Should Brazil’s central bank be selling foreign reserves?

Márcio Garcia, 25 September 2013

a

A

The US dollar’s rise in August and the Brazilian Central Bank’s (BCB) interventions in forex markets have started a debate about whether the BCB should keep on intervening as it has been doing, mostly via currency derivatives markets, or if it should also be selling its international reserves.

The objective of Brazil’s foreign-exchange intervention

Before discussing the best way for the BCB to intervene, it is necessary to establish the intervention’s goals. Since May, when the Federal Reserve announced it would taper off QE3, capital has been flowing back from emerging markets to developed economies. The reduction of capital inflows, together with limited prospects for Brazilian exports, explain the depreciation of the Brazilian real, together with most other emerging markets’ currencies. In this context, the BCB’s interventions may only smooth the exchange-rate path towards its new, more depreciated, equilibrium.

The forex interventions are not meant to establish a floor for the exchange rate, but to provide the needed liquidity for the depreciation to take place without excess volatility and overshooting – which may entail unnecessary economic costs. If the goal were to establish a limit to the depreciation, the BCB would have to use monetary policy, significantly raising the interest rate, as it used to do before the floating-exchange-rate regime was adopted in 1999.

Methods of intervention

Having established that the forex intervention’s goal is to provide liquidity to the currency markets, we may deal with the following question – which is the best way to intervene? Past currency interventions in Brazil have been conducted through three instruments:

  • Currency swaps sales (equivalent to sales of US dollar futures).
  • Auctions of US dollar credit lines.
  • Sterilised sales (those that do not affect the interest rate) of international reserves (Garcia 2013).

In the most recent episode, the BCB has not used the latter option, which has led to a debate about whether or not to return to sterilised sales of reserves.

Brazilian foreign-exchange markets

First, some background on forex markets in Brazil and BCB interventions is in order. The Brazilian real is a nonconvertible currency. Foreign-currency deposits at Brazilian banks are not legal. Derivatives markets in Brazil that trade contracts related to the US dollar settle the trades in real. Likewise, BRL NDFs (non-deliverable forwards of Brazilian real), traded in the US, are settled in US dollars. Because of legal restrictions in the Brazilian forex spot market, a large currency futures market developed at the BM&F Bovespa exchange. As a result, the liquidity of the first-to-mature US dollar futures contract is many times larger than the spot market’s.

The BCB has a long history of intervening in currency derivatives markets. Since crises in Brazil are usually associated with large depreciations of the domestic currency, domestic agents seek yields in foreign currency as a systemic hedge. Also, during crises, speculation increases the demand for yields in foreign currency. Realizing that those yields in foreign currency could be achieved through domestic assets and non-deliverable-currency derivatives, since there was no need for actual exchange rate transactions, the Brazilian government has, for many decades, provided domestic public debt indexed to the US dollar and sold currency swaps (Lopes 2003). Note that this is different from other experiences of interventions in derivative markets by emerging markets’ central banks, as the speculative attack against the Thai baht in 1997 (Lall 1997), since the BCB or the Brazilian Treasury only have to disburse domestic currency, never actual hard currency.

The transition mechanism of foreign-exchange intervention

Second, let us examine the dynamics and interdependence of futures and spot currency markets in Brazil. Imagine that the BCB sells currency swaps, akin to selling US dollar futures. The immediate effect is a reduction of the US dollar futures price, but not of the spot dollar price. This decrease in the difference between futures and spot dollar prices – the forward premium – means that it becomes cheaper to hedge against the real’s depreciation. This opens an arbitrage opportunity for the banks, who borrow short-term dollars from abroad and invest them in cupom cambial. The cupom cambial is the onshore dollar rate implied by interest and currency futures, equal to the domestic interest rate in real minus the depreciation hedge’s cost, the forward premium, i.e. a trading strategy that yields US dollar returns in Brazil, with the settlement being in real.

Figure 1 shows the one-year-onshore-dollar rate, the cupom cambial, and the corresponding one-year US rate on the left-hand scale, as well as the exchange rate measured in real per US dollar (an appreciation of the real is a decrease of the exchange rate) on the right-hand scale.

Figure 1. Brazilian onshore US dollar rate (cupom cambial), US one-year interest rate, convertibility risk, and exchange rate (BRL/USD)

The BCB US dollar futures sale, at first, does not directly affect the spot dollar price. However, by increasing the cupom cambial, it opens a transmission channel to the spot market, via banks, who bring spot dollars to profit from the increase in cupom cambial. This transmission channel is, in general, quite efficient. However, there are two factors that may disturb its operation:

  • The first obstacle is capital controls, which were in place until recently, when the international scenario was one of abundant capital inflows (Chamon and Garcia, 2013). Today, such controls have been removed and no longer disturb the perfect arbitrage between the futures and the spot US dollar markets.
  • The second obstacle, much more harmful, is the general perception that dollar futures contracts may be poor substitutes for spot dollars.

