1999 Fiscal Conference - Sustainable Public Sector Finance in Latin America - Comments: de Oliveira Barbosa - Capital Markets and Deficit Finance in Brazil
Capital Markets and Deficit Finance in Brazil
FÁBIO DE OLIVEIRA BARBOSA
he purpose of this paper is to discuss the role of capital markets in deficit financing with specific reference to the Brazilian case. After this short introduction, the paper is divided into two parts. The first part summarizes fiscal policy in Brazil over the last few years, focusing on the implementation of a vast fiscal agenda despite some short-term setbacks. The second part describes recent developments in budget financing, particularly in strategies for public debt management for both domestic and external capital markets. The final section concludes by providing some perspective on future deficit financing and fiscal policy.
The Fiscal Agenda: Main Challenges and Results
Currently Brazilian fiscal policy is undergoing a dramatic change. In 1999, the primary result targeted was a surplus of 3.1 percent of gross domestic product (GDP). Of this amount, the central government was slated to accrue 2.3 percent of GDP, 0.4 percent should come from state and local governments, and state-owned enterprises should add another 0.4 percent.
The results, presented in Chart 1, confirm the Brazilian government’s commitment to achieving the required fiscal targets. The targets have been met for four consecutive quarters beginning in December 1998, and recent numbers indicate that the year-end target will be reached as well.
Figures for the central government, comprising the National Treasury, the Social Security System, and Central Bank operational accounts, showed a primary surplus of 2.66 percent of GDP in the period from January through October (see Chart 2). Last year, the primary surplus was 0.65 percent of GDP during the same period. The increase, equivalent to 2.1 percent of GDP, was largely achieved through expenditure reduction and enhanced tax revenue collection. It is also important to stress that this result has been achieved despite modest growth in GDP in 1999 and an increase in the social security deficit, which jumped from 0.6 to 0.8 percent of GDP in the same period.
On the expenditure side, despite some rigid rules, budget execution has been strictly in line with targets. Mandatory expenditures, such as transfers to states and local governments and social security benefits, grew by 15.0 and 9.7 percent, respectively, in nominal terms in 1999. Discretionary expenditures, the only remaining area of spending that could be cut, were reduced by 10 percent, demonstrating the major fiscal effort under way.
State-owned enterprises (SOE) moved from a deficit of 0.20 percent of GDP in 1998 to a surplus of 0.64 percent in the first ten months of 1999. The same trend was detected in the result for the state and local governments (SLG), which showed a 0.37 percent surplus through October 1999. This compares very favorably with the deficit of 0.27 percent of GDP during the same period in 1998.
In sum, looking at the short-term flows, it is clear that the government’s fiscal stance improved in 1999 compared with the previous year, when the primary balance was not consistent with the macroeconomic framework and the external scenario. However, analysts should not underestimate the efforts that the country has made over the last four or five years with respect to structural reform. In fact, Brazil has implemented a vast and impressive fiscal agenda despite, as mentioned before, precarious short-term fiscal flows.
The first item of this agenda is privatization. Brazil has been implementing one of the largest privatization programs ever. Since 1991, it has rendered US$87 billion in proceeds and debt transfers, an amount that would currently be equivalent to 16 percent of GDP.
Privatization was crucial from the fiscal standpoint due to two of its major effects: the permanent reduction of interest charges as a result of debt redemption, and the elimination of potential equity investments that could be required from the Treasury (as a shareholder) if companies remained in government hands. For instance, in the case of the steel and power sectors, the Treasury’s capital support for state-owned enterprises in 1986 and 1987 was equivalent to 2.5 percent of GDP.
Privatization not only has reduced outstanding debt but also has generated some positive “collateral effects” for the country, such as the reinforcement of foreign direct investment flows (FDI). In 1994, FDI in Brazil was US$2.6 billion. In 1998, it was US$26.1 billion, and in 1999, despite the floating of the exchange rate, FDI had reached US$24.8 billion in October and was expected to be around US$27 billion by year-end. According to the Central Bank of Brazil, FDI flows directly related to the privatization program totaled US$22.8 billion over the last four years (see Chart 3).
