Chapter 12 Peru: Foreign Exchange Intervention under Financial Dollarization

Adrián Armas

Author:

Adrián Armas

0000000404811396

https://isni.org/isni/0000000404811396 International Monetary Fund

Search for other papers by Adrián Armas in
Current site
Google Scholar
PubMed

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; monetary policy design; exchange rate volatility; building monetary policy credibility; floating exchange rate regime; Exchange rate arrangements

Download PDF (448 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 explores the case for foreign exchange intervention in Peru. The Central Reserve Bank of Peru abandoned the exchange rate peg in August 1990 and moved to a managed floating exchange rate regime. Since its inception, the floating regime has been subject to intervention by the central bank to smooth out exchange rate volatility. One key element of the motivation by the Central Reserve Bank of Peru to intervene in the foreign exchange market is the fact that Peru has a high degree of financial dollarization, although on a clearly decreasing path. The Peruvian case provides a long experience of discretionary-type intervention, based on daily assessments of foreign exchange market conditions. This type of intervention has enabled the Central Reserve Bank of Peru to run a countercyclical monetary policy, in particular, during critical periods such as the global financial crisis and the latest episode of quantitative-easing tapering that coincided with a strong fall in commodity prices (2013–15).

Introduction

Foreign exchange intervention by emerging market central banks has been a key policy issue in recent decades, especially as these banks have increasingly adopted managed floating exchange rate regimes, moving away from fixed or pegged regimes. This was also true in Peru.

After two major financial crises—the 1997–98 crisis that hit emerging market economies and the 2008 global financial crisis—foreign intervention policies have endured in emerging markets. This fact has renewed analytical interest in academic circles, which had neglected intervention policies because of scant evidence on their effectiveness in developed countries and their theoretical irrelevance in a world of perfect capital mobility (Backus and Kehoe 1989).

The new academic literature instead assumes imperfect capital mobility, as advanced by Cavallino (forthcoming) or Fanelli and Straub (2018), for example. These two reports suggest that optimal intervention policies “lean against the wind,” which is tantamount to reducing exchange rate volatility.¹ Recent evidence on central bank intervention supports these theoretical findings; for example, Daude, Levy-Yeyati, and Nagengast (2016) and Fratzscher and others (2019) provide important evidence on the effectiveness of foreign exchange interventions, and that smoothing exchange rates is the main objective of intervention in emerging market economies.

This chapter examines the Peruvian case of more than 25 years of sterilized foreign exchange intervention under a managed floating exchange rate regime. It delves into issues such as the rationale of Peruvian central bank intervention within its overall monetary policy framework, the discretionary and high-frequency intervention implementation, existing evidence about its effectiveness, and the optimality of leaning-against-the-wind intervention in the face of uncertainty regarding shocks.

The next section explores the policy background and explains the rationale for foreign exchange intervention. The chapter then assesses evidence for the effectiveness of central bank foreign exchange intervention, followed by intervention procedures in Peru, and concluding remarks.

A Managed Float Since 1990

Peru has run a managed floating exchange rate regime since 1990. In August 1990, a series of liberalization reforms was the onset of a stabilization process. The reforms were put in place to tame hyperinflation and reverse a dramatic output collapse that had started in the mid-1970s. The Central Reserve Bank of Peru was first among the region’s central banks (along with the Central Bank of the Dominican Republic) to abandon its pegged exchange rate regime.

During the 1970s and 1980s (and earlier), Peru had many exchange rate pegs, a heavily controlled multiple exchange system, and a closed financial system.

From August 1990 to December 1991, monetary authorities abandoned all exchange controls, restored full convertibility, liberalized current and capital accounts, and established a managed floating exchange rate regime.

