W O R L D B A N K W O R K I N G P A P E R N O . 4 9
Enterprise Size, Financing
Patterns, and Credit
Constraints in Brazil
Analysis of Data from the Investment
Climate Assessment Survey
Anjali Kumar
Manuela Francisco
THE WORLD BANK
W O R L D B A N K W O R K I N G P A P E R N O . 4 9
Enterprise Size, Financing Patterns,
and Credit Constraints in Brazil
Analysis of Data from the Investment
Climate Assessment Survey
Anjali Kumar
Manuela Francisco
THE WORLD BANK
Washington, D.C.
Copyright � 2005
The International Bank for Reconstruction and Development/The World Bank
1818 H Street, N.W.
Washington, D.C. 20433, U.S.A.
All rights reserved
Manufactured in the United States of America
First Printing: April 2005
printed on recycled paper
1 2 3 4 5 07 06 05
World Bank Working Papers are published to communicate the results of the Bank's work to the
development community with the least possible delay. The manuscript of this paper therefore
has not been prepared in accordance with the procedures appropriate to formally-edited texts.
Some sources cited in this paper may be informal documents that are not readily available.
The findings, interpretations, and conclusions expressed herein are those of the author(s)
and do not necessarily reflect the views of the International Bank for Reconstruction and Devel-
opment/The World Bank and its affiliated organizations, or those of the Executive Directors of
The World Bank or the governments they represent.
The World Bank does not guarantee the accuracy of the data included in this work. The
boundaries, colors, denominations, and other information shown on any map in this work do
not imply and judgment on the part of The World Bank of the legal status of any territory or
the endorsement or acceptance of such boundaries.
The material in this publication is copyrighted. Copying and/or transmitting portions or all
of this work without permission may be a violation of applicable law. The International Bank
for Reconstruction and Development/The World Bank encourages dissemination of its work
and will normally grant permission promptly to reproduce portions of the work.
For permission to photocopy or reprint any part of this work, please send a request with
complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
MA 01923, USA, Tel: 978-750-8400, Fax: 978-750-4470, www.copyright.com.
All other queries on rights and licenses, including subsidiary rights, should be addressed to
the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA,
Fax: 202-522-2422, email: pubrights@worldbank.org.
ISBN-10: 0-8213-6129-5 ISBN-13: 987-0-8213-6219-0
eISBN: 0-8213-6130-9 ISSN: 1726-5878
DOI: 10.1596/978-0-8213-6129-0
Anjali Kumar is Lead Financial Economist in the Finance cluster of the Latin American and
Caribbean Region of the World Bank. Manuela Francisco is Consultant to the World Bank on
leave from the University of Minho.
Library of Congress Cataloging-in-Publication Data has been requested.
Contents
Preface v
Introduction 1
Firm Size, Financing, Access to Credit, and Credit Constraints 10
Financial Institution Ownership and Access to Credit 16
Financial Access as an Obstacle to Growth Compared to Other Variables 18
Conclusion 19
Appendix 23
References 57
LIST OF TABLES
1. The Dataset: Characteristics of Sample Firms 7
2. The Dataset: Alternative Classifications of Firm Size 9
3. Firm Size and Sources of Finance: Working Capital and New Investments 11
4. Bank Ownership: No. and Percentage of Firms by Ownership Category 16
5. Access to Credit and Credit Constraints--Breakdown per Type of Bank 17
6. Firm Size and Finance Related Obstacles to Growth 19
A.1. GDP, Population, and Branch Density per State 23
A.2. The Dataset (Size, Region, Industry, Manager's Education, Sales Growth) 24
A.3. Definition and Construction of Variables 25
A.4. Source of Finance--Working Capital 28
A.5. Source of Finance: New Investments 30
A.6. Overdrafts, Credit Lines and Trade Credit 32
A.7. Firm Size and Number of Banks Firms Do Business with 34
A.8. Size, Region, Education, Industry, and Sales Growth Effects
on Access to Credit and Credit Constraints 36
A.9. Reasons for Not Applying for a Bank Loan and Reasons
for Bank Loan Rejection 38
A.10. The Importance of Collateral and Shares of Collateral 40
A.11. Regression Results--Firm Characteristics, Performance
and the Probability of Having a Loan 42
iii
iv Contents
A.12. The Impact of Firm Size on the Likelihood of Having a Loan: Model 2 44
A.13. The Likelihood of Having a Loan According to Its Duration 46
A.14. The Impact of Bank Ownership on the Firm's Likelihood of Having
a Loan--Model 2--Sample Split by Bank Ownership 48
A.15. The Impact of Bank Ownership on the Firm's Likelihood of Having
a Loan--Model 2--Consolidated Sample 50
A.16. Probability of Having a Loan from a Public Bank or a BNDES Credit Line 52
A.17. Obstacles to Growth--Firm Size and Other Factors 54
A.18. The Relative Importance of Obstacles to Growth and Firm Size 55
Preface
T his paper investigates the importance of firm size with respect to access to credit,
relative to firm performance, and other factors which may affect creditworthiness,
such as management education, location, or the industrial sector to which the firm
belongs. The principal findings are that size strongly affects access to credit, compared to
performance as well as other variables, suggesting quantitative limitations to credit access.
