Contagion or isolation?

Given the anxiety in the markets over the recurrent problems facing major emerging market economies such as Argentina, we are trying to assess the impact of a possible stress scenario in important emerging market countries crisis situations on the Brazilian economy.

Therefore, the next step is to anticipate how the Brazilian economy is likely to react to a possible external shock and large doses of contagion. This is why we have decided that the best way to analyze the impact of a rupture of this magnitude would be to verify the behavior of economic indicators in three previous but similar shocks – the Asian crisis of 1997, the Russian crisis of 1998 and the devaluation of the real in 1999.

Obviously, this requires us to bear in mind the significant differences in Brazil’s macroeconomic fundamentals during the periods in question. It is from precisely this viewpoint that we argue that the contagion of a possible crisis in Argentina or any other large emerging market country should not be so traumatic for Brazil. Finally, it is important to recognize that the fiscal effort, both in short-term performance and in structural reform, and the conduct of sound monetary policy and probable support of international institutions will make a difference both in the magnitude of the contagion and its duration.

The support given by the new agreement with the International Monetary Fund, under which Brazil will be allowed to draw down $15 billion between September 2001 and December 2002 in anticipation of possible problems in Argentina, clearly demonstrates the intention of the IMF to try to isolate Brazil from external volatility in order to preserve the progress made so far.

The key variables for quantifying the magnitude of the crises were country risk indicators as measured by the price of the C-Bond, Brazil’s most liquid Brady bond, and the EMBI-Brazil. We analyzed each of the three periods using the C-Bond and the EMBI-Brazil. We analyzed the main difference in the effective spread of the two benchmarks – moving averages of one month and three months – and used this figure as the measure of the severity of the crisis. The results in the table on the following page show that using this criterion, the worst impact came during the Russian crisis, followed by the devaluation of the real and the Asian crisis. In the average of the three crises, the deterioration in the spread between the current level and the one prior to the event, was 466 basis points.

We also analyzed the duration of each crisis, taking the number of days it took for the spread to return to pre-crisis levels. In this case, the Asian crisis was the longest-lasting crisis, lasting 92 working days, but we should recognize that during the Russian crisis, the spread did not fall back fully due to the onset of the real crisis. Recovery on average took 73 working days, or three-and-a half months.

From the viewpoint of monetary policy, one can see that because of the need to protect its managed exchange rate regime, rising interest rates in Brazil were always out of proportion to the deterioration in country risk indicators. During the Asian crisis, the ratio between the increase in interest rates and the deterioration in country risk was almost nine times. This ratio was far lower in the other two crises, but on the other hand, the interest rates prevailing at the time were much higher than during the Asian crisis. In each of the three periods, interest rates rose about 4.2 times more than the deterioration in the country risk indicator, which means that in each moment there was a rise of about 20 percentage points in the Selic rate, Brazil’s benchmark interest rate.

As we have just mentioned, the analysis of similar events in the past is fundamental for us to try and forecast Brazil’s economic behavior in a hypothetical crisis triggered by Argentina, or some other large emerging market economy. But the acknowledgment of parametric changes in the Brazilian economy today is necessary for us to avoid errors of judgment. We can therefore highlight a series of improvements in the country’s economic fundamentals, which differentiate the present situation from previous periods and which therefore generate different responses in economic policy and even in the markets.

Exchange Rate Policy
The first major parametric change that we can underscore is exchange-rate policy. Under the managed rate regime adopted in the previous crises, the economic policy instruments available in Brazil were far more limited than today. Monetary policy had to fully absorb the external shocks, which made international crises endogenous. The reflection of this peculiarity could be felt in the reaction of interest rates, which as we have mentioned, were 4.2 times greater than the deterioration in country risk.

With a floating rate regime, most of the international volatility has been absorbed by the exchange rate, and so monetary policy has shifted to inflation targeting. This requires the central bank to raise or lower interest rates to reduce the secondary effects of exogenous shocks on other prices in the economy without imposing such a great cost in terms of economic growth as was the case under the managed exchange rate system.

As well as being able to absorb shocks using both monetary policy instruments and the floating exchange rate, Brazil’s public finances currently allow the government to use fiscal policy instruments to increase its primary surplus targets. It can also reduce the effects that rising interest rates and a falling currency have on public sector debt.

To emphasize this improvement, it is worth recalling what happened after the Asian crisis broke, when the Brazilian government drew up a fiscal adjustment program that consisted of 51 measures, involving an increase in revenues and spending cuts, but which in practice led to a heavier tax burden and higher spending, bringing about a minute change in the primary fiscal balance.

The perception that the government was not capable of implementing a fiscal adjustment that could compensate for the rise in interest rates dramatically intensified the judgment of insolvency risk in the public finances, aggravated by the increase of almost nine percentage points in the debt-to-GDP ratios in 1997-1998. The debt ratio rose from 34.5% of GDP in 1997 to 43.3% of GDP in 1998. From 1999, after the approval of the Fiscal Stability Plan (PEF), the Brazilian public sector succeeded in moving to a primary surplus from a primary deficit, improving its short-term fiscal performance. Furthermore, it altered the structure of the tax system by enacting the Fiscal Responsibility Law, implementing a civil service reform, introducing an actuarial factor when calculating INSS retirement benefits that are paid to private-sector contributors and concluded a restructuring of state and municipal debts. These measures improved the credibility of Brazil’s public finances and made it possible for fiscal policy to absorb the impact of an external shock.

Instruments in Selected Periods
The table shows the extent of the constraints on Brazilian economic policy during the Asian and Russian crises. The ability to use a greater range of the support of multilateral agencies leads us to believe it would be possible to deal with an Argentine crisis today at a far lower economic and social cost.