Unlike other inflation stabilization episodes, Peruvian stabilization was based on the strict control of monetary aggregates. To make the central bank credible, it was crucial to grant it full independence and to break free of the fiscal dominance of policy that was prevalent in the two previous decades. This meant, for example, that any form of central bank credit to the government was forbidden. At the time, exchange rate–based stabilization was ruled out because of the scarcity of international reserves, and because a series of exchange rate–based stabilizations had failed in the 1980s.²

When the 1990s began, reserves were almost depleted. After the stabilization, the central bank was committed to remonetize the economy after a fall in the demand for domestic currency and currency substitution during hyperinflation. During and after inflation stabilization, a natural way of remonetizing was the purchase of foreign currency through direct spot interventions. As long as stabilization was credible, the increasing domestic currency demand reflected a private sector portfolio shift from foreign to domestic currency. At that time, given that the stabilization program was based on avoiding internal borrowing by the central government, no other liquid assets were available that were suitable for monetary operations, including traditional sovereign bonds.³

Nonetheless, building monetary policy credibility was a slow and steady process: it took all of the 1990s to bring inflation down to international levels. This disinflationary process through the control of money aggregates was impaired by increasing instability in the relationship between monetary aggregates and inflation.⁴

Despite the reduction of inflation levels, dollarization remained high, as other conditions were not yet in place at the time (such as anchoring long-term inflation expectations by adopting an inflation-targeting regime).⁵ Figure 12.1 shows the high degree of dollarization in the banking system during the 1990s. It is important to note that firms both in the tradable and nontradable sectors had dollarized assets and liabilities, and in most cases, the currency mismatch was high. In addition, an important share of government debt was denominated in US dollars.

The existence of financial vulnerabilities resulting from currency mismatches called for intervention to (1) accumulate reserves to provide international liquidity when needed, and (2) minimize sharp and unexpected exchange rate depreciations, which would bring about deleterious effects of currency mismatches in the economy.

Figure 12.2 describes the evolution of net international reserves and the behavior of the Peruvian nuevo sol (sol hereafter) during the 26 years since 1992, after hyperinflation. The figure shows the massive growth of international reserves, from virtually zero to around $60 billion. Shaded areas in the figure depict episodes of falling reserves—which are tantamount to periods of foreign capital outflow and exchange rate depreciation.

This episode of initial international reserve accumulation ended in mid-1998 with the advent of a series of financial and other crises that affected emerging market economies across the globe: the Asian crisis in 1997, an extreme El Niño phenomenon in Peru in early 1998, and the Russian default in September 1998. The latter was the most damaging of these negative shocks, as it brought sudden, short-term capital outflows that froze the credit market and forced the closure or merging of half of the banking system.⁶ The exchange rate depreciated 18 percent from June 1998 to September 1999, producing a strong contraction in an economy with 82 percent of loans denominated in foreign currency (Figure 12.1).

The other shaded areas in Figure 12.2 correspond to the global financial crisis of 2008–09 and the period of monetary policy tapering in the United States, which started with the taper tantrum in May 2013. During the global financial crisis, the sol depreciated 18 percent; during the tapering period, it depreciated as much as 35 percent (Table 12.1).

^{Table 12.1.}

Selected Episodes of Foreign Exchange Intervention

Source: Central Reserve Bank of Peru.

^{Table 12.1.}

Selected Episodes of Foreign Exchange Intervention

	Emerging Market Spillover Crisis (Asia, Russian Federation) and El Niño Phenomenon	Global Financial Crisis	Tapering and Strong Fall of Commodity Prices
Year of initial shock	1997	2008	2013
Depreciation	18% (Jun. 1998-Sept. 1999)	18% (Apr. 2008–Feb. 2009)	35% (Apr. 13–Feb. 2016, 34 months)
Credit dollarization	82%	52%	42%
Average GDP growth during shock and 1 year later	2.5% (1997–2000)	6.2% (2008–10)	3.9% (2013–16)
Net purchases of interna-tional reserves 1 year before initial shock	2.9% of GDP (1996)	7.1% of GDP (2007)	44.4% of GDP (2012)

Source: Central Reserve Bank of Peru.

^{Table 12.1.}

Selected Episodes of Foreign Exchange Intervention

	Emerging Market Spillover Crisis (Asia, Russian Federation) and El Niño Phenomenon	Global Financial Crisis	Tapering and Strong Fall of Commodity Prices
Year of initial shock	1997	2008	2013
Depreciation	18% (Jun. 1998-Sept. 1999)	18% (Apr. 2008–Feb. 2009)	35% (Apr. 13–Feb. 2016, 34 months)
Credit dollarization	82%	52%	42%
Average GDP growth during shock and 1 year later	2.5% (1997–2000)	6.2% (2008–10)	3.9% (2013–16)
Net purchases of interna-tional reserves 1 year before initial shock	2.9% of GDP (1996)	7.1% of GDP (2007)	44.4% of GDP (2012)

Source: Central Reserve Bank of Peru.

