend to adjust by halting reserve accumulation, which, over time, can bring the stock of reserves in line with reduced prudential needs.

The level of international reserves appears to be adequate in Latin America. Using either the IMF’s adequacy of reserves accumulation (Figure 2.1) or the ratio of international reserves to short-term debt (the so-called “Guidotti-Greenspan rule”), in all cases the level of international reserves for the Latin America 5 (LA5) countries seems appropriate (Figure 2.2). In the case of Peru, however, the stock of reserves seems large by both metrics.11

A number of challenges are involved in assessing whether reserves are excessive. To begin with, the probability of a sudden stop may be a function of the level of reserves; for example, investors that would otherwise “rush to the exits” in an adverse shock may decide to keep their positions if they feel the central bank has enough reserves to smooth that shock. This seems to be an empirically relevant channel, as shown by the predictive power of reserves in the early warning literature, as discussed earlier. Alternatively, the presence of reserves may encourage risky liability structures; for example, borrowers taking on short-term external debt because they can count on the central bank to provide foreign exchange liquidity if they were to face tighter global financial conditions (Kim 2008).

More generally, much of the benefit of reserves stems from the option of deploying them in the event of distress (rather than from their actual deployment). In game theoretical terms, their use “off the equilibrium path” can bring many benefits, even if reserves are not deployed. The majority of countries in Latin America have not experienced a major homegrown crisis since the early 2000s. The decline in currency mismatches, and the large stock of reserves accumulated, certainly played a significant role in building that resilience. However, even if the prudential benefits of reserves are very large, they are likely subject to diminishing returns. For example, in a standard buffer–stock savings model, an additional dollar buys less and less in terms of consumption insurance at the margin. Similarly, the prudential benefits of accumulating reserves past an adequacy level are likely to decline at the margin (whereas the moral hazard effects on private sector risk-taking behavior may not).

Intervention to Attenuate Financial Stability Risks

Advanced economies with floating exchange rate regimes often have a “benign neglect” view of the exchange rate. This is supported by a long history of exchange rate swings, sometimes sizable, without adverse effects for financial stability.

Among emerging market central banks, however, financial stability concerns feature much more prominently. They typically dislike sharp movements in the exchange rate, particularly those that involve a sharp depreciation. Currency mismatches on corporate and financial balance sheets are a major source of financial fragility. They played a central role in currency crises in the region, including high profile cases such as Mexico in 1994, Brazil in 1999, and Argentina in 2001. When currency mismatches are present, sharp movements in the exchange rate can easily render a borrower insolvent—including the government. And if mismatches are present in the financial sector, the shock can easily gain a systemic dimension. These mismatches have declined over time, due to tighter financial supervision and regulation, and a greater awareness of the risks involved among borrowers. But there is a genuine fear that pockets of vulnerability may emerge during times of distress. For example, large firms suffered heavy losses in Brazil and Mexico because they used complex foreign exchange derivative products.

In principle, a depreciation that is not warranted by perceived fundamentals could be self-correcting, to the extent that the overshooting of the exchange rate increases the expected returns in local currency. That should entice investors to keep, or even increase, their local market exposures. However, in practice, it is feared that sharp depreciations can create adverse dynamics in the foreign exchange market, beyond what is warranted by fundamentals. The exchange rate’s automatic stabilizer role may thus break down, resulting in disorderly conditions, as discussed in IMF (2015). Several factors can contribute to adverse dynamics in the foreign exchange market, including the fears of unknown currency mismatches (that is, even if mismatches are small, investors may still flee because they believe, or expect that other investors believe, that mismatches are potentially serious).

The case for intervention under those circ*mstances is fairly uncontroversial, unless the extent of intervention is perceived to be excessive; that is, if the central bank is perceived to be resisting the movement to a new equilibrium.12 We should bear in mind that it is very difficult to assess the equilibrium exchange rate in real time, and whether movements, even if sharp, should be considered “excessive.” In the limiting case, excessive intervention could become unproductive if it facilitates capital flight that would not otherwise take place under a more depreciated exchange rate. In general, there can still be a case for some intervention to help smooth the impact of a permanent shock if that helps prevent financial stability risks from materializing. However, the central bank should remain mindful of the moral hazard that this can create (such as encouraging excessive risk-taking behavior by the private sector because of an implicit “put”).

Conversely, central banks may also intervene to slow the pace of appreciation if they fear that it is moving the exchange rate away from fundamentals and setting the stage for an eventual correction that could be disruptive. This type of intervention is discussed in more detail later in the chapter.

Inflation Pass-Through

Another motive for intervening is the concern of pass-through to inflation. Exchange rate pass-through has declined over time, as central banks in the region have established their credibility—and despite marked increases in import ratios (see Carriere-Swallow and others 2016). However, the sheer size of a sharp depreciation can still have nonnegligible effects on inflation, even under a small rate of pass-through. Moreover, there may be concerns that the effects may be nonlinear, and become stronger after a large depreciation, with some threshold level of depreciation after which the pass-through increases.