Looking at short-versus long-term loans, the impact of size on access to credit is greater
for longer-terms loans. Further, looking at the ownership of the lending institution, it is
found that public financial institutions are more likely to lend to large firms. Finally,
examining the role of financial constraints relative to other constraints faced by the firm,
it is found however that financial access constraints may have a less significant differen-
tial impact across firms of different sizes than other constraints though cost of finance as
a constraint is very important.
The authors are grateful to Thorsten Beck, Gledson Carvalho, Soumya Chattopadhyay,
Marianne Fay, Luke Haggarty, Patrick Honohan, Leora Klapper, Leonid Koryukin, John
Nasir, Maria Soledad Martinez Peria, Mark Thomas, and Jos� Guilherme Reis for their
valuable comments on earlier versions.
v
Introduction
Should firm size affect the ability of a firm to access external capital for growth? If access
to external financing is based on current performance, or expected future performance--
that is, on returns or expected returns--size per se should not have an impact on access
to external finance. Yet in many countries it is perceived that small firms face particular
disadvantages in the credit market.
This paper examines the extent to which firm size affects financing patterns and
restricts access to finance in one country, Brazil, based on an Investment Climate Survey
of 1642 firms constructed in 2003, which includes firms in thirteen Brazilian states (out of
27) and nine industrial groups. The following key questions are addressed: (i) whether
small firms financing patterns differ from large firms, and whether small firms have less
access to credit and face more credit constraints than larger firms; (ii) the importance of
firm size, compared to performance, or other factors, in assessing access to credit and credit
constraints; (iii) whether credit provision criteria are different for fixed capital (long-term
loans) and for working capital (short-term loans), (iv) whether bank ownership--public,
private or foreign--impacts differentially upon on credit provision across firm sizes, and
(v) the role of credit constraints relative to other constraints, in relation to firm size.
The present section discusses the questions examined, reviews results of former studies
on firm size and access to finance, and discusses the data sample and the variables used in
the present investigation. Section 2 investigates financing patterns by firm size and ana-
lyzes differentials in access to credit, evaluating the role of size, among other factors, as
a constraint to financial access. Section 3 examines the differential impact of financial
institutions' ownership on the provision of credit to firms of different sizes. Section 4
investigates the role of financial access as a constraint to growth, relative to other factors,
for firms of different sizes. Finally, Section 5 presents overall conclusions.
1
2 World Bank Working Paper
Firm Size, Performance, and Characteristics:
Impact on Financing and Access to Credit
Studies of the extent to which firm size affects financing patterns, at the cross country
level, have looked primarily at differentials in debt equity ratios, and results suggest that
size does affect financing patterns (Demirgu�-Kunt and Maksimovic 1999). Large firms
have more long-term debt as a proportion of total assets compared to smaller firms, and
are more likely to use external finance compared to small firms (Beck, Demirgu�-Kunt,
and Maksimovic 2002, 2003). More disaggregated investigations of sources of finance
have also looked at the use of trade credit, finding that large firms are significantly asso-
ciated with less trade credit finance (Demirgu�-Kunt and Maksimovic 2001). The greater
use that smaller firms make of trade credit is more prominent in countries where the
legal infrastructure is weak. As the legal infrastructure strengthens, across a spectrum of
countries, the use of trade credit is reduced for all firm sizes. Moreover, comparing bank
financing and trade credit, these studies suggest that size plays a larger role in access to
bank financing than in access to trade credit. In the present study, data from the Invest-
ment Climate Survey on Brazil permits disaggregation of sources of financing into a
wider spectrum, beyond debt and equity finance, or bank finance versus trade credit. It
also permits the separation of financing sources for short and long term capital.