During these periods of volatility induced by global forces, foreign exchange interventions had been important to moderate excess fluctuations in the exchange rate. The key rationale for intervention was the avoidance of financial vulnerabilities stemming from the heavy dollarization of the financial system.

The literature on financial stability points to the exchange rate as a key asset price that can trigger financial crises.⁷ As such, foreign exchange intervention to tame exchange rate volatility has been part of monetary policy design and requires high international reserves to allow purchases or sales on demand. In tranquil times, with capital inflows, purchases of foreign currency can dampen currency appreciation, and at the same time, build international reserves for harder times.

High international reserves have allowed more active foreign exchange interventions to face negative external shocks, such as the global financial crisis, US monetary policy tapering, and the sharp decline of commodity prices around this period (which continued until February 2016).

A remarkable period of intervention started in 2003 and ran through 2007, when US dollar purchases reached 9.6 percent of GDP (Figure 12.3). This unprecedented reserve accumulation was abruptly followed by the global financial crisis. Intervention was so large that it averted a recession that, otherwise, would have been like the one during the 1998–99 emerging economies crises.

Because figures represent yearly flows, they mask a huge selling period from June 2008 to February 2009. In that time, $7.1 billion (about 6 percent of 2009 GDP) were sold to avoid triggering negative balance sheet effects. After the global financial crisis, capital inflows and reserve accumulation resumed, which ended by the reversal of capital flows brought about by the tapering of quantitative easing in the advanced economies.

That tapering period is another remarkable episode of leaning against depreciation winds, with interventions of almost 5 and 6 percent of GDP in 2014 and 2015, respectively.

Overall, the long history of intervention episodes in Peru shows distinct features. In the early 1990s, a key driver of intervention was the accumulation of international reserves, which primarily reflected domestic currency remonetization. The period ends with the financial crisis in the emerging market economies at the end of the 1990s. Thereafter, a period of capital inflows into emerging markets took place; as a result, sizable international reserves were accumulated. In this period, US dollar purchases had a precautionary motive—that is, the increase of buffer stocks to face capital flow reversals.

Motivation for Intervention

The Central Reserve Bank of Peru follows an inflation-targeting regime in conducting its monetary policy. In other words, it uses its main policy interest rate to gear the inflation rate toward the inflation target. Inflation targeting is flexible in the sense of Svensson (2000), because it allows deviations from the inflation target whenever shocks do not move long-term inflation expectations away from target. These instances occur, for example, when short-lived supply shocks hit. Failing to conduct flexible inflation targeting would mean unnecessary output volatility.

The monetary policy framework considers the dynamics of output and its macroeconomic determinants. Financial dollarization is a key force behind the contractionary effects of exchange rate depreciation. In short, the contractionary effects are due to currency mismatches in the balance sheet of nontradable firms and households. Sometimes these contractionary effects can be larger than the usual textbook expenditure-switching effects, which are expansionary. Periods of sharp and large depreciation can imply overall contraction, because of large-scale balance sheet effects.

As such, monetary policy design includes a set of tools to control dollarization risks in a financially vulnerable economy (Figure 12.4). One of these tools is sterilized foreign exchange intervention, which induces high-frequency portfolio shifts across currencies, but it does not affect the level of domestic currency liquidity consistent with the policy rate. By reducing exchange rate volatility, foreign exchange intervention minimizes the risk of triggering perverse balance sheet effects, which would otherwise unleash excess output volatility.

Actions and communication by the central bank have allowed agents to understand the monetary policy design—in particular, the distinction between policy rate moves for inflation forecasting and foreign exchange intervention to smooth exchange rate volatility at high frequencies.

There are other ways to frame monetary policy design in Peru, apart from the usual flexible inflation-targeting regime. For example, it is said that monetary policy is usually made in a highly uncertain environment, and monetary policy should therefore respond not only to the most likely outcome but also to risk scenarios.

In Peru, the risk management approach is based on uncertainty about dollarization risks. To avoid low risk and the potentially damaging outcomes arising from balance sheet effects, the central bank uses sterilized foreign exchange interventions, as depicted in Figure 12.4.

