To: posjim who wrote (7685 ) 9/9/1998 5:23:00 PM From: Steve Fancy Respond to of 22640
Merrill Lynch Bulletin 2 of 2 8 September 1998 Francis Freisinger Manager, Latin American Economics Latin America Economics: Brazil Hikes Rates But Can It Deliver On the Fiscal Side? ú As we expected, accelerating capital outflows have forced the Brazilian Central Bank to raise interest rates. The overnight rate (Selic rate) will rise from 19% to 29.75%, with effect from September 8. The move will buy the government time by temporarily slowing down capital outflows. But without additional measures the increase is probably not enough to generate net capital inflows. ú Whether this time is measured in days or weeks depends on two factors: First, can the government send sufficiently strong and clear signals about fiscal measures given the political constraints of the elections on October 3, and second, the state of the global financial markets. ú Preliminary reports suggest that outflows on Friday September 4 were around $3bn. That would imply that international reserves currently stand in the $55-57bn range, compared to $70bn at the end of July. Reserves will continue to fall even without capital flight by domestic Brazilian investors. Reserves could easily fall below $50bn in October even in a good scenario. ú New fiscal measures should do more than offset the extra debt service costs of the rate hike-we estimate these at around 0.22% of GDP per month. They ideally would have to make progress towards reducing the deficit to below the level expected before the interest rate hike. Unfortunately, we find it difficult to see where such dramatic fiscal measures could come from in the short run. The risk of disappointment is thus real in our view. ú If there is disappointment, or if world capital markets deteriorate further, it is possible that capital outflows will accelerate once again, and at that point it will be difficult to avoid floating the exchange rate. ú Despite recent market speculation, we do not believe the government will adopt additional administrative controls on capital outflows. Apart from strong denials this weekend by Central Bank authorities, we believe that the costs of such measures would significantly outweigh any benefits. ú At best, if the fiscal package is adequate and the world markets are more favorable, interest rates should begin to fall after the election. However, we do not expect rates to fall back to their previous levels until well into 1999. ú Last week we revised our annual GDP growth forecast down to 1.0% for 1998 and 0.4% for 1999. These forecasts incorporate a monetary tightening and assume that the government can produce sufficient fiscal measures to stabilize expectations. They imply a contraction in GDP in Q4 1998 and Q1 1999 of around 0.4% quarter-on-quarter, seasonally adjusted. This is roughly twice the contraction that occurred following the last period of monetary tightening, when rates were raised to 43%. 1) The Central Bank has raised rates by suspending rediscount operations at the 19% TBC rate with effect from Tuesday October 8. It will force the SELIC overnight rate up to the 29.75% TBAN level as a minimum. But rates could easily rise further if the authorities desire. 2) The move demonstrates that official commitment to the current exchange rate policy remains strong, after several weeks of apparent denial about the seriousness of the global situation. But the interest rate hike on its own will not be sufficient to stabilize the situation. 3) The interest rate hike will tend to discourage speculation against the Real, by increasing the carrying cost of short positions. But a comparison with the interest rate hike of October 1997 suggests that the increase may be insufficient to attract inflows. Then, the government raised the TBC and the SELIC rate to 43%. At that time, the implied annualized currency depreciation in the non-deliverable forward (NDF) market was then trading in 18-25% range (compared to 7% suggested by the exchange rate mini-band). But in recent days, the implied depreciation has been much higher, in the 32-39% range. In other words, from the perspective of an offshore investor, an overnight interest rate of 30% will still be highly unattractive until the risk premium in the NDF market has come down dramatically. The temporary exemption on the 15% withholding tax granted this August to foreigners makes the current overnight rate comparable to just 34.5% last October. 4) We believe that interest rates are going to stay at current levels or higher at least until November. Note that it was only in February 1998, three months after the onset of the previous crisis, that large-scale capital inflows arrived. This time round, given the deteriorating world environment, the political uncertainties generated by the election and the inevitable delay in implementing a fiscal package, it seems likely that it will be at least as long before the government can count on capital inflows. Meantime, reserves will continue to fall. 5) The fiscal impact of higher interest rates will be much more rapid than last October because of the structure of domestic debt. We estimate that 70% of the outstanding R$310bn of domestic debt is floating linked to domestic interest rates. That implies an extra R$2bn monthly spending. To put that in perspective, the entire monthly federal payroll is only R$3.5bn. 6) We believe Brazil's fiscal deficit was not acceptable even before the interest rate hike. So ideally a fiscal package would have be big. It ought to go well beyond simply offsetting this interest rate expense. Moreover, it should focus on reducing expenditure, although further tax increases are also going to be necessary. 7) The problem is that spending remains extremely rigid downwards in Brazil, since key constitutional reforms are still not fully in place. The only obvious short-term measure we can see would be drastic cuts in the "other" spending line at the federal level, which has risen rapidly in recent years. This spending line could relatively easily be cut by R$3-4bn in 1998. 8) Unfortunately, such a spending reduction would be far from sufficient. Yet big reductions in payroll spending will only be possible with the completion of enabling legislation for the administrative reform, which has been delayed until after the election. In the longer term, the social security reform offers the other key to fiscal savings. 9) Moreover, Brazil has a credibility problem as a result of the failure of the fiscal package announced last November to produce any improvement in the primary balance. Indeed the primary balance deteriorated by nearly 1.5% of GDP after November. In order to make any future package credible, binding fiscal targets are going to be crucial. That will require congressional approval, as will any tax increase. In other words, we cannot expect key fiscal measures to be implemented until after the election. Our big concern is that they will disappoint the market.