Inflation Targeting in Brazil: Lessons from the First Year

Inflation Targeting in Brazil: Lessons from the First Year

Preliminary

Inflation targeting in Brazil: Lessons from the first year

Claes Berg

Chief Economist

Sveriges Riksbank

2000-06-24

I would like to thank you for inviting me to this conference evaluating the first year of inflation targeting in Brazil. As inflation targeting is becoming increasingly popular in emerging market economies, it is important to take stock of the experiences so far in order to come up with concrete improvements and useful policy advice for the future. Brazil has introduced a framework for inflation targeting, similar to that in UK, Sweden and other industrial countries and thus has the potential of providing a suitable model for other developing countries, not least with regard to quantitative modelling and transparency.

As I understand, the decision to eventually implement inflation targeting in Brazil was brought about by a confidence crisis related to the global financial crisis 1997-1999. The confidence crisis generated capital flight from emerging markets. In Brazil, fiscal and balance of payments weaknesses were already apparent at the time of Russia´s debt default in August 1998: the consolidated fiscal position showed a primary deficit (i.e. before interest payments), the bulk of government domestic debt consisted of short-term financing and the current account deficit was approaching 5 percent of GDP. Brazil had to let its currency – the real – float on January 15, 1999. In February, the real plummeted to 2,15 to the dollar, from 1,20 at the beginning of the year (see Figure 1). In the new, highly unstable, environment, the authorities had to find a new nominal anchor for monetary policy, which involved a stronger and more transparent commitment. In this context, inflation targeting emerged as an appropriate regime to achieve price stability.

On March 8, the Brazilian government and the IMF jointly announced a new economic programme for 1999. The programme aimed at minimising the passthrough of the devaluation on inflation, while allowing for some nominal appreciation of the exchange rate. As a first move, Banco Central do Brasil raised interest rates from 39 percent – the level they had reached before the float – to 45 percent. In order to further calm markets and reduce inflation expectations, an interest-rate-bias concept was introduced, which meant that the Bank could lower rates between meetings without calling for a new vote by COPOM (the Monetary Policy Committee).

All in all, Brazilian monetary policy performs well in the area of transparency and accountability. The main tools in this regard are the official inflation targets and the quarterly Inflation Report. Given that the January devaluation represented a once-and-for-all change in the price level, with no further upward pressure, the authorities decided to opt for successively declining inflation targets: 8 percent in 1999, 6 percent in 2000 and 4 percent in 2001, all with a tolerance range of +/- 2 percentage points. So far, the inflation target has been met, and actual inflation figures and market expectations are converging. The Broad Consumer Price Index, IPCA, was chosen for the purpose of gauging inflation targets (see Figure 2). The index covers a sample of families with personal income between 1 and 40 minimum wages and has a broad geographical basis.

The first Inflation Report was published in June 1999. The Reports are both detailed and technically advanced, including fan charts of inflation and GDP growth which reflect the varying degrees of uncertainty surrounding the central tendency measures of forecasts over the medium term. The analysis assumes a constant interest rate and the time horizon spans from six months to two years. The central bank bases its forecasts on a range of indicators including surveys of market expectations, short-term forecasting models and structural econometric models of the transmission mechanisms of monetary policy. Naturally, assessments of prospects for relevant economic variables such as the exchange rate, the fiscal policy stance and aggregate demand and supply are also included in the analysis.

Moreover, in order to promote transparency, edited notes of COPOM meetings are released and published on the web with a one-week lag (until a few months ago, there was a two-week lag). These minutes, which are about five pages long, contain information on the Committee’s discussion and assessments of aggregate demand and supply conditions, the international environment, prices, the money market, open market operations and, most importantly, inflation prospects and monetary policy guidelines. Importantly, the forward-looking bias (downward, upward or neutral) is always mentioned in COPOM minutes, which increases predictability and openness with regard to policy decisions.

During its first year of inflation targeting, the central bank has successively lowered interest rates from 45 percent in early March 1999 to 19,5 per cent in July 1999 and 18,5 per cent today (cf. Figure 3). Thus, interest rate policy has been prudently cautious over the past year, making sure that expected inflation is well within the target band for end-2000 before further monetary easing is pursued. Indeed, with a floating currency regime and less reliance on short-term capital flows, rates are now not only lower, but also more stable than before. Along with the improving trend of the fiscal and external accounts, the successful interest rate policy has contributed to the strengthening of the exchange rate.