Two main points arise in the conduct of such risk management policy. First is the usual moral hazard discussion. By intervening, the central bank provides public insurance, which induces private agents to take even more risks or to not seek private sources for hedging their financial risks. It is important to highlight that any policy option taken by central banks is seldom neutral: it is often argued, for example, that by reducing interest rates, monetary authorities induce more risk taking by banks. The approach of the Central Reserve Bank of Peru has been to foster hedging alternatives, mostly at the micro level, but to never neglect its macroeconomic stability mandate.

A pure free float is probably not credible for a central bank in an economy with high currency mismatches, because monetary policy always wants to avoid extreme risks to the economy (a lesson from the global financial crisis). As the extent of currency mismatches declines, the exchange rate could become more flexible (point B₁ closer to point A₁ in Figure 12.5); however, this process of diminishing currency mismatches will take time.

The second point in the risk management framework relates to the accumulation of enough international reserves to conduct intervention operations of either a positive or negative sign. High international reserves overcome a key problem that financially dollarized economies face—namely, the absence of a lender of last resort denominated in US dollars. Therefore, high international reserves provide a self-insurance mechanism to the economy. Like any insurance mechanism, its optimality depends on policymakers’ risk aversion, the history of balance of payments crises, and the shortage of international liquidity. In Peru, monetary authorities have taken a conservative approach to reduce the risk of economic crises, given the historical lessons delivered by economic setbacks of the past, in particular, hyperinflation and the full loss of international liquidity at the end of the 1980s.

In the Peruvian case, the argument for foreign exchange intervention to face episodes of real exchange rate misalignment has not been relevant. Only two episodes of important misalignment have occurred in the past 25 years. The first was observed before the financial crises in emerging market economies during 1997–98 and the second before the taper tantrum in 2013.

The overall perspective of risk management uses a set of tools summarized in Table 12.2. Both high international reserves and foreign exchange interventions are used to dampen shocks (ex ante) and confront shocks (ex post) in the face of risk or crisis events.

^{Table 12.2.}

Macro Risk Management Tools

Source: Central Reserve Bank of Peru.

^{Table 12.2.}

Macro Risk Management Tools

Types of Policies	Risks Internalizing	Shocks Dampening	Shocks Confronting
High reserve requirements on US dollar liabilities	Yes	Yes	Yes
High international reserves		Yes	Yes
Foreign exchange intervention		Yes	Yes

Source: Central Reserve Bank of Peru.

^{Table 12.2.}

Macro Risk Management Tools

Types of Policies	Risks Internalizing	Shocks Dampening	Shocks Confronting
High reserve requirements on US dollar liabilities	Yes	Yes	Yes
High international reserves		Yes	Yes
Foreign exchange intervention		Yes	Yes

Source: Central Reserve Bank of Peru.

When a shock hits, it is uncertain whether it is transitory or permanent. The history of shocks can reveal the normal ups and downs in prices and quantities, but it is not a good guide for disruptive financial crises, which, almost by definition, are unforecastable. Therefore, foreign exchange interventions that lean against the wind may be optimal for a financially dollarized economy.

To illustrate leaning against the wind, Figures 12.5 and 12.6 show a case of negative external shock, such as a taper tantrum or falling terms of trade. Point A₁ in Figure 12.5, panel 1, is the result of a depreciation in the fully floating case. The lower point, B₁, is the result of a leaning-against-the-wind policy. When the shock hits, there is uncertainty not only about the duration of the shock (permanent versus transitory) but also about whether exchange rate depreciation is expansionary or contractionary.⁸

If the shock turns out to be permanent, the exchange rate would remain at point A₂ in the future, as depicted in panel 2. Leaning against the wind means that the exchange rate will eventually achieve this A₂ level, but at a slower pace. The small cost of the leaning-against-the-wind policy is that the exchange rate is not working as quickly as a shock absorber provided by the fully floating case. However, the benefit may be strong, because the central bank avoids triggering harmful balance sheet effects.

The lower exchange rate change to B₁ brought about by foreign exchange intervention depends, of course, on the effectiveness and extent of intervention. In turn, the size of intervention depends on the assessment of the shock; is it perceived as permanent or transitory? How mismatched are household and firm balance sheets? How fundamental is the shock? Floating exchange rates also tend to overshoot, not only for the reasons described in Dornbusch (1976) on the adjustment of sticky prices but also because of agents’ panics and irrational exuberance.⁹

Therefore, the passage from point B₁ to B₂ (equal to A₂) lets the exchange rate achieve its new fundamental equilibrium, but slowly, avoiding unnecessary jumps that may wreak havoc within a financially dollarized economy.

