THE EFFECTS OF CREDIT UNIONS ON BANK RATES IN THE BRAZILIAN MARKET. Keywords: credit unions, interest rate, competitiveness, market share

THE EFFECTS OF CREDIT UNIONS ON BANK RATES IN THE BRAZILIAN MARKET Emmanuel Sousa de Abreu1 Abstract This article examines the relationship between cr...
Author: Ross Sherman
8 downloads 3 Views 668KB Size
THE EFFECTS OF CREDIT UNIONS ON BANK RATES IN THE BRAZILIAN MARKET Emmanuel Sousa de Abreu1 Abstract This article examines the relationship between credit unions' market share and the interest rates charged by Brazilian banks. The relationship was evaluated from two possible assumptions: that the largest credit unions' market share can influence interest rates, or that credit unions would be motivated by environments with higher interest rates due to the lower competitiveness. To measure the level of market share, it was used, as a proxy, the relationship between portfolio of credit unions and banks, considering more than one thousand (1,000) municipalities in which credit unions had values in the types of transactions analyzed. It was evaluated two types of operations, payroll deductible loans and personal no deductibles loans, trying to assess whether or not the difference in collaterals give more or less competitiveness. In order to investigate the relationship between these variables, it was used a test of nonparametric correlation, the Spearman's Correlation Coefficient. The results showed that, evaluating the Brazilian market, the interest rate on payroll deductible loans has positive relationship with the Market Share of credit unions, and also that, considering only the state of Rio Grande do Sul, there is a negative correlation between the credit unions' market share and the bank's interest rates on personal no deductibles loans. It was concluded that the influence of credit unions in the Brazilian financial system depends on the characteristics of the market being analyzed. Thus, in environments with low participation, the credit unions are not able to impact the system, and their participation increases more easily in markets where there are high profits margins, however, in markets with credit union expressive participation, they are able to impact the market prices in some financial operations. Keywords: credit unions, interest rate, competitiveness, market share.

1.

Introduction

The concentration of the Brazilian banking industry in recent decades has raised concerns about its effects on competition within the National Financial System. This fact has been emphasized and has brought up the role of credit unions in the financial market and therefore in the development of the national economy. In this context, one of the major advantages brought up by the defenders of a higher participation of credit unions in the market is that they increase the market competition and provide more efficiency to the system by reducing the prices of financial services offered to the consumers. Despite the extensive literature on market structure in the banking industry, the empirical studies in the Brazilian market about the increase on competitiveness provided by credit unions according to their market share are incipients, especially when taking into account its effects on reducing the price of the financial services, such as the loans interest rates. In the international context, there are some empirical studies, such as the articles written by Tokle and Tokle (2000) and Feinberg (2001), that have found an influence of credit unions on the banking loans rates offered in the markets analyzed by them. These articles are opposed to the view that all credit markets have national coverage due to technological innovations which 1

The views expressed in this work are those of the author and do not necessarily reflect those of the Brazilian Central Bank or its members. The author is Operational Manager of Special Studies in the Central Bank of Brazil. He has a Master of Science in Accounting degree.

1

enable a balance between the local markets prices. However, Simons and Stavins (1998) present evidences suggesting that the financial markets are still predominantly composed of regional structures. In this way, it is reasonable to assume that some financial products have a more local nature than others. Thus, it is also important to ascertain whether or not credit products with different characteristics are related to a higher or lower participation of the credit unions in the financial market, for example, the existence of the collaterals in the loans products. Taking into account the international studies which have demonstrated that credit unions increase competition in the financial system, the idea that this sector could provide greater competition and, consequently, could reduce the interest rates charged was brought to the Brazilian academy and market. Thus, considering that the level of participation of credit unions in the market is a potential factor of efficiency in the financial system, we can make the following question: is there a correlation between the interest rates of the credit transactions and the credit union market share in the banking industry? The following alternative answers to this question were formulated: Ho: The greater the participation of credit unions in a given market, the lower the interest rates of banking credit operations due to the increase in the system efficiency caused by the increase in competition. Hi: The higher the interest rates of the credit transactions, the greater the market share of the credit unions in a particular market because higher rates would be an entry opportunity with high profit margins. Thus, the general objective of this research is to conclude whether or not the participation of credit unions in the market has a relationship with the banking interest rates. 2.