In assessing the factors which would affect access to credit, traditional theory would
suggest that in well-functioning credit markets, lenders would base their decisions on the
overall financial soundness of firms and on expected performance and projected cash
flows, adjusted for risks and transaction costs, rather than upon firm size. Measures read-
ily available for expected performance, adjusted for risks, are difficult to construct, how-
ever at a very simple level, many authors have found that greater sales and profits are
associated with greater access to credit (for example, Bigsten and others 2003; Topalova
2004). In addition, firms with increasing sales, increasing turnover (sales/assets) ratios,
lower volatility of sales or lower liabilities to assets ratios, would be expected to have greater
access to credit and less credit constraints.
Yet, empirical studies have also found that smaller and younger firms are more credit
constrained than larger and long established firms. Bigsten and others (2003) also report
that small firms are less likely to obtain a loan than large firms. Levenson and Willard
(2000) find that constrained firms are smaller, younger, and more likely to be owned by
their founders. Furthermore, Levy (1993) reports that lack of access to finance emerges as
the binding constraint for smaller and less established firms.1
Several reasons have been pointed out why access to credit may be affected by firm size in
addition to performance. First, greater constraints may be faced by small firms due to market
imperfections,intheformofgreaterinformationalopacity.Thoughnotuniquetosmallfirms,
this may be considerably more relevant because of relatively poor quality and provision of
financial information. This leads to greater difficulties in credibly conveying their quality or
the quality of their projects (Binks and Ennew 1996). Small firms, and especially small young
1. This analysis presents however two caveats. One is that empirically it is difficult to disentangle
creditworthy firms from non-creditworthy firms and therefore it is unclear if higher constraints are well
justified or not. Moreover, a survival bias hides important information regarding non-surviving firms
whose failure may result from credit constraint.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 3
firms, lack the long credit history of larger and longer established firms. Also small firms do
not have publicly-known contracts (supplier, customer, or labor-related), and do not trade
securities that are continuously priced in public markets. Moreover, unlike large firms their
performance is not regularly assessed by independent market analysts, and they may be unable
to provide audited financial statements (Berger and Udell 1998; Saito and Villanueva 1981).
External financial agents must consider the provision of finance under imperfect and asym-
metric information (Berger and Udell 1994) related both to the ex ante evaluation of the pro-
ject and the firm and the ex post monitoring of performance. Information is particularly
important for debt financing, where the lender is not a beneficiary of upside gains, but is a
potential loser in the event of downside firm failure. It has been argued that such information
asymmetries, and thus adverse selection and moral hazard, lead to credit rationing (Stiglitz
and Weiss 1981); a situation where, with a given total supply of credit, some entities are unable
to obtain a loan at any interest rate. Such credit rationing may explain the credit constraints
that small firms face (Lung and Wright 1999; Berger and Udell 1994).
Second, to the extent that the adverse effects of information asymmetry may be
reduced by the provision of collateral (Angelini and others 1998; Berger and Udell 1994)
it is argued that smaller firms face greater difficulties. Larger firms tend to own more assets
for collateral. Also in large firms, managers' investments in the firm can also constitute a
pledge of performance (Bester 1987; Binks and Ennew 1996). In the case of small (unlisted)
firms pledged collateral is often of a personal nature (Avery and others 1998). Greater
reliance on personal assets may discourage investments at the margin as they imply addi-
tional risk (Binks and Ennew 1996).
Third, in addition to informational opacity, small firms may be associated with real
risk differentials compared to large firms, since they are known to have a high failure rate
compared to larger firms (Lund and Wright 1999; Gertler and Gilchrist 1994). Small and
especially new firms and may also have relatively more volatile earnings due to less oppor-
tunities for diversification of their output or client base (Chittenden and others 1993;
Hughes and Storey 1994; Klapper and others 2002). Smaller firms may thus be less likely
to survive economic downturns (Gertler and Gilchrist 1994). Evidence has shown that
small business closures occur in the first three years of operations (Bank of England, 1994).