As an example, suppose a company had bought dollar futures, expiring in January 2014, in order to assure the total disbursement, in real, of a US dollar payment in reference to imports or external debt service. The company’s expectation is that the Brazilian futures market, which settles in real, will allow it to fix today the cost, in real, of the US dollar amount needed next January. Normally, it is perfectly reasonable to expect this to happen. Nevertheless, Brazil has experienced severe crises in the past, in which “São Paulo dollars” (dollar futures) became poor substitutes for greenbacks.

Imperfect substitutability of dollar futures

In 2002, during the peak of the crisis before Lula’s first inauguration, the exchange rate reached almost four real per US dollar (now, more than a decade later, it is in the range 2.2–2.3 real per US dollar, see Figure 1). Normally, US dollar futures prices are higher than the spot price. At that time, however, the opposite happened. The US dollar futures price was cheaper than the spot price, i.e. the currency futures market was in backwardation, demonstrating serious doubts about dollar futures.

What could these doubts be?

  • The fear that the new government would default on the public debt.
  • The possibility that, facing a sharp shortage of international reserves, the BCB would impose quota restrictions on foreign exchange transactions.

In the quota restrictions scenario – which has happened in the past – only a few targeted transactions, deemed essential, would be allowed to clear at the official exchange rate (which is used as a benchmark for the clearing of futures contracts). All the remaining forex transactions would be put in a queue (or go to the black market, paying a hefty premium). In this case, the dollar futures contract would be a very poor substitute for the spot one.

Under normal conditions, US dollar futures in Brazil are perfectly arbitraged with the (same maturity) BRL NDF in the US. However, during the 2002 crisis, the US dollar futures price in real became much lower (cheaper) than the US dollar price in real implied by the BRL NDF. This premium is called convertibility risk. Figure 1 also displays the convertibility risk, together with the onshore US dollar rate implied by interest and currency futures, the cupom cambial. While the convertibility risk usually hovers around zero, it tends to increase during international crises. Note how both series jump to extremely high levels during the crisis of 2002.

One may rightfully argue that, compared to 2002, the current crisis is much milder, especially after the respite given by last week FED decision not to taper yet, and Brazil is much less vulnerable. Nevertheless, one specific event during the 2008 crisis highlights the importance of eliminating another component of uncertainty.

The importance of being willing to use reserves

In October 2008, when Brazil had already achieved a much more solid economic footing, including investment grade status, while announcing the program to confront the international credit crunch after Lehman’s demise, the Brazilian Finance Minister conveyed the message that President Lula had determined that international reserves should not be used. According to Valor Econômico (10 October 2008): “On Monday [6 October], while announcing with Meirelles [Henrique Meirelles, then-governor of the BCB] measures in order to deal with the crisis, finance minister Guido Mantega revealed that president Lula had commanded the economic team to spare the reserves – the country’s major insurance against a liquidity crisis. The market understood his statement as a presidential veto on using the reserves. On the following day, Tuesday, the dollar rose sharply and created an atmosphere of panic in the market, which has led Meirelles to ask for Lula’s authorisation to spend the reserves.”

To sum up, under normal circumstances, the BCB can intervene in the forex markets through US dollar swaps sales, without making use of its international reserves. What the BCB must never do is to give the impression that it will be reluctant to use foreign reserves, if need be. To credibly commit to provide the reserves, the BCB has to show both willingness to use the reserves as well as ability to do so. On the short run, with more than US$ 370 bi of international reserves (more than 16% of GDP), this does not seem to pose a problem. Nevertheless, the ability to provide the reserves will depend upon the return of a consistent macroeconomic policy stance, together with structural reforms, which are fundamental to assure Brazil medium and long-term external and domestic balance.

References

Chamon, M and M Garcia (2013), “Capital Controls in Brazil: Effective?”, PUC-Rio Discussion Paper No. 589.

Garcia, M (2013), “Brazil: Did Inward Capital Controls Work?”, VoxEU.org, 1 March.

Garcia, M (2013), “Can sterilized FX purchases under inflation targeting be expansionary?”, PUC-Rio Discussion Paper No. 606.

Lall, S (1997), “Speculative Attacks, Forward Market Intervention and the Classic Bear Squeeze”, IMF WP/97/164, December.

Lopes, F (2003), “Notes on the Brazilian Crisis of 1997-99”, Brazilian Journal of Political Economy, Vol. 23, n 3(91), July-September.

Topics: Exchange rates, International finance
Tags: Brazil, capital flows, Central Banks, derivatives, exchange rates

Associate Professor at PUC-Rio, Brazil; Visiting Scholar - Sloan School - MIT and NBER

Vox Talks