Theoretical Reference

International studies emphasize that the competition established by credit unions can impact the interest rates of the local banks credit operations (Tokle and Tokle, 2000; Feinberg, 2001a; Feinberg, 2001b). In the beginning of his studies, Feinberg (2001a), started with a traditional economic logic, and defined the following assumptions for the model: a) The increase in the market share of the two largest banks leads to an increase in the Lerner Index (LI-Lerner Index)2. b) An increase in the credit unions market share leads to a fall in LI. c) The increase in the price elasticity of demand is accompanied by a fall in LI. d) Reductions in (b) and (c) are bigger, in absolute values, when there is an increasing in the market share of the largest banks. 2

Lerner Index: Describes the market strength of a business.

. Where P is the market price and the

marginal cost CM. The index is inversely negative the formula of price elasticity of demand.

2

The empirical results found by Feinberg (2001a) suggest that rates of loans with collaterals (vehicles) offered by banks in small markets are significantly affected by the market share held by the credit unions. On the other hand, in transactions without warranties, the author found no significant relation to the fees charged by the banks. Later, in Feiberg (2001b), the author adds others evidences, such as: a) In regions where banking regulators or other factors make credit unions market insertion easier , banking customers tend to pay lower interest rates; and b) The influence of the credit unions on banking rates tend to be stronger the higher the market concentration. These findings in the international studies provided the basis for the formulation of the null hypothesis (H0). In Brazil, the study of Annibal and Koyama (2011a) approaches this line of research when it compares the interest rates of the personal loans charged by the credit unions to the rates charged by the banks in order to verify the existence of significant difference between them and if the participation of credit unions in similar operations has any effect on the interest rates charged by the banks. According to the authors, the results revealed that the fees charged by credit unions were significantly lower, but there was no evidence that credit union competition was enough to reduce the fees charged by banks in the same type of credit line. It is noteworthy that the study used the number of banking service posts per capita as a measure of Credit Unions’ Market Share, in a particular municipality. Campos and others (2011) also perform a comparison between the Brazilian banks and the credit unions, however the authors have focused on the analysis of the differences between the prices of the banking services charged by the two segments. For these authors, the credit unions have the following characteristics: a) greater flexibility and differentiated products; b) ability to compete with the banking market; c) in some banking services the credit unions practice higher prices than banks and in others banking services, lower prices. In addition to it, Silva (2011) provides more information about the Brazilian Credit Union System (BCUS). Below, we emphasize some of his conclusions that will be useful in the discussions of this article, namely: a) the weight of the BCUS in the credit supply and in the capillarity varies significantly depending on the credit modality and on the federative unit; b) the credit unions would be more focused on granting credit to the lower level of the population, by providing personal loans and microcredit; c) the BCUS also would collaborate to improve the macroeconomic efficiency of the Brazilian Financial System due to its concentration in the south region of the country and not in the southeast region, where there is a greater banking concentration.

3

The different average rates charged by traditional financial institutions, including credit unions according to the Brazilian regions are highlighted in Sebrae (2012). The study evaluated the different average rates charged by financial institutions in different regions of Brazil and between different products granted to Micro and Small Enterprises (MSEs). In Sebrae’s study, it appears that some regions have practiced better interest rates according to the product they have offered, and there is no a geographical region that has linearly showed better rates than others in all kinds of products. Below, the different rates by region for two different products are shown as an example. Figure 1 - Monthly average interest rate of discounted trade bills (%)

North

Northeast

Southeast

Brazil

Midwest

South

Source: Sebrae (2012)

Figure 2 - Monthly average interest rate on the overdraft banking

North

South

Brazil

Southeast

Midwest

Northeast

Source: Sebrae (2012)

Thus, it is noted that some regions, on average, have better positions depending on the products offered. This is an important factor and indicates that the Market Share of the cooperative sector should be analyzed according to participation in a particular product or service, and not according to the absolute participation, for example, participation in total deposits, total operations credit or according to the number of service points in a given region. Perhaps, that was the reason that made the study of Annibal and Koyama (2011a) has not found a relationship between the participation of cooperatives in the market and the market interest rate. Another aspect that needs to be taken into account in the analysis of the results shown in the item 5 of this work concerns the characteristics of the Brazilian banking system, especially regarding to the aspects of concentration and competitiveness. 4