Privatization has also brought new players to the stock market, which is a very important aspect of the overall process of financing future economic growth. Finally, the Brazilian economy is already benefiting from the efficiency and productivity gains brought by privatization of former SOEs, which were usually overstaffed and inefficient.
The second item on the structural agenda was administrative reform. This program was designed to provide greater flexibility in human resource management, allowing for the required adjustments through the elimination of job stability and the single legal regime for civil servants and through the establishment of more rational criteria to adjust civil service salaries. A major change has already taken place even though additional legislation is required (for example, draft bills on the parameters for public employment and the dismissal of civil servants, and a constitutional amendment setting salary ceilings for the three branches of government). Nevertheless, a reduction of the potential long-term deficit is projected from enhanced control of personnel expenditure—one of the largest budget components.
The third major area was reform of the social security system. The milestone here was the change from a concept of time of service—established some sixty years ago, mirroring the Italian “Carta del Lavoro”—to a concept of time of contribution and minimum age as parameters for retirement. Also, the partial benefit granted for early retirement was removed. Those changes are expected to smooth and reduce the projected deficits for the social security system, until the results reach sustainable levels.
The second stage of social security reform is under way and involves several important aspects. These include reworking the regulatory frameworks between public sector entities and pension funds for public employees, instituting civil servant contributions (including those of retired civil servants) to public pension funds, and, finally, establishing more adequate parameters for private workers’ pension funds.
The fourth item on the reform agenda, which is not usually treated as such although it could be said that it has the same relevance in terms of structural change and long-run fiscal effects, is refinancing of state government debt. A major program has been implemented in this area since 1996. So far, the Treasury has signed contracts with twenty-five of the twenty-seven states, and the two remaining states (Tocantins, which is a new state, and Amapá, in the northern part of Brazil) are not crucial to determining the core fiscal result.
The total amount refinanced was around R$120 billion (equivalent to US$70 billion). Under the agreements, the states are committed to generating a primary surplus equivalent to 13 percent of net current revenues, and ceilings have been set for their debt-to-net-revenue ratio. The agreements also limit the states’ access to international borrowing from multilateral institutions and through bond issuance.
Several states initiated privatization programs, which included the sale of electricity, gas, water, and sewage companies. Since 1996, states took in US$31.6 billion from privatization (including debt transferred to the new shareholders), out of which São Paulo State’s share corresponded to approximately US$13.3 billion.
These contracts included several targets related to state banks, which were closed, transformed into development agencies, or privatized. Five major financial institutions were privatized (BANERJ, CREDIREAL, BEMGE, BANDEPE, and BANEB). BANESPA—one of the largest Brazilian commercial banks and previously owned by the government of the state of São Paulo—is going to be privatized in the first half of 2000. Several other smaller state banks were closed.
This effort has had two effects. First, together with the government’s program to strengthen and restructure the financial sector (PROER), privatization of state-owned banks is helping to promote a healthier financial system by eliminating institutions that were usually inefficient and undercapitalized from the economic arena. Second, it closes the important “window of financing” that states have used in the past, when their banks performed the role of “lender of last resort” to finance budget deficits.
A similar phenomenon occurred when some state-owned power companies used their cash flows to finance budgetary expenditures unrelated to their core activities. On some occasions companies would buy energy from federal power generation companies without paying for it, thereby transforming a state deficit into a federally financed deficit. It was a very comfortable and efficient source of funding because it is hard to imagine the federal government shutting off the power to a major city. In addition to causing fiscal problems, this procedure reduced the market value of federal power generation companies. As previously noted, the states’ privatization programs are addressing those concerns.
All things considered, one can easily attribute at least part of the swing in state and local government finances of 0.64 percent of GDP to the effectiveness of the fiscal programs monitored by the Treasury. A lot remains to be done, but it is clear that the potential for SLG finances to damage the country’s fiscal policy has been minimized.