By contrast, larger firms can potentially be more diversified and thus better protected
against economic fluctuations (Brewer and others 1996; Saito and Villanueva 1981).
Furthermore, larger firms are usually older and better established, which itself demonstrates
their survival under market competition.
Such differences between large and small firms are translated into higher bank trans-
action cost of lending to small firms. These real transaction cost differentials refer to search,
information, evaluation, monitoring as well as higher risk. Saito and Villanueva (1981)
estimate the real cost of lending to small firms being approximately twice that of lending
to large firms. In the present study, the extent to which small firms face greater credit con-
straints is empirically examined, and the importance of size differentials is compared with
variables reflecting firm performance, adjusted as far as possible for risk.
Other Factors Affecting Access to Credit
Looking at other variables which could affect firms' access to finance, it has been suggested
that there may be an "industry effect." Banks may favor firms of specific industries as clients,
4 World Bank Working Paper
lending more to `growth' industries (Rajan and Zingales 1998). An alternative explanation
for an industry effect is that some industries are more likely to depend on external financ-
ing than others, depending upon initial project scale, cash flows and requirements for con-
tinuing investment (Rajan and Zingales 1998; Bigsten and others 2002).2 Industrial effects
could thus be hypothesized to arise from factor intensity differentials, so that more capital-
intensive firms, with higher credit needs, may face proportionally greater constraints.
There may also be a "regional effect" so that financial access differentials in different
firm locations can arise from differentials in bank density across regions, which themselves
may reflect differentials in income and levels of economic activity. In Brazil there are sharp
income differences between the five main regions, where the Southeast is three times as
rich as the Northeast in per capita income terms. Kumar and others (2004) find that there
is a large variation in branch density across different regions of Brazil. While the South and
Southeast are relatively well branched, access to bank branches is relatively limited in the
North and Northeast. Well branched regions, and as a consequence, greater ratios of banks
per firm would be expected to ease physical access and also lower information asymmetry
problems and as a result ease credit access.3
Next, there may also be an "ownership" effect of the firm (private domestic, private
foreign, or state) and credit access. Foreign firms may have more access to credit and less
credit constraints than domestic private firms. Foreign firms are usually highly visible, well
known and publicly listed and traded. Previous studies in Brazil suggest that foreign firms
outperform domestic counterparts (Willmore 1986). State firms may have more credit
access (especially from public banks) relative to private domestic and private foreign firms.
If it is argued that state firms are generally obliged to make their financial situation public,
decreasing the agency costs associated with information asymmetries, such firms would be
expected to have superior access. One the other hand, if access to credit depends on per-
formance, state owned firms have often been shown to perform less well than private firms
(for example, Majumbar 1998; Vinning and Boardman 1992) which would suggest that
state firms should be more credit constrained than private firms.
The extent to which different levels of managerial education affect access to credit and
credit constraints is also explored. This has not been addressed in previous empirical stud-
ies. However, various authors have raised the importance of managerial education. Jensen
and McGuckin (1997) maintain that variations in firm performance are largely associated
not with traditional characteristics such as location, industry, size, age, or capital, but rather
with intangibles specific to the firm such as the managerial capital of the firm or the skill
of its workforce. At the individual level, Kumar (2004) found a strong education effect in
explaining access to financial services in Brazil. We expect that firms with more educated
managers have more access to credit than firms with less educated managers, as a result of
their ability to smooth complicated loan application procedures, presenting positive finan-
cial information, and/or building closer relationships with banks. Furthermore, better edu-
cated managers are more likely to have managerial skills in finance, marketing production,
and international business that would lead to firm's growth.
2. Another industry specific hypothesis could be to check for differential effects of government poli-
cies, which sometimes aim to promote specific sectors of the economy. In Brazil, government policy has
offered credit incentives to export oriented industries for example.