In this way, Tavares (1985) reports that the national banking industry was under an explicit guideline for concentration from 1967/68, policy that was ratified in the 70's by the creation of the Commission for Companies Mergers and Acquisitions (Cofie) in 1971. The author demonstrates, with empirical data, that the measures have provided effective results and that the Brazilian banking sector increased its degree of concentration in the period from 1968 until 1981. Besides, Morais and Silva (2006) have evaluated the bank concentration in Brazil after the Real Plan3, and also identified the existence of concentration. The authors have justified this result on the fact that the volume of total deposits and total assets in the Brazilian financial market is concentrated in a small number of banks, which means that there is a high correlation between the Market Share and the amount of assets variables. In the same way, Paula Freitas (2010) has evaluated the impact of banking concentration in Brazil after the Real Plan and has observed the concentration of banking and financial activities in more developed regions of the country, what has caused a worsening of the distribution of credit in the poorest regions, which is consistent with other studies in the same line of research (Vasconcelos and others, 2004; ALMEIDA, 2007; Menezes and others, 2007). As for the current level of concentration of the Brazilian system, Dantas and others (2011) indicate, analyzing the period from 2000 to 2009, that: a) based on the parameters of the USA authorities (HHI), there was not found, in the whole sample period , evidence of concentration in the Brazilian banking market in relation to total assets, loans and banking deposits; b) the indicators exceed the limit to be considered as "moderate concentration" market in all parameters, considering only the end of sample period used in the study. Thus, based on the criteria of regulatory agencies, which tend to be more restricted and judicious, Dantas and others (2011) found a degree of "moderate concentration" at the end of the period. Another important issue for the analysis of the results brought in the item 5 of this work is related to the “credit leaking” which tends to occur in the Brazilian environment. This transfer of deposits from a peripheral region to the major centers is brought in the works of Vasconcelos and others (2004) and Sorgato and others (2010). They have shown that banks tend to allocate their resources, acquired at remote locations, in the more profitable environments. These evidences in the literature, banking concentration and "credit leaking", demonstrate that the Brazilian banks have the ability to move their resources to regions of their interest and not necessarily by the pressure done by the increase in the number of credit unions in the national markets.

3

The Real Plan ("Plano Real", in Portuguese) was a set of measures taken to stabilize the Brazilian economy in 1994. The Plano Real intended to stabilize the domestic currency in nominal terms after a string of failed plans to control inflation.

5

As a conclusion, the evidence brought by the international literature of negative correlation between the market share of credit unions and the interest rates can not be directly applied to the Brazilian environment. This fact should be taken into consideration during the formulation of the methodologies for assessing correlations. These aspects have established the basis for the formulation of the alternative hypothesis (Hi). 3.

Delimitation

The research to be developed has a very specific scope, given that it is not meant to characterize or explain the results to be found. Its goal is to collect results of the existing studies and verify if they also occur in the Brazilian environment. That is, we tried to verify whether what is proposed in the existing international literature and commonly theoretically established for the Brazilian environment can be empirically considered too. The goal is, by starting from a Popperian point of view (whose understanding believes that science can not prove anything, just "disprove"), to test if the participation of credit unions in the Brazilian market provides enough competition to imply a reduction in the market interest rates. If the statistical tests result in a positive correlation, the assumption that the largest share of credit unions in the markets necessarily induces to a reduction in the interest rate in the evaluated region will be denied. And in this case, there will be other possible explanations for the empirical results found. At the same time, this result could strengthen evidence that market characteristics determine the relationship of the Market Share of credit unions in bank interest rates. On the other hand, if we find out that a negative correlation exists, we can not necessarily affirm that this is due to greater competition, however it can be considered a possible answer to the evidence found. 4.