In sum, the fiscal agenda that has been implemented over the last few years is a very important one, involving major structural changes that address all the relevant components of a sound fiscal policy. It is noteworthy to stress that most of this agenda calls for constitutional amendments, which—after passing the Senate and House of Representatives’ Justice and Constitution Committees—require a 60 percent majority in two rounds of voting in each of the two houses to receive final approval. So far, twenty-two constitutional amendments have been approved, most of them in the last six years, clearly demonstrating Brazil’s commitment to a new fiscal policy.
It was also true, however, that by the end of 1998 the gradualism in fiscal policy was no longer acceptable, particularly after the failure of the 1997 fiscal package. A tougher fiscal policy needed to be put in place.
This very strong message was conveyed by domestic and international capital markets through the volatility of many variables. Increased spreads for emerging economies, reduced access to new lending, and shortened maturities all affected the overall budget structure and public debt management strategy. These effects are precisely the subject of the next part of this paper.
Budget Financing and Debt Management Strategies in Recent Years
The bulk of the federal budget deficit in Brazil is financed by domestic capital markets, mostly through bonded debt. After 1994, which marked the beginning of the Real Plan and the successful process of macroeconomic stabilization, it was clear that the federal government’s debt management strategy should be adapted to the new environment. By that time, public debt structure was almost fully indexed—most of it to inflation indexes, the exchange rate, or to floating short-term interest rates; nominal, fixed-rate instruments represented no more than 6 percent of total outstanding debt. At that point, the basic strategy was to modify the domestic debt structure by gradually increasing the share and average maturity of fixed-rate instruments.
This strategy was implemented cautiously and gradually, since the economy was just starting to consolidate after the stabilization process began and uncertainty prevailed regarding inflation and nominal interest rates. Also, there was very modest activity in forward markets for longer-dated instruments. In early 1995, only two-month fixed-rate instruments were auctioned, and it was two and a half years until the Treasury could sell two-year fixed-rate notes—in September 1997. These were the longest-term notes ever sold in the domestic market. At that point, the share of fixed-rate instruments in total federal domestic debt was about 64 percent, and the duration had increased significantly.
The second auction for the two-year note was scheduled for October 30, 1997. However, in that very week the Asian crisis took place, bringing instability to international capital markets and particularly affecting the emerging market economies. Brazil also faced turbulence in its domestic capital markets, with uncertainties about the sustainability of the Real Plan and the virtually fixed exchange rate regime. The Central Bank raised short-term interest rates sharply in order to cope with market pressures against the economic policy framework.
Due to the increased duration of the instruments that had occurred, the cost of public debt was not significantly affected by monetary policy tightening. The Treasury continued to issue fixed-rate securities but, instead of a two-year maturity, the offer structure was gradually shortened because of the market’s reluctance to buy longer-dated securities. Maturity was first shortened to one year, then six months, then three months, and finally to two months at the first successful fixed-rate instrument auction after the Asian Crisis in mid-December, 1997. The maturity-lengthening process restarted, and by March 1998 one-year fixed-rate public debt instruments were being auctioned.
The process was interrupted again in mid-1998 in the wake of three sorts of problems. The first was turbulence in international capital markets associated with Russia. Second, 1998 was an election year in Brazil and the opposition party candidate was running ahead of the president, increasing uncertainties about the economic outlook and thus sparking volatility in the domestic capital market. The Central Bank target for overnight rates (the SELIC rate) was also reduced—a move that was not received well by the market. Finally, the spread between Central Bank rates and effective market rates widened, directly affecting the Treasury’s ability to sell fixed-rate securities.