Also, the exchange rate remains an important focal point people use to assess the strength of the economy. A large depreciation may adversely affect confidence and price- or wage-setting behavior, even in the nontradable sector. If the shock to the exchange rate is permanent or highly persistent, the economy will need to cope with and adjust to it. If a large swing in the exchange rate is perceived to be temporary, then there could be a case for using foreign exchange intervention to counter that overshooting and to limit its impact on inflation, inflation expectations, and relative prices more generally—which could in fact affect resource allocation.

This motive for intervention is less controversial, to the extent that it is consistent with the monetary policy objective (meeting the inflation target) and is driven by the response to a perceived temporary shock to foreign exchange markets. The case for using intervention is stronger if it has a more immediate effect on the exchange rate and avoids the need to adjust the policy rate in response to exchange rate developments (central banks typically want to adjust the policy rate gradually and predictably, making it a less suitable instrument to respond to high-frequency fluctuations). However, inflation-targeting countries need to credibly highlight that foreign exchange intervention is subordinated to interest rate policy to avoid misperceptions and potential confusion about the central bank’s objective (which, if not well articulated, could be a significant cost of intervention).

Managing more Persistent Shocks

Typically, sustained intervention is associated with capital inflows, since the fear of running out of reserves (or seeing them drop below prudential metrics) eventually limits the willingness to sell foreign exchange. However, the question of how to respond to persistent depreciation pressures will likely become more and more pertinent if the global financial cycle reverses, and countries experience sustained capital outflows. This is somewhat uncharted territory, and the discussion is left for Chapter 6.

One of the main concerns in the face of persistent capital inflows is the loss of external competitiveness and Dutch disease considerations. These concerns tend to be labeled as mercantilist. It should be acknowledged, however, that it can be quite costly for an economy to adjust rapidly to the new equilibrium exchange rate, and for workers to move from the tradable to the nontradable sector and back, following a persistent appreciation that eventually reverses. Tradable firms may be credit constrained, go out of business, and only slowly be replaced by new entrants once the cycle reverses. The presence of currency mismatches in nontradable sector firms will only compound such a problem.

Sustained capital flows can also exacerbate prudential concerns. While the risks of capital flows are typically associated with “hot money” flows that can quickly reverse, persistent inflows can fuel credit and asset price booms, which often result in crises. These are of concern for countries with shallow financial markets. Persistent flows may be even riskier, since the longer the climb, the larger the potential fall. Risks can be amplified if the domestic financial system does not allocate the easy money toward productive uses, and instead uses it to finance consumption booms or asset price bubbles.

Despite these legitimate concerns, it is not clear whether intervention is an adequate tool to manage persistent shocks, for several reasons. First, it becomes harder to make the case that intervention is used to prevent an overshooting of the exchange rate, as opposed to resisting the movement to a new equilibrium. There are also concerns that intervention may become less effective over time. By smoothing the shock to the exchange rate, the central bank may encourage more inflows during the boom phase (as investors expect the exchange rate to continue to appreciate, which increases their expected gains). The opposite is also true, and intervention may stimulate outflows when the capital flow cycle turns (as investors want to take advantage of the delayed adjustment to flee at a more favorable exchange rate). To the extent that intervention is perceived to be costly, it may become a less suitable tool for dealing with permanent shocks, even in models where it continues to have traction (Ghosh, Ostry, and Chamon 2016). In these circ*mstances, there is a stronger case for adjusting the macro policy stance in response to the shock. Intervention may play a supporting role, at best. A full-fledged discussion of how to manage capital flows is beyond the scope of this chapter, which focuses only on intervention. For that discussion, please refer to IMF (2012).

Conclusion

This chapter discusses motives for foreign exchange intervention under a flexible exchange rate regime.13 It emphasizes that even if intervention seems desirable, its cost must also be considered, and the exact nature of those costs remains a subject of debate.

Many point to the interest rate differential as a measure of the cost of holding reserves. Yet that does not give a complete picture, as it fails to consider the change in the exchange rate, which can make intervention even more costly due to the forward premium puzzle (in which the higher interest domestic currency would tend to appreciate). Likewise, to the extent that the central bank leans against the wind, and intervenes when the exchange rate overshoots, the resulting valuation effect may reduce the costs. Furthermore, the interest rate differential fails to factor in differences in credit risk. Perhaps more important, it abstracts from the fact that international reserves can reduce risk premiums, not only for sovereign borrowers, but also for corporate and financial borrowers.14

Although settling this debate is beyond the scope of this book, it seems reasonable to assume that these costs are not minimal. Authorities would therefore typically need to make a compelling case for the benefits of intervention.

The policy framework and the policy mix can also influence the adoption of alternative countercyclical policies, including foreign exchange intervention. For example, an economy that is well integrated with global financial markets and is experiencing overheating may fear that raising interest rates to cool domestic demand could stimulate larger capital inflows. These inflows can fuel domestic credit expansions and stimulate demand. They can also contribute to asset price inflation—or even unsustainable asset price bubbles—and inflationary pressures. This would result in appreciation pressures, which could drive the policymaker to consider foreign exchange intervention, along with macroprudential policies—or even capital controls as a more frequent instrument among the usual elements in the policymaker’s toolkit. Chapter 6 revisits these issues.