Method

The methodology proposed for this study is similar to the methodology employed in the studies of Feinberg (1999) and Feinberg (2001a). It differs, however, in the types of data used and in some particularities of the implementation and the analysis of the correlation tests, mainly due to the peculiarities of the data, the market characteristics and the sampling distribution observed in the Brazilian environment. In essence, however, the studies converge in evaluating the relationship between market share of credit unions and the rates charged on loans with and without collateral in local environments. 4.1. Data base The data used in the research came from the Credit Information System of the Central Bank of Brazil (SCR). This system is used in Brazil to record and to query information about the loans, sureties and guarantees and credit limits granted by financial institutions to individuals and companies in the country. This base was also used by Annibal and Koyama (2011a) to obtain the interest rates of their research. 6

This data collection system was instituted by the National Monetary Council (CMN) and it is maintained by the Central Bank of Brazil (BCB), which stores the information forwarded and also regulates the process of correcting and updating the database by financial institutions4. As established by Brazil (2008), in his art. 3, it is considered credit operations, for the purposes of information provided to SCR: I - loans and financing; II - advances; III - leasing operations; IV - co-obligations and collaterals; V – non-cancelable unconditionally and unilaterally credit commitments by the institutions mentioned in art. 4; VI - operations written off as losses and credits contracted with resource to release ; VII - other transactions involving credit risk, including those that have been negotiated with substantial retention of risks and benefits or control, in accordance with Resolution nº. 3,533, of January 31, 2008.

As will be more fully explained ahead, the data used in the research are aggregated by segment, by type of transaction and in certain prescribed period, making it impossible to identify the institutions involved or the individual operations. This procedure was essential, given the issue of confidentiality of personal banking data guaranteed by the Brazilian law. 4.2. Variables and Sample Data In order to measure the relationship between the interest rates and the market share of credit unions in a given region, two metrics were used. The first was represented by the municipal average rates of credit operations performed by multiple banks segment, while the second was represented by the ratio between the total value of financial operations of the cooperative sector and the total financial value of multiple banks (Market Share by type of operation). We have used the average interest rates of the following types of loans: personal loans and payroll deductible loans. The data was obtained on December 2012. The use of these two operations aimed to verify the existence of distincts relationships between the participation of cooperatives operations with low level of guarantees and those with a strong guarantee instrument, for example, the direct rebate on the payroll of the debtor. The two metrics that were used are described below. a) Credit Unions Market Share (Market Share-MS): is calculated by the ratio of active portfolio of the cooperative sector with the active portfolio of the commercial banks segment.

The active portfolio is calculated by totaling the values of the loans that are expired and the ones that are to expire.

4

More information is available on: http://www.bcb.gov.br/

7

∑ Two types of market participation were established in this study based on the type of the credit operation. First, the market share of the payroll deductible loans (MSc) and second, the market share of the personal loans (MSs). b) Annual Hired Rate Weighted by Active Portfolio (Txp): a relationship between the sum of the result of weighting the contracted rate of operation by the value of active portfolio of operations and the segment’s total active portfolio. ∑

Just as for the variable Market Share, two variables of average interest rates by multiple banks according to credit subtype were also established. One concerning the rate of payroll deductible loans (TXC) and another rate of personal loans (Txs). The variables above were obtained for two Brazilian regions and for two different types of operations, forming four data samples, as indicated below: - Sample 1: Brazilian Municipalities with balance in payroll deductible loans; - Sample 2: Brazilian Municipalities with balance in personal loans; - Sample 3: Municipalities of Rio Grande do Sul state with balance in payroll deductible loans; - Sample 4: Municipalities of Rio Grande do Sul state with a balance in personal loans. The state of Rio Grande do Sul was chosen because it is the Brazilian state that has the largest credit unions market share and therefore, allows the representation of a Brazilian market that comes closest to the international markets in terms of market share. 4.3. Descriptive Analysis and the Chosen Statistical Method The methodology used to evaluate the association between the variables was based on non-parametric correlation test, the most widely well-known method is the Spearman (ρ). This statistical variable sorts the data and then applies the Pearson (r) equation to those data, as described below. ∑

To calculate this coefficient, initially, we ordered one of the variables in an ascending ranking order and corresponding to each element will be obtained by the number of elements in S whose rank is higher than that is calculating the least number of elements whose rank is lower thereto. According to Bunchaft and Kellner (1999), the ordinal correlations cannot be interpreted in the same way as the Pearson correlations mainly due to the fact that the square of the index 8

cannot be interpreted as the ratio of the variance common to the two variables. The statistics can be defined according to the equation below.

