Brazilian Central Bank: Why Fraga chose inflation targeting

The Brazilian central bank is facing a huge challenge with its decision to make inflation-targeting the cornerstone of monetary policy. With the failure of the previous policy aimed at preserving the exchange rate - which was in place between March 1995 until the real's devaluation in January - it is arguable that the bank under new governor Armínio Fraga had little choice but to go down this road.

The only other serious alternative - targeting monetary aggregates - would be difficult to implement in Brazil because of the high instability of money demand. Yet inflation targeting will hardly be any easier to bring off with the central bank facing a paucity of data on transmission mechanisms. This is needed to help decide what amount of money will feed through to what level of inflation.

Fraga, talking to journalists in London recently, explained why inflation targeting was chosen. "In reality if you are thinking about monetary systems and exchange-rate systems it's really a two step decision. "One is the exchange-rate system you want and what is appropriate for a given country: fixed or floating?"

"If it's fixed then your monetary policy is given by that sole objective of keeping the fixed exchange rate. My experience as an academic and a practitioner is that if you are going to do that it has to be really fixed in order to have a system that is rigid and robust such as Hong Kong's or Argentina's. If not, and you are going to go floating, then you have to decide how you are going to run your monetary policy, how you are going to anchor expectations, what academics call the nominal anchor.

"Here the choice is between a monetary anchor which is quickly disappearing from the face of the earth [because] with financial innovation the monetary aggregates are very unstable; what I call the 'just do it approach' where you have no anchor and just do it; and then inflation targeting. Inflation targeting to me is the superior method because it's very focused while at the same time allowing some room for discretion because you need it in this very creative world we live in."

The problem with inflation targeting is that it requires a method of inflation forecasting that takes account of how monetary transmission works and the lags in the system. The central bank needs to know precisely the effects on aggregate demand and inflation of short-term interest rates, reserve requirements and credit restrictions.

Says analyst José Carlos de Faria of ING Barings in a recent report: "In Brazil because of the poor quality of the time series on economic activity and especially because of the many changes in the economic regime, it is very difficult to estimate the econometric relationships necessary to describe the transmission mechanisms. Many times in the past, for instance, the government resorted to credit restrictions to curb consumption: these included hikes in taxes on financial transactions and more mundane measures like limiting the number of instalments allowed for the financing of durable goods. Because variables that cannot be easily quantified move so often, it is difficult to isolate the effect of real interest rates on economic activity."

Another difficulty with inflation targeting is that it can set up a vicious circle in the economy. If, for example, a poor harvest led to a rise in prices, the central bank would tighten monetary policy provoking a contraction in demand that would reduce the level of output even further and cause a recession.

On balance, however, most analysts agree with Fraga that inflation targeting is the best solution. "The major drawback of [monetary aggregates] lies in the high instability of money demand in Brazil due to technological and financial innovations, changes in taxes and regulations, and all sorts of structural changes," says the ING Barings report. "This leads to a low level of correlation between money supply and inflation."

As well as domestic factors there are also external ones to consider such as capital flows and exchange rate volatility. The very low impact of the real devaluation on inflation after forecasts of 1999 inflation had ranged between 15% and 80% (the expected figure is now 8%) shows how poorly the relationship is understood.

Says Fraga: "The ultimate goal of monetary policy is to look after inflation. To the extent that any outside event affects that, indirectly, we will take that into account. If indeed capital flows dry out a bit one could argue that there might be pressure on the exchange rate.

"In fact this has just happened due to both international factors [the crisis in Argentina and fears of interest rate rises in the US] and to internal events in Brazil such as in the last two weeks the political events in the state of Pernambuco. All of these have an impact on the markets and on expectations and they certainly don't help. We will keep an eye on inflation and ultimately that is what is going to guide us but you have seen the exchange rate go from 1.65 or so to 1.75 now down to 1.73 today and these are the things we are going to be looking into going forward.

"We live in an integrated world and for the most part these external factors have been helpful to us recently. Every now and then they turn around and we have to take whatever we get. But the main goal in the end will be what the implications are for inflation and if they seem to hurt inflation expectations then we have to take that into account."

Up to the devaluation Brazil's economic regime was something of a halfway house, with a pegged exchange rate and some capital controls. With the currency now floating many analysts feel that logically the capital controls should go too.

"We still have quite a lot of capital controls in Brazil and we have been gradually removing them, " says Fraga. "Our overall objective in terms of our balance of payments is that we think it's natural for a country in our position to run a capital-account deficit but we feel the markets will keep an eye on us and it has to be well financed. So the goal of our policy is to prevent the financing of these deficits with short-term money. Obviously an economy needs a cushion of short-term money to function and we are no exception and we welcome that. But what we don't want is to get addicted to a lot of short-term borrowing.