3. A state level analysis is not attempted in this paper.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 5
Bank Relationships, Bank Ownership and Access to Credit
Looking at the extent to which access to credit may be affected by the lender, studies have
pointed out that closer banking relationships could reduce transaction costs that emanate
from information asymmetries. Closer banking relationship can facilitate the flow of infor-
mation between borrower and lender, easing the bank's assessment of managerial skills,
business prospects, firm needs and resources. The better informed the bank the more it will
be able to apply prospects-based lending methods rather than collateral-based lending
(Binks and Ennew 1997). Closer relationships could be established through longer associ-
ation, uniqueness of association, or interaction over multiple financial products, that allow
the bank to learn about the firm's cash flows (Peterson and Rajan 1994). There is a broad
empirical literature with evidence that closer relationships (length of the relationship or
exclusive relations) are associated with lower credit constraints. Chakravarty and Scott
(1999) find that the relationship duration and the number of activities between households
and lenders significantly lower the probability of being credit-rationed. Cole (1988) finds
that a lender is more likely to extend credit to a firm that has an existing savings accounts
and other financial services. Also Peterson and Rajan (1994) report that the length of the
relationship has a positive and significant impact on credit availability. Ferri and Messori
(2000) report that close customer relationships between local banks and firms promote a
better allocation of credit in the North and Center of Italy but worse in the South.4
One measure used to proxy the closeness of bank relationships is the extent to
which such relationships are unique. Peterson and Rajan (1994) and Cole (1998) find
that firms that borrow from multiple banks are charged at significantly higher rates and
face lower availability of credit. These results are interpreted to suggest that multiple
relationships decrease the value of the private information generated by the potential
lender (Cole 1998). However, on the contrary, it has also been argued (Binks and Ennew
1996) that the vast majority of small firms do not need a close relationship with their
banks because they require standard services. Furthermore they state that banks need
to be selective when developing relationships since such services are costly in terms of
people and time. The present paper investigates the extent to which unique banking
relationships affect access to credit.
Another factor which may differentially affect access to credit for firms of different
sizes may be the ownership of the lending financial institution. Foreign banks may provide
more credit to large corporate firms for two reasons; first, foreign banks tend to "cherry
pick" good clients with the offer of superior services, and second, foreign banks are usually
located in large financial centers away from small firms (Berger, Goldberg, and White 2001;
Clarke and others 2001). Clarke and others (2001, 2002) find that foreign bank penetration
improves financing conditions for enterprises of all sizes, but this process seems to benefit
4. There are also studies that focus on the role of firm-lender relationships and the pricing of credit.
In Diamond (1989), Peterson and Rajan (1993), and Boot and Thakor (1994) it is predicted that loan
interest rates should decline over time though Greenbaum et al. (1989), and Sharpe (1990) maintain that
lenders charge lower interest rates in early periods. Empirically, studies have found contradictory results.
Peterson and Rajan (1994) find that the length of the relationship has no effect on the cost of credit. Berger
and Udell (1995) find that the cost of borrowing in credit lines decreases with long term bank--borrower
relationships and that collateral is less frequently required. The impact of bank relationships and the cost
of credit is not examined in the present study.
6 World Bank Working Paper
larger firms more. Public banks on the contrary may have a closer association with small
firms as they are often mandated to ease credit to small and new firms as a mean of over-
coming perceived market failures.
Other Factors Affecting Access to Credit
Heterogeneity of firms in terms of access to credit may also arise due to other characteristics,
which we broadly group under three categories: competitiveness, credibility, and capacity
for innovation. Competitiveness may be reflected in age, where survival suggests that firms
are at least as competitive on average, as other existing firms. Being an older firm should
also lower informational opacity (Frazer 2004).5 Another indicator of competitiveness, in
a global sense, is whether firms are exporters or not. Firms' transparency and credibility
should clearly affect their access to credit, and some researchers have pointed out that
formal sector firms may be deemed more transparent, or firms which are members of a
group or trade association (Binks and Ennew 1996). Finally, innovation and technological
change are majors drivers of economic growth (Solow 1957). At the firm and industry level,
recent contributions have found strong links between technological change and produc-
tivity, and between R&D and a firm's growth (Long and others 2003; Griliches 1998, for a
survey). Innovative capacity may be suggested by the education of the workforce as human
capital influences growth (Barro and Sala-i-Martin 1995), Lucas (1988), and Romer
(1990). The results of Laursen and others (1999) corroborate this thesis. They find that the
availability of a high fraction of employees with higher education was in general conducive
to growth.
Data and Sample Characteristics
Table 1 summarizes the sample composition according to region, industry, ownership,
manager's education, and sales growth. Looking at a simple parameter to measure firm
performance, about 65 percent of firms claimed to have increasing sales over