Where: ρ = Spearman correlation coefficient; = Difference between the ordinances (between positions corresponding values of x and y); = Number of pairs of values (of sorts). 5.

Results

After we have justified and explained the choice of the statistical methodology, we pass now to evaluate the results generated by the tests. Below, the tables with the information about the performed tests are presented. The tables show the coefficient of correlation between the two variables (Ms and Tx). After, we show the significance value of the coefficient and finally, we present the sample size. First, we analyze the sample of the personal loans for all the Brazilian municipalities. Correlation – Personal Loans (Brazil) Spearman's rho (ρ)

Average Interest Rate (%) Market Share (%) Correlation Coefficient Sig. (2-tailed) N

,013 ,663 1045

It is observed that the value of two-tailed significance for ρ is well above 0.05, which indicates that one cannot say that there is a relationship between increased participation of credit unions in the Brazilian personal loans and the rates practiced in these operations. This result is similar to the result found by Feinberg (2001a). He found no significant relationship between the Market Share of credit unions and the interest rates of the credit operations without any type of collaterals. Importantly, the relationship was not significant even with a large sample of 1045 municipalities in which credit unions had balances on personal loans operations. On the other hand, the results obtained in the payroll deductible loans sample were different. The results found for Brazilian municipalities are showed in the table below. Correlation – Payroll Deductible Loans (Brazil)

Spearman's rho

Annual Interest Rate Average - Market Share Correlation Coefficient Sig. (2-tailed) N

,155(**) ,000 650

** Correlation is significant at the 0.01 level (2-tailed).

Thus, unlike the personal loans, the data of this sample indicated significant values. It is observed that the value of bilateral significance for ρ is well below 0.05, even lower than 0.01, which enables us to state that there is a significant relationship between the increased 9

participation of credit unions in Brazil with payroll deductible loans and the rates used by multiple banks in these transactions. It is important to note that the coefficient of correlation was positive (0.155), confirming that as the average interest rate practiced by multiple banks increases, there is a corresponding increase in the market share of credit unions in the city. This result is at odds with empirical results found in countries with a higher share of credit unions in the financial system, since in these countries the evidence indicates a negative relationship between measures due to the increase of competitiveness in the industry (TOKLE and TOKLE, 2000; FEINBERG , 2001a; FEINBERG, 2001b). However, one can consider this result consistent by observing the characteristics of the Brazilian financial system considered as a whole, mainly due to the low participation of the cooperative sector in the financial system (about 2% of total assets). In this sense, it is understood that the high rates charged by banks in these counties would allow the entry of credit unions in these operations and, therefore, the credit union highest market share. At the same time, low market share of this segment (SILVA, 2011), however, would not be able to force a reduction in the rates charged by banks, due to the banking high concentration and the national characteristics of Brazilian commercial banks. As will be seen ahead, the results are different when we evaluate the state with the largest market share of credit unions alone. In this way, we analyze below the correlation obtained in the Brazilian market which has the largest share of the credit unions, represented by the municipalities of the state of Rio Grande do Sul. As explained earlier, we had as a goal to verify whether in environments where cooperatives have high market share, there would be another form of correlation between the variables, notably the hypothesis of increasing competition and, therefore, reducing the interest rates charged by banks. Correlation – Payroll Deductible Loans Operations (Rio Grande do Sul) Average Interest Rate (%) Market Share (%) 1,000 Spearman's rho (ρ) Correlation Coefficient Sig. (2-tailed) . N 54

,265 ,053 54

Regarding the payroll deductible loans operations of Rio Grande do Sul, the significance value for ρ was found nearby, but was higher than 0.05, which indicates that one cannot say there is a relationship between increased participation of credit unions in Brazil with payroll deductible loans operations and the interest rates charged on these transactions. Correlation –Personal Loans (Rio Grande do Sul) Annual Interest Rate Average - Market Share Spearman's rho (ρ) Correlation Coefficient 1,000 Sig. (2-tailed) . N 101 * Correlation is significant at the 0.05 level (2-tailed).

