Brazilian financial model highlights potential shortcomings

Written by Armando Castelar.

business day south africa

THE idea of expanding "concessionary finance" in South Africa, by providing loans at below market rates to favoured industries and infrastructure projects, is part of a policy thrust through which the government aims to boost the investment rate and influence the direction of the economy.

The Industrial Development Corporation is primarily in charge of this. A major inspiration behind this policy initiative is the Brazilian Development Bank (BNDES), which is widely credited with promoting Brazilian economic success through concessionary finance policies. It is considered a key instrument in implementing the Brazilian federal government's industrial and infrastructure policies.

Up until the 1970s, the main focus of BNDES was financing public companies. In the 1980s, it began to save failing companies and became, according to a phrase coined at the time, "a hospital for loss-making companies". In line with a general change in Brazilian politics in the 1990s, BNDES was then put in charge of managing the privatisation process, in addition to carrying out its regular lending activities. However, BNDES became wary of lending money to Brazilian state-owned enterprises, as the money was not always paid back. When infrastructure-building was privatised, BNDES began to fund infrastructure development once again. In recent years, BNDES has become involved in financing a very wide range of activities, from shopping centres and trucks to exports and hydropower plants.

BNDES is the main provider of long-term financing in Brazil. Concessionary credit accounted for more than 30% of total Brazilian credit for nearly all of the period between 1997 and 2007, equivalent to 17% of Brazil's gross domestic product (GDP). In 2010, the value of BNDES loans was more than three times the amount provided globally by the World Bank. BNDES has seen the value of assets under its control quadruple from $83bn in 2003 to $318bn last year: an average annual rate of increase of almost 20%. Last year, infrastructure and industry received the biggest shares of BNDES's disbursements, at 39% and 34% respectively.

BNDES is partly funded by a special payroll tax. The bank has to pay interest on the amounts that are lent to it by the unemployment insurance fund, but it does not have to pay the capital amount back. Additional funding takes the form of loans from the Treasury, which are long-term and are offered at a lower cost than market rates.

A review of the literature on the role of public or state banks in development reveals that government banks are more common in poor countries with underdeveloped financial systems, interventionist governments and poorly protected property rights. Government banks tend, in general, to be less productive than private banks. The Brazilian case supports this conclusion, as privatisation has generally improved bank performance there. In Italy, it was found that the stronger a political party was in the area in which a borrower firm was located, the lower the interest rate charged by government banks, suggesting corrupt and politicised banking practices. In many developing countries, government banks lend more and perform worse in an election year.

Overall, the literature suggests that government banks generate less financial development, lower expansion of per-capita income and less productivity growth than private banks. These adverse effects are strongest in poor countries. This backdrop is critical to understanding the substantial limitations of the BNDES model.

First, although only 6% of the projects BNDES finances are with big companies, 71% of all disbursements go to large companies, with the 10 largest accounting for 36% of the total. Many of these large companies are established multinationals, such as Vale and Petrobras. Such companies could easily raise loans from Brazilian commercial banks and international institutions.

Second, there has recently been a large increase in the funds BNDES receives from the Treasury; last year, half of the money channelled to BNDES came from the national budget. The Treasury borrows at 11% but lends to BNDES at 6%. In effect, the bank receives an annual $10bn subsidy from the Brazilian government, which is then largely channelled to big, creditworthy companies.

Third, while BNDES's disbursement as a proportion of total Brazilian investment has increased dramatically, from 5% in 1995 to 25% in 2010, the overall investment rate has remained relatively static, at 15%-20% of GDP. In other words, the low-interest credit the bank provides is not stimulating new investment. Instead, established companies are drawing increasingly on the cheaper credit BNDES provides, instead of going to the more expensive private banks.

Fourth, concessionary finance also has the effect of raising Brazilian interest rates, as the state has to borrow in order to fund the subsidy BNDES provides. Further, the scale of concessionary financing means monetary policy has less power in Brazil. BNDES interest rates are not affected by central bank determination, so monetary policies seeking to shift the rate at which individuals and firms borrow affect only two-thirds of the credit market that is represented by the loans BNDES does not provide.

Perversely, credit subsidies are more likely to be accessed by bigger firms, meaning small businesses and poorer households have to borrow at higher rates.

Finally, there is little accountability with regard to which projects BNDES funds, even though it is a public bank. There is no public discussion on the Treasury allocations to BNDES and, apart from reviews of procurement procedures, there are no external assessments of the bank. A six- or seven-person board appointed by the president makes all the decisions about who should receive what credit. The bank hides behind secrecy laws, which should not be applicable to subsidised financing.

Originally, the role of the bank was to strengthen the underdeveloped Brazilian banking sector and to provide funding for the development of Brazilian infrastructure. This may have been an important factor in Brazilian development for a while but, as the economy and the banking sector developed, BNDES increasingly took on the role of a large bank financing large companies at cheap interest rates. This created fewer opportunities for private banks to develop and has entrenched a huge subsidy, which is channelled towards large corporations from the tax contributions of the working class. At the same time, critical issues in the development of the Brazilian economy, such as the rate of investment and the productivity of the economy, are largely unaffected by concessionary finance and are relatively ignored.

If either the Brazilian government, or any other government hoping to learn from the Brazilian example, wants to promote growth and development, it should work on improving the investment climate and the quality of the schooling and post-schooling systems.

Therefore, a country such as South Africa, which has a sturdy and deep financial sector and many cash-rich companies reluctant to increase their investments until they face a more certain future, needs to more critically evaluate the BNDES story if it wishes to justify the expansion of concessionary finance.

Published in Business Day, South Africa, on August 2, 2012