Recent data show that financial markets will be the main channel of contagion in the event of a crisis in a major emerging country, with the deterioration in country risk indicators. However, given the improvement in Brazil’s macroeconomic fundamentals that we have set out above, it is likely that the country risk indicator would not worsen to the same extent as it has in previous crises and that recovery will be faster.

We have therefore adopted two hypotheses for the likely performance of the country risk factor in the case of a crisis. Under the first hypothesis of strong contagion, country risk would grow to the same magnitude as in previous crises, that is, a rise of 466 basis points. The second, more optimistic, hypothesis assumes that foreign investors will be able to distinguish between Brazil and the country in crisis. We also assume that investors will differentiate between Brazil today and the situation it was in prior to its fiscal adjustment program. We believe that country risk will deteriorate by 233 basis points, or half the rate of the previous crisis.

In both cases, we consider that the duration of the stress will be shorter than the average of previous crises, which lasted three-and-a-half months. Under the first hypothesis, country risk indicators would return to normal within three months, remaining stable up to the end of the period under consideration. Under the more optimistic hypothesis, the country would return to its current levels in two-and-a-half months and would fall a month later to average level prior to the worsening of the external crisis.

There is an additional reason to expect a faster recovery from previous crises: the perception that the markets are simply waiting for a clearer definition of the Argentine economic scene before issuing buy and sell orders for Brazilian assets, recognizing the merits of local economic policy in achieving stability.

Because Brazil is operating a floating exchange rate combined with an inflation-targeting policy, we do not expect that monetary policy reaction will have to be as extreme as it was in previous crises. Our view is that monetary policy will not be used to force a return of exchange rates to levels prior to the crisis, which under the previous managed exchange rate regime would require a rise of up to 20 percentage points in the Selic rate. We simply expect a correction in the Selic rate in the same proportion as the deterioration in the country risk indicators, which according to our simulations, would require an increase in the Selic rate from 19% now to a level close to 21.3%-23.7% in the case of weak and strong contagion respectively.

The duration of such a mini-interest rate shock would be consistent with the hypothesis of a rapid decline in country risk. We believe that the Selic rate could return to the same pre-crisis levels three months after it first began to rise.

Given the decline in country risk indicators and domestic interest rates, and assuming a hypothesis of a further round of US interest rate cuts this year, we can also estimate a likely pattern for the exchange rate under scenarios of strong and weak contagion. The more pessimistic hypothesis would lead to an overshooting in the exchange rate that would drive the ³ down to a level close to R$2.80 per dollar. The currency would later return to a level of about R$2.50 to the dollar. With the more optimistic scenario, the initial devaluation would drive the real down to R$2.65 per dollar and then revalue to R$2.45 to the dollar by the end of the year. We began this exercise with a base exchange rate of R$2.50 per dollar.

The hypothesis of rapid recovery and confidence in Brazilian assets, with temporary impacts on the exchange rate and less pronounced rises in interest rates is crucial for the economy to avoid falling once again into a perverse debt dynamic, which occurred in 1997-1998. The government must at all costs prevent – and this is what the floating exchange rate is for – the economy from falling into recession, since this would make it harder to continue running a primary surplus because of the high elasticity between GDP growth and tax revenues. The view that fiscal policy would be insufficient to offset the declining exchange rate would quickly set off a vicious circle as country risk deteriorated, further impacting the exchange rate and driving interest rates up.

Our simulations indicate that a temporary increase in interest rates would not substantially affect the outlook for real GDP growth this year, which is admittedly not particularly favorable. In this case, according to our scenarios and incorporating a rise in the projected primary budget surplus to 3.6% of GDP this year, we believe the debt-to-GDP ratio would reach 53.7% to 54.8%, depending on which of the two contagion hypotheses we choose. This would add 4.4 to 5.5 percentage points to the debt-to-GDP ratio this year.

We do not expect the market to see this rise as being caused by a deterioration in fiscal policy, since the quality of the public debt has improved substantially this year. The average maturity of the federal debt has increased by 40% this year, to 22.06 months from 15.8 months. The average duration has improved by 53% by rising to 9.7 months from 6.3 months in the first half of this year. Furthermore, the shift in the debt-to-GDP ratio, the main objective of the fiscal adjustment, would have been caused by an external shock – a crisis in Argentina – and not a more lenient fiscal policy, and markets would be more forgiving in their analysis of Brazil’s public finances.

The same would apply to the probable non-compliance with this year’s inflation targets. The central bank has set itself a target of keeping inflation within a 6% rise in the IPCA price index this year and our projections indicate that in a scenario of strong contagion, inflation could reach 7.1% or 6.7% in a less pessimistic scenario.

However, no one can accuse the central bank of pursuing an expansionist monetary policy. The fact is that the deviation from the target can be explained mainly by factors beyond the bank’s control, such as a price increases in the highly regulated power industry, the drought at the beginning of the year and the depreciation of the real. There is no evidence of generalized price increases, which could drive the inflation rate up permanently. Monetary policy should, as we have noted, be adjusted to prevent price increases caused by exogenous shocks from contaminating the rest of the economy.

Our view is that the more optimistic scenario will be borne out by events, if indeed there is another emerging market crisis involving a major country such as Argentina, which we are simply taking as an example for this exercise. As we have noted, macroeconomic indicators would suffer a substantial decline in the short term but since we do not expect a prolonged period of contamination, the outlook for economic growth and fiscal sustainability will not come under threat and so increase considerably the chances of maintaining stability after such an event.

Octávio de Barros is chief economist at BBV Banco in São Paulo. Fábio Akira is the bank’s senior economist.