-,196(*) ,049 101

10

In another way, differently from the payroll deductible loans transactions, the data for this sample indicated significant values. It is observed that the value of ρ for significance are presented below 0.05, which enables us to state that there is a significant correlation between increased participation of credit unions with personal loans and the rates used by the commercial banks. However, it turns out that the correlation coefficient was negative, confirming that, as the market share of credit unions in the municipality increases, there is a decrease in the average interest rate applied by multiple banks and vice versa. This result is consistent with the empirical results found in countries with a higher share of credit unions in the financial system, since in these countries the evidence indicates a negative relationship between measures due to the increase of competitiveness in the industry (Tokle and Tokle, 2000; Feinberg, 2001a; Feinberg, 2001b). The final results of the tests are summarized in the table below: Summary of Results

Market Brasil Brasil Rio Grande do Sul Rio Grande do Sul

Type of Credit Operation Payroll Deductible Loans Personal Loan Payroll deductible Loan Personal Loan

Result Sig. values Hi – Correlation + not significant not significant Sig. values Ho – Correlation -

Ho: The greater the participation of credit unions in a given market, the lower the interest rates of banking credit operations due to the increase in the system efficiency caused by the increase in the competition. Hi: The higher the interest rate of the credit transactions, the greater the market share of the credit unions in a particular market, because higher rates would be an entry opportunity with high profit margins. 6.

Conclusions

The results presented here suggest that a) the inclusion of credit unions, considering the Brazilian local markets as a whole, lacks the ability to reduce the rates charged by the commercial banks; b) credit unions tend to increase their participation in places where there are high interest rates because it means there is competitive advantage due to high profit margins. It was interpreted from the literature presented, that the characteristics of the Brazilian market justify the results obtained in this research, mainly due to the concentration of the Brazilian financial system and due to the process known as "credit leaking." These findings were observed from empirical data obtained in 1,045 Brazilian municipalities that demonstrate that there is not a negative correlation between the market share of credit unions and the average interest rate practiced in municipalities. Rather, there is a significantly positive correlation between interest rates of payroll deductible loans and the Market Share of credit unions in this type of operation. As for the personal loans, the results were not significant and, therefore, there were not conclusive results. 11

On the other hand, restricted to a smaller market, such as the Rio Grande do Sul state analysis, which has the largest market share of credit unions in Brazil, we have obtained a different result from the rest of the country. We have found a negative correlation between the banking interest rate and the Market Share of the credit unions with personal loans operations, which meets the international literature and reinforce that in environments with greater participation of cooperatives, an increasing in the market share of the segment in some types of operations may impact the banking interest rates. References ALMEIDA, D. B. D. C. Bancos e concentração do crédito no Brasil: 1995-2004. Cedeplar. Belo Horizonte/MG, UFMG: 89, 2007. ANNIBAL, C. A.; KOYAMA, S. M. Cooperativas de Crédito: taxas de juros praticadas e fatores de viabilidade. Trabalhos para Discussão. Brasília: Banco Central do Brasil, 2011a. Disponível em: . Acesso em 19 jun. 2013. ANNIBAL, C. A.; KOYAMA, S. M. Pesquisa Trimestral de Condições de Crédito no Brasil. Trabalhos para Discussão. Brasília: Banco Central do Brasil, 2011b. Disponível em: . Acesso em 19 jun. 2013. BRASIL. Resolução CMN nº 3.658, de 17 de dezembro de 2008. Altera e consolida a regulamentação relativa ao fornecimento, ao Banco Central do Brasil, de informações sobre operações de crédito. Leis, Brasília, DF, 2008. Disponível em: . Acesso em: 18 jan. 2013. BUNCHAFT, G.; KELLNER, S. R.O. Estatística sem mistérios. 2ª ed., v.2, Petrópolis: Vozes, 1999. CAMPOS, T. S. M.;SOUZA, M. C; CAMPOS, F. C., et al. O Cooperativismo: uma Análise das Taxas Cobradas nos Serviços Prioritários de Bancos Versus Sistema de Cooperativas de Crédito. VIII Simpósio de Excelência em Gestão e Tecnologia. Anais..., 2011. DANTAS, J. A; MEDEIROS, O. E; PAULO, E. Relação entre concentração e rentabilidade no setor bancário brasileiro. Revista Contabilidade e Finanças, v. 22, n. 55, p. 5-28. 2011. DRAPER, N.; SMITH, H. Applied regression analysis. 2ªed. New York: John Wiley & Sons, 1981. FEINBERG, R. The Effects of Credit Unions on Market Rates for Unsecured Consumer Loans. Filene Research Institute (University of Wisconsin-Madison) e Center of Credit Union Research. 1999. FEINBERG, R. The Competitive Role of Credit Unions in Small Local Financial Services Markets. The MIT Press. The Review of Economics and Statistics, v. 83, n. 83, pp. 560-563 2001a. FEINBERG, R. The Effects of Credit Unions on Bank Rates in Local Consumer Lending Markets. Filene Research Institute (University of Wisconsin-Madison) e Center of Credit Union Research.2001b. 12