On the other hand, if the shock turns out to be transitory, the exchange rate would return to the pre-shock level depicted in Figure 12.6. Here the cost of the leaning-against-the-wind policy is zero, because there is no need for the exchange rate to work as a shock absorber. Nevertheless, the gain can be sizable, because the central bank avoids excess volatility.

Regarding the overall policy framework, the central bank monetary policy analysis made through its formal quantitative model takes into account the aforementioned causes and effects.¹⁰ For example, the evolution of the exchange rate considers equations such as

\begin{matrix} s_{t, t + 1} = (1 - c_{e}) E_{t} [s_{t + 1}] + c_{e} s_{t - 1} & (12.1) \end{matrix}

and

\begin{matrix} i_{t} = i_{t}^{*} + (s_{t, t - 1}^{e} - s_{t}) + p r e m_{t} + ξ_{t}, & (12.2) \end{matrix}

where $s_{t, t + 1}^{e}$ is the expectation of the next-period (t + 1) exchange rate based on information available in the current period. However, this expectation is not purely rational. Instead, it comprises a fully rational component (E_t[s_{t + 1}]) and an inertial component (s_{t - 1}). Agents know that the central bank is always up for intervention whenever an important shock hits and see that the size of intervention is large relative to market turnover; therefore, the evolution of the exchange rate is well explained by sluggish expectations, as shown in equation (12.1). Parameter c_e measures how effective and important foreign exchange interventions are to induce inertia in exchange rate expectations.

In equation (12.2), i_t is the domestic policy rate, $i_{t}^{*}$ is the foreign policy rate, prem_t is the risk premium that corresponds to assets denominated in domestic currency, and ξ_t is a shock to the nonarbitrage equation.

When equation (12.2) is solved forward, and exchange rate expectations are substituted by equation (12.1), the spot exchange rate s_t depends on its past value and the rational expectations of current and future interest rate differentials, current and future values of risk premiums, and current and future values of shocks. Again, the higher the c_e, the more inertial the spot exchange rate.

On the other hand, the assessment of the effects of exchange rates relies heavily on the evolution of the output gap:

\begin{matrix} y_{t} = ... + α_{q m} (q_{t}^{m}) - α_{q b} (Δ q_{t}^{b}) + ..., & (12.3) \end{matrix}

where y_t is the output gap relative to trend GDP, $q_{t}^{m}$ is the effective multilateral real exchange rate gap, and $Δ q_{t}^{b}$ is the bilateral (relative to the United States) exchange rate depreciation. We abstract from all remaining terms in the output gap equation. Equation (12.3) shows the standard expenditure-shifting effect associated with a higher effective real exchange rate, together with the negative balance sheet effect linked to the bilateral real exchange rate depreciation.

If parameter α_qb is small relative to α_qm, then the overall effect of an exchange rate depreciation is positive. In contrast, when α_qb is large, the overall contractionary effect prevails. Furthermore, it is possible to have nonlinearities in the above equation if α_qb is a time-varying parameter that depends on exchange rate jumps. If the exchange rate change is small, α_qb may also be small and constant, but when the jump is large, α_qb may also change to a higher level.

Overall, the overarching motive for foreign exchange intervention in Peru is to avoid excess exchange rate volatility that would trigger negative financial and real effects in the economy.

Effectiveness of Intervention

Following the lean-against-the-wind approach explained in the previous section, a first glimpse of the effectiveness of intervention compares the sol with other currencies in the region.

Figure 12.7 shows the evolution of the Peruvian sol, together with the Brazilian real and the Colombian peso, all against the US dollar. The sol and the real are on the left axis because of their comparable levels; the peso/US dollar are on the right. These currencies tend to move in the same direction because of global factors. The sol moves concurrently with other currencies but at a milder pace, suggesting that daily intervention operations effectively influence the daily spot rate.

The smoother sol/US dollar path can be verified by computing cumulative changes during the different episodes of the past 15 years (Table 12.3). As can be seen, Peru’s cumulative appreciation or depreciation rates have been much lower.

^{Table 12.3.}

Appreciation and Depreciation of Currencies at Comparable Episodes (Cumulative percentage changes)