Despite a nearly 50 percent nominal depreciation of the real, the occurrence of supply chocks such as drought as well as substantial adjustments in domestic oil product prices and other administered prices, inflation for 1999 was within the Bank’s tolerance range of 6-10 percentage points. During the autumn, inflation picked up somewhat, mainly due to the impact of higher oil prices, tax increases and a weaker currency. However, during the last three months, inflation has been on a downward trend - supported by the appreciation of the real and sluggish domestic demand – and reached 6,5 percent (y/y) in May. Thus, the Bank’s target of 6 percent (+/- 2 percentage points) for 2000 looks safely within reach.

Similarly, inflation expectations have come down as the performance and credibility of inflation targeting has gained momentum. In May 2000, they reached approximately 6,1 percent for 2000 and 4,5 percent for 2001. Thus, market expectations have gradually converged to the official targets and private actors are increasingly adapting to a flexible exchange rate system. In fact, market forecasts on the day of COPOM meetings are highly correlated with actual rate decisions as a result of the flexible arrangement with forward-looking biases. In this way, monetary policy is predictable and paves the way for an increasingly stable political and economic environment.

In spite of inflation targeting, exchange rate developments still play a vital role for the Brazilian economy, which remains quite vulnerable to the external environment due to big shares of assets and liabilities denominated in dollars and the dependency on capital inflows[1]. Indeed, inflation performance is still to a large extent influenced by developments in the currency markets, which has been very apparent over the past year (cf. Figure 1). Moreover, the central bank has occasionally, as provided for in the IMF agreement, been intervening in foreign exchange markets in order to boost the value of the real.

All in all, the Brazilian economy has performed remarkably well over the past year.

Despite pessimistic growth forecasts in the beginning of the 1999, ranging between –3 and –6 percent, GDP eventually turned out positive at 0,8 percent. This was due to the very strong fourth quarter GDP figure of 3,8 percent, driven by investment, industrial production and net exports (see Figure 4). Growth was somewhat slower in the beginning of 2000, reflecting still sluggish domestic demand after the severe fiscal tightening during the previous year, but is nevertheless expected to reach a healthy 3,5-4,0 percent this year.

In order to come to terms with deep-rooted fiscal weaknesses, a comprehensive three-year programme was worked out with the IMF in September 1998. The objective of the programme was to achieve high and rising primary surpluses in order to stabilise the debt-to-GDP ratio and put it on a declining path. Progress has been quite remarkable: ever since the announcement of the targets, Brazil has fulfilled them by a comfortable margin. The primary surplus for the central government is forecasted at 2,6 percent for 2000. Moreover, the current account deficit is projected to decline to about 3,5 percent of GDP (US$23 billion), which will be more than fully financed by net FDI. Another encouraging development was Brazil’s decision in early April to pay back the IMF emergency assistance, which signals the authorities’ confidence about the prospects for capital flows and the broad external accounts.

In conclusion, the implementation of inflation targeting in Brazil is off to a very promising start. However, I would nevertheless like to take this opportunity to deliver a few words of caution. With such a short period of success, credibility is still fragile. The perhaps most difficult issues lie ahead and are related to other areas than monetary policy: further adjustments in the fiscal deficit, a comprehensive tax reform and reforms in the pension, health and education systems. However, as the new monetary policy regime provides for continuous adjustments of the interest rate in light of inflation prospects, imbalances in the economy can now be prevented in a much better way than before. With its technically sophisticated framework for inflation targeting, promising track record of transparency and crucial contribution to Brazil’s remarkable economic performance during the past year, the central bank sets a good example for other emerging markets to learn from and follow.

Before ending I will pose some questions. Is it true that the market therefore focuses primarily on the minutes, rather than on the Inflation Report, when analysing the future direction of policy. It would be interesting if someone in the COPOM would like to comment on this. I would also like to know how you arrive at a certain inflation forecast in the inflation report: what are the respective roles of the staff and the members of the COPOM in writing the report? Which were the main considerations when the COPOM decided to cut interest rates? Was actual inflation figures more important than the forecast for future inflation? Was a stable real a prerequisite for rate cuts?





1

[1] However, the dependency on capital inflows is much less pronounced with the new exchange rate regime.