FREITAS, A. P. G.; Paula, L. F. R. Concentração Regional do Crédito e Consolidação Bancária no Brasil: Uma Análise Pós-Real. Revista Economia. v. 11, n. 1. p. 97-123, 2010. HUMPHREY, D. B.; LAWRENCE B. P. "Banks' Responses to Deregulation: Profits, Technology, and Efficiency," Journal of Money, Credit, and Banking, p. 73-93, Fev, 1997. MENEZES, M.; CROCCO, M.; SANCHES, E.; AMADO, A. Sistema financeiro e desenvolvimento regional: Notas exploratórias. In Paula, L. F. & Oreiro, J. L., editors, Sistema Financeiro: Uma Análise do Setor Bancário Brasileiro. Elsevier-Campus, Rio de Janeiro, 2007. KWOKA, JOHN E ., Jr. The Effect of Market Share Distribution on Industry Performance, this REVIEW, pp. 101-109, Fev, 1979. RHOADES, STEPHEN, A. The Effect of Nonbank Thrift Institutions on Commercial Bank Profit Performance in Local Markets. Quarterly Review of Economics and Business, 1987, pp. 16-28. Ed. "Special Issue on Industrial OrganizationTopics in Banking," Review of Industrial Organization, fev, 1997. SAVING, T. Concentration Ratios and the Degree of Monopoly. International Economic Review, pp. 139-146, 1998. SEBRAE. Atuação das cooperativas de crédito no âmbito das Micro e pequenas empresas no Brasil. Sebrae – Serviço Brasileiro de Apoio às Micro e Pequenas Empresas Brasília, 2012.

SILVA, C. A. T; MORAIS, M. C. Concentração do Setor Financeiro após o Plano Real. 6º Congresso USP de Controladoria e Contabilidade. Anais..., 2006. SILVA, M. A. Impacto do Sistema Cooperativo de Crédito na Eficiência do Sistema Financeiro Nacional. Trabalhos para Discussão. Brasília: BCB, 2011. Disponível em: http://www.bcb.gov.br/pec/wps/port/TD246.pdf. Acessado em abril de 2013. SIMONS, K.; STAVINS J. Has Antitrust Policy in Banking Become Obsolete? New England Economic Review, pp. 13-26, 1998. SORGATO, L. A. A.; FERREIRA JUNIOR, R. R. ; SANTOS, P. R. Concentração bancária e desigualdade regional: o caso do Nordeste brasileiro. In: Encontro Regional de Economia, 2010. Anais eletrônicos... Fortaleza, 2010. Disponível em: http://www.bnb.gov.br/content/aplicacao/eventos/forumbnb2010/docs/concentracao.pdf Acesso em: 03/05/2013.

TAVARES, M. A. R. Concentração bancária no Brasil: uma evidência empírica. Revista de Administração de Empresas. TOKLE, R. J.; TOKLE, J. G. The Influence of Credit Union and Savings and Loan Competition on Bank Deposit Rates in Idaho and Montana. Review of Industrial Organization, pp. 427-439, 2000. VASCONCELOS, M. R.; FUCIDJI, J. R.; SCORZAFAVE, L. G.; Assis, D. L. D. O todo e as partes: Uma análise da desigualdade de crédito entre os estados brasileiros e os determinantes do crédito bancário com a aplicação de dados de painel. Economia e Sociedade, v. 22, n. 1. p. 123149, 2004.

13

Suggest Documents