Efficiency and productivity of Brazilian banks: A new approach based on two-stage network DEA.
| Autor | de Mello Valerio, Victor Eduardo |
-
Introduction
This paper studies the performance of the Brazilian banking industry, applying the data envelopment analysis (DEA) model to general multi-stage systems (GMSS) (Kao, 2014), with panel data from the largest commercial banks. Unlike other DEA models, GMSS permits simultaneous computation of the efficiency of a production system and its corresponding internal processes without prior identification of the weights of aggregated products and inputs, but rather by the conventional definition of a ratio between aggregated output and input. Furthermore, GMSS permits relaxation of the assumption that internal processes do not consume exogenous inputs or do not produce exogenous outputs. These properties are particularly useful in studying a complex production system, comprised of a network of internal processes, as financial institutions are classified.
The credit market is strategic for the economic growth of emerging countries, insofar as financial intermediation activity boosts the flow of scarce resources to the productive sector (Levine et al., 2000; McCaig and Stengos, 2005; Bohacek, 2007). This importance is attenuated in emerging countries, which generally do not have a developed capital market, giving the banks a worrying responsibility for increased liquidity (Tecles and Tabak, 2010). In contrast, the impact of intermediation activities on economic growth depends on the degree of financial sector development (Eggoh and Villieu, 2014). Because of this, research on the performance of the banking sector is essential (Berger and Humphrey, 1997; Paradi and Zhu, 2013; Liu et al., 2013b; Shyu etal., 2015).
Although a considerable amount of research applies DEA models to analyze the performance of the banking industry (Liu et al., 2013b; Guo et al., 2017; da Silva Fernandes et al., 2018; Li et al., 2018), few have focused on Brazil (Wanke and Barros, 2014). Brazil has the largest and most complex financial system in Latin America, and its banking industry has experienced an accelerated process of deep restructuring since the first half of the 1990s (Tecles and Tabak, 2010).
This restructuring originated with changes in the Brazilian regulatory framework. The purpose was to adjust the credit market to the new macroeconomic context of inflation stabilization, combined with a fiscal adjustment, and additionally to overcome the critical situation of the financial sector. The sector was characterized by highly leveraged institutions, and credit portfolios lacking strict supervision of risk (Barros and Wanke, 2014).
Therefore, regulatory reforms focused on encouraging privatization, accompanied by the sector opening to greater participation by foreign banks. What suffered were efforts to decrease the presence of public banks, hitherto focused on financing redistributive policy and fiscal policy. The whole process involved a large number of merger and acquisitions (M&A) (Wanke et al., 2017). The plan was that the banking system's recovery would favor development and minimize the country's systemic financial instability.
Almost twenty years since the start of the restructuring process, evidence indicates that the Brazilian banking industry is close to its consolidation. However, the intense use of M&A as a privatization and economic liberalization mechanism has resulted in an increase in the degree of banking concentration (Belaisch, 2003), characterized by an abrupt decrease in the number of institutions and the emergence of large banks and financial holding companies. Consequently, competition is decreasing (Yildirim and Philippatos, 2007).
Based on the foregoing, the motivation of this article can be described as follows. First, we evaluate at which efficiency levels the Brazilian credit market has been consolidating recently. Although similar studies exist, analysis of the Brazilian banking system, comprised of a financial intermediation process and a revenue process connected in a network, has not yet been explored. Second, using the Malmquist index, we carry out an intertemporal analysis of change in productivity and technology for each semester in a four-year period. This application extends the literature by combining the concepts of the Malmquist Index to network DEA-GMSS, measuring changes in the banking system and their internal processes in an integrated manner. Finally, we examine how the financial environment impacted the performance achieved by Brazilian banks through truncated Tobit models. This method is necessary to better understand the impact of the bank restructuring.
The remaining paper is organized as follows. Section 2 explains the conceptual framework of Brazil's banking industry. The purpose of Section 3 is to explain the proposed conceptual model used to analyze banking institutions. Section 4 intends to clarify the applied methodology. Section 5 details data analysis. Sections 6 and 7 present empirical results and discusses their policy implications, respectively. Section 8 concludes.