In this environment, a new reduction in the maturity profile structure of the offer was required in order to allow the Treasury to keep selling fixed-rate securities. However, it was appropriate for the Treasury to go into a critical position in terms of the “perceived” refinancing risks by further shortening the average maturity. There were already some signs of concern from international and domestic capital markets about public debt profile. So there was a trade-off between duration and the Treasury’s refinancing risk, and the decision was to minimize the latter by using floating-rate instruments. The strategy was successful in keeping the average yield at very comfortable levels, but it did so at the cost of a sharp reduction in domestic debt duration.
In 1999, the Treasury continued to issue floating-rate instruments (one-and two-year notes), and by April fixed-rate instruments represented no more than 4 percent of total domestic debt. The Treasury gradually resumed issuing shorter fixed-rate instruments while monitoring the total debt average maturity (then at about ten months) in order to avoid unnecessary noise in this area (see Chart 4 for September figures).
As for external debt, the Treasury’s borrowing program is not driven by budget financing demands. Since 1995, when Brazil returned to international capital markets, the Treasury has borrowed about US$18.8 billion in nineteen transactions.1 Out of this total, around US$5.3 billion were bond exchange transactions and did not involve new money. In net terms, the Treasury borrowed some $13.6 billion in cash over almost five years (averaging about US$2.7 billion per year).
In this context, international borrowing is not really relevant for budget financing, which is mostly done through the domestic capital market. Of course the country faced a very difficult situation in late 1998 and early 1999, and the support package arranged by the international community was extremely important to the economy’s recovery and for dealing with the balance-of-payments constraints. However, this was an atypical situation, and the Treasury’s strategy towards international capital markets follows a different approach.
In fact, the first objective is to establish benchmarks in strategic markets (dollar, euro, and yen) in order to build and maintain some reference of Brazilian sovereign risk for the relevant segments of the yield curve and to pave the way for other Brazilian borrowers, both public and private. Establishing such benchmarks is important because the domestic capital market in Brazil does not yet provide long-term financing at rates compatible with the expected rate of return on projects.
A second objective is to develop a solid investor base for Brazilian issues in order to attract new investments and to demonstrate the country’s profound transformation over the past five years. It could be said that the Treasury’s external borrowing program has played a catalyzing role in the success of the privatization process and the massive FDI flows with which it is associated.
The third objective is related to balance-of-payments requirements. Since 1995, two of the nineteen transactions have been denominated in yen, twelve in euro or other European currencies, and five in global bonds. The turbulence involving emerging economies in general, and Brazil in particular, deeply affected country access to international capital markets. The magnitude of change can been seen in the widening spreads for sovereign bonds. In June 1997, the Treasury issued a global, thirty-year bond with a 395 basis point spread; two years later, a global, ten-year bond was bearing an 850 basis point spread.
However, more important than the yield is the access to the market itself, which is central to financing the current account deficit. So, when borrowers faced several “closed doors,” with a very low roll-over rate for credit lines and bonds, the Treasury played a key role in providing the appetite for Brazilian risk in several transactions throughout 1999, mainly in the new and promising eurobond market. The goal was accomplished, and domestic borrowers, despite the cost increase, are gradually returning to the international capital markets.
Perspectives for Fiscal Policy, Deficit Financing, and Debt Management Strategy
Brazil’s fiscal policy target for 2000 is an overall public sector primary surplus of 3.25 percent of GDP. The federal government budget was sent to Congress with a primary surplus of 2.65 percent of GDP, and, for the first time, this target was established by a legal instrument (the Budget Guidelines Law).
The privatization pipeline is very strong. In October, an important legal obstacle to the privatization of the electrical generation facility, FURNAS, was removed. Two other electrical generation facilities, CHESF and ELETRONORTE, are also scheduled to be sold in 2000. The federal government is concluding negotiations with the state of São Paulo to speed up the sale of its state-owned bank, BANESPA. CESP-Paraná will most likely be divided into two parts and also privatized next year, together with other power distribution companies owned by the states of Paraiba, Pernambuco, Pará, Maranhão, and Paraná.
Regarding debt management, the objective is to increase the share and the maturity of fixed-rate securities. The core of the Treasury’s strategy is to lengthen the duration of the debt in order to increase the predictability of debt service, reduce budget risks, improve the effectiveness of monetary policy, and develop the domestic capital market. As of late 1999, this has not been an easy task, given that forward markets are liquid only for the next six months and that the possibility of a one-year note is not there yet. However, as in the past, this process will be carried out very cautiously.
The Treasury has not issued dollar-linked securities since January 1999, and there is no intention to do that in the short run. Although it could be appropriate in a virtually fixed exchange rate regime, it would not fit well in a floating exchange rate regime. The Treasury’s goal is to use the predictability of the exchange rate to issue long-term securities (three to five years), extend the average maturity of public debt, and provide a sort of domestic benchmark until fixed-rate instruments can again be established.2
The development of the domestic capital market is a major priority for the federal government. To achieve this, one important measure is to reduce the number of auctions. There are now four auctions every week, and it is clear that this heavy schedule reduces investor interest in the events and does not support the development of the secondary market. However, the new debt auction schedule will be implemented very carefully so as to preserve the Treasury’s refinancing risk at safe levels.
In the same context, the Treasury is planning to reduce the number of debt instruments offered to the market. Currently, securities can be linked to the reference rate set by the Central Bank (TR), the dollar, price indexes, or a fixed rate. It is likely that this diversity will lead to the pulverization of the secondary market, with negative effects on its liquidity as well as an increase in the issuer’s cost of financing in the primary market.
The Treasury and the Central Bank of Brazil intend to publish monthly auction schedules, hopefully three or four months ahead of time, in order to increase predictability, attract more investors, and improve liquidity in some specific segments. The implementation of the “reverse auction” not only will enhance liquidity for longer-dated debt instruments but also will help to adjust the public debt profile to the new environment. Some market incentives to public debt dealers will be established, with more strict performance criteria in order to transform them into effective marketmakers.
Finally, the Treasury is planning the implementation of a “domestic exchange” involving the so-called privatization currencies. When several SEOs were sold or closed down, the Treasury took over their liabilities and refinanced them through privatization currencies. These are long-term bonds (the average maturity today is around fifteen years) bearing below-market interest rates. These could be called the Brazilian domestic Bradys. The problem is that there are seventy-five different bonds on the market, most with very low liquidity, which reduces their potential to operate as benchmarks for long-term transactions.
The plan is to hold an auction in which the Treasury would offer new five-, ten-, and fifteen-year securities in exchange for the privatization currencies, on a strictly voluntary basis. Cash could be accepted, but there has still been no decision regarding floors or ceilings for this portion. However, this will not be the focus of the auction since the idea is to improve the liquidity of these instruments, reduce the number of different instruments traded in the market, and at the same time provide a sort of benchmark. The new security will be indexed to the same price index used with most privatization currencies, which means there will be no increase in the share of indexed domestic debt.
The new bond issue may prove to be an important instrument for funds, insurance companies, and others who would like to hedge long-term indexed liabilities. At the same time, it could help the Treasury to lengthen the average maturity of the outstanding domestic debt.
In sum, the Treasury’s debt management strategy is to increase duration, which will be done basically through the increased issuance of fixed-rate securities. At the same time, the federal government (the Treasury and the Central Bank of Brazil) will be working to develop domestic capital markets adequately, focusing on the establishment of an effective yield curve covering all relevant maturities.
Until that goal is accomplished, it is important to keep borrowing from international capital markets as the Treasury did in 1999, with three-, five-, and seven-year euro transactions, as well as with the 2004/2009 global bonds. A new yen transaction was planned for the beginning of 2000, once that market appears to be opening up for emerging market risk.
The Treasury will continue to be present in those markets, but not as a heavy borrower. In fact, the Treasury’s external debt amortization profile is comfortable and enjoys the availability of a deep and liquid domestic market that, although requiring further improvements, already provides necessary budget financing.