An Exchange Rate Band for Brazil

The major purpose of this article is to argue that, in the present macroeconomic situation, Brazil could adopt an exchange rate band very well defined between 1.8 and 2.5 (real against dollar). This new exchange rate regime would be able to reconcile – or at least minimize – the economic policy dilemmas related to growth, inflation and the balance-of-payments.

Many analysts tend to say that, with an open capital account, that is, under free international capital mobility, the only way for the Brazilian Monetary (and Exchange) Authorities to avoid this permanent speculative attack “in favor” of the Brazilian currency would be to reduce drastically interest rates (or to impose a huge “Tobin tax”). After all, there is no doubt that international investors are “attacking” the Real mainly because of the two-digit rate of interest in Brazil.

In a previous article published a few days ago (May 22) here in this blog, we discussed the taxation alternative, that is, the so-called “Tobin tax”, designed – as Tobin himself used to say – “to put some sand in the wheels of the international monetary system”. Now, we are going to present and discuss an exchange rate band alternative.

The Real has already been appreciating for more than five years and it seems that the market believes that the expected devaluation of the real is zero, given the good situation of the external accounts. Interest parity theory would suggest then that interest rates for Brazilian securities denominated in reais should be basically equal to rates for Brazilian securities denominated in dollars or to interest rates for USA securities plus a country-risk premium. But it is well-known that Brazil’s Central Bank follows a strict inflation-target monetary regime, which requires at this moment very high interest rates, entirely designed to combat or avoid inflation, independently of their impact on the exchange rate.

In any fashion, after the amazing appreciation of the Brazilian currency (from 4.0 in 2002 to 1.6 in 2008) and also after the world financial crisis in the last few months – with a supposed comeback of risk aversion and high volatility – one would normally ask ourselves if the so-called “speculators” would not begin to rethink the Brazilian interest parity situation, by making an extremely simple comparison: after all, for short periods such as one month or even three months, the interest differential is less than 1% monthly or 3% quarterly – against potential devaluations of the currency that might theoretically achieve easily two digits in the same periods.

In other words: what happened recently with international financial markets would supposedly reinforce the “fear” or at least the perception that, after a long period of appreciation of more than 100%, which lasted more than 5 years, the exchange rate scenario for the real might revert in the remaining seven months of 2008.

This would mean that, even without any interest rate reduction in Brazil or in spite of a new interest rate increase (as it recently happened), the speculative attack “in favor” of the Real should be somewhat decelerated, simply by the influence of the new international financial situation, which at least calls the attention of international investors to the fact that the volatility of exchange rates is basically the same as the volatility of stock and commodity prices – these are financial assets that can move easily by 10% in one single month (or one single week), in clear contrast to interest rates, which normally stay in a very short range between 0.1% and 1.0% per month.

What is even more interesting in the case of exchange rates (or commodity prices) is the overshooting phenomenon studied by Rudiger Dornbusch in the seventies and by Jacob Frenkel later on.

The foreign investor needs – in order to make a profit – a local interest rate higher than the sum of the “lower” interest rate (from the country where funds are borrowed) and the actual devaluation of the local currency. Pure arbitrage tends to produce therefore an equilibrium relationship – called international interest rate parity – between interest differentials and exchange devaluation.

In practice, what happens is that, in order for the market to produce this clear and certain devaluation expectation, it is necessary that the level of the exchange rate appreciates too much, up to a point which goes much beyond the equilibrium position.

This is what is happening in Brazil. The Dornbusch overshooting. However – unless an exchange rate band is announced – the overshooting might go too far. In the same fashion that Alan Greenspan and Robert Shiller used to talk about “irrational exuberance” for the USA stock market, one can say that there is a certain “irrational complacency” from the part of international fund managers. They simply do the carry-trade and neglect the exchange risk. Maybe this happens because, at the end of the day, they get 20% of the profits but the losses (when and if they occur) belong 100% to the client investors.

The fact is that financial flows dominate the exchange rate market in the short run and tend to generate these abnormal phenomena called overshootings, producing financial asset bubbles. Notice that financial flows represent very minimal parcels of huge inventories of financial assets (trillions of dollars) managed by hedge funds and institutional investors.

In consequence of all that, the Brazilian currency keeps being attacked in the middle of 2008, in spite of a world credit crisis, just because interest rates are high in Brazil – and interest rates probably will go up even more in the next few weeks, due to the inflation target model.

An exchange rate band should be able to stop the speculative attack, with interest rates staying where they are in Brazil or even going up more. The lower limit should be slightly higher than the present level of the market, let us say 1.8, otherwise one would be giving a prize – rather than a small punishment – to recent speculation. And the higher limit – something like 2.5 – is necessary and easy to be defended with the existing level of international reserves, in order to minimize the inflationary consequences of actual devaluations within the band, by signaling a maximum percentage range for the band of less than 40%.

In other words: the Central Bank buys dollars at 1.8 and sells dollars at 2.5. Without an exchange rate band, one might see the free market exchange rate going down (up) to nominal levels similar to 1994-1998 (one to one or even less), leading to an additional dramatic appreciation of the real effective exchange rate. On the other hand, with the introduction of a band with a lower limit of 1.8, the reversal of expectations could and should be almost immediate.

Therefore, with a large exchange rate band, at the lowest level of this band the probability of a devaluation of the real reaches almost 40%, becoming much greater than any possible interest differential or any possible additional appreciation. This is, in our opinion, an interesting and adequate way to stop and revert the overshooting, by containing this contradictory “speculative attack” against strong currencies with low interest rates and in favor of weak currencies with high interest rates. (The other alternative, which we discussed in another article here on May 22, is the Tobin taxation on capital flows).

It is important to reinforce the coherence and consistency between the concept of a large exchange rate band and the inflation targeting monetary regime. In both cases, we are talking about rules and about minimizing the discretionary power and the “art” of central bankers, which take arbitrary numerical decisions about interest rates, but tend to fear any type of intervention related to numerical “decisions” about the exchange rate.

In other words, without the monetary regime of inflation targeting and without the exchange rate band, the Central Bank becomes too dependent on the charisma and the art of persons like Greenspan or Bernanke or – in the case of Brazil – of a group of brilliant economists who are directors of the Central Bank.

Here, with respect to the exchange rate, it is the same story. Free international capital mobility produces bubbles and overshootings. The way to avoid it – or minimize it – is to decide about two numbers (the limits of the band), something that Central Bankers should not fear to do because, after all, they already decide – in an arbitrary fashion – the number for the basic interest rate.

5 Responses to "An Exchange Rate Band for Brazil"

  1. Vitoria Saddi   May 29, 2008 at 8:52 am

    Professor,I don’t understand how one could have control over both instruments: exchange rate and interest rates? Given the impossible trinity we can’t control both, right?

  2. John Hugh   May 29, 2008 at 8:54 am

    I was under the impression that fixed exchange rates don’t work in practice. If you take Brazil during the 1990s or Argentina in the same period, we notice that the speculative forces tend to win the battle and the government is forced to devalue the currency. Why is your proposal different?

  3. Anna Martins Fontes   May 29, 2008 at 8:55 am

    Don’t you think that the Brazilian government is already imposing an informal exchange rate target? If you observe the reserve accumulation and the amount of sterilized intervention it seems that the Central Bank is trying to defend the Real without imposing a formal exchange rate target. What do you think?

  4. Nicolas Magud   May 29, 2008 at 11:16 am

    Antonio Carlos,Thanks for an interesting post. I have the following comments. On the one hand, both exchange rate bands and capital controls have been shown in the literature (and im practice) not to work. On the other hand, from a long-term perspective, one of the most important things for the central bank is to build credibility. The latter takes a lot of time. So far, the Brazilian Central Bank has done a very good job in this aspect, showing how tough it is in defending price stability over anything else. I personally value that a lot. Credibility and a tough anti-inflationary credentials will help boost long-term and sustainable growth. The market will know that the bank is concerned with it, and consequently is anti-inflationary policy will be positively affecting inflation expectations and will reduce time-inconsistent behavior. The more so when global inflation is increasing–you don’t want domestic inlfation to separate form the former from above. The exchange rate will accommodate to clear the market. And, a country that is growing consistently over time is expected to have a stronger currency anyways.This is my view, so I will very much appreciate your comments.Thanks,Nicolas

  5. Lemgruber   May 29, 2008 at 2:57 pm

    I will try to comment some of the comments made on May 29.VitoriaThe trinity continues impossible, but the overshooting is going too far. Like a bubble.John HughAn exchange rate band is entirely different from a fixed exchange rate, particularly a large band of 40%Anna Martins FontesThe degree of overvaluation and overshooting indicates that the Government does not have an informal exchange rate target.NicolasThere is no reason for an exchange rate bank to have an impact on the credibility of the Central Bank. On the contrary.