-
Overview of Brazilian banking industry
Brazil's credit market was largely made up of public banks until the early 1990s. It represented 50% of total bank assets and 55% of total credits granted. Most loans were characterized as long-term. Based on this, some authors argue that public banks were important in funding investment, and thus to the country's economic growth (Ness, 2000; Beck et al., 2005).
However, in that time, banks were characterized by high administrative expenses, personnel expenses, and operational expenses. They depended chiefly on inflationary transfers to survive in the market. Brazil experienced double-digit inflation levels starting in the late 1950s, reaching a maximum level of about 2,708.17% per year in 1993. Thus, the inflationary transfers to the Brazilian banking system were about 3.4% of gross domestic product (GDP) in the corresponding period (Franco, 2000; Nakane and Weintraub, 2005).
The critical situation of the banking sector was further intensified by macroeconomic changes introduced by the inflation stabilization plan, called the Real Plan. Inflation effectively converged at low levels. Thus, one of the main sources of public banks' income, and of the banking industry in general, was minimized. Inflationary transfers decreased to 1.8% of GDP in 1994, reaching 0.03% of GDP in 1995. Specifically in relation to banks, inflationary transfers followed the same movement, decreasing from an average of 35% of total revenues in 1993 to close to zero in 1995 (Franco, 2000; Nakane and Weintraub, 2005).
In addition, the Real Plan planned to maintain the basic interest rates at high levels. The purpose was to signal to the market the economic policy's commitment to stabilizing inflation. As a result, other real interest rates in the economy also increased, including Brazil's banking spreads Beck et al. (2005). Aiming to replace inflationary transfers, the banking industry greatly increased its financial intermediation activities. However, it did so without conducting a rigorous study of risk with borrowers, and without strict supervision by the Central Bank of Brazil (Ness, 2000; Nakane and Weintraub, 2005).
Accordingly, to contain the increase in financial intermediation and its corresponding impact on stimulating aggregate demand, the Central Bank increased the percentage of reserve requirements. In 1995, the rate of reserve requirement of new demand deposits increased from 40% to 100%, while the rate of reserve requirements for time deposits was set at 20%. As a result, many companies went bankrupt, either due to decreased demand, or from increased debt from the increased bank spreads. In addition, the banks' credit portfolio deteriorated significantly, with a simultaneous decrease in the financial sector's liquidity (Ness, 2000; Nakane and Weintraub, 2005).
Given the catastrophic situation in its banking industry, with high expenses and low revenues, rising defaults and decreased liquidity, Brazil's Central Bank implemented a broad and aggressive normative project to restructure the financial sector. Strictly speaking, this was divided in two phases (Tecles and Tabak, 2010; Staub et al., 2010). In the first phase, based on a 1987 law, the Temporary Special Administration Regime (RAET), the Central Bank assumed the responsibility to extra-judicially liquidate banks or to place them under special temporary management in the event of problematic liabilities. It simultaneously provided funds to insure customers' deposits (Barros and Wanke, 2014; Ness, 2000). In the second phase, two regulatory measures were drafted: the Program of Incentives for the Restructuring and Strengthening of the National Financial System (PROER) in 1995, and the Program of Incentives for the Restructuring of the State Public Financial System (PROES) in 1996.
After these acts, the public sector's share in the Brazilian credit market has been minimized. Some small public banks were liquidated. In addition, the tool of M&A was used widely. Tax incentives were created and credit lines subsidized for both domestic and foreign banks to acquire public banks (Barros and Wanke, 2014; Ness, 2000). The Central Bank also passed Resolution 2,303 in 1996, allowing banks to increase revenues by charging fees for services.
The current situation of the Brazilian banking industry, since the regulatory changes implemented by the Central Bank, points to a process of bank consolidation, since no significant changes are observed in the number of banks and/or bank branches. Moreover, the sector composition related to bank ownership has shown some stabilization (Faria Junior et al., 2006; Almeida and Jayme Jr., 2008).
In contrast, several studies show that, due to the mechanism of M&A, consolidation of the banking system occurred at the cost of high levels of bank concentration, with negative impacts on the competitive structure of the credit market (Barros and Wanke, 2014; Yildirim and Philippatos, 2007)...
Para continuar a ler
PEÇA SUA AVALIAÇÃOCOPYRIGHT GALE, Cengage Learning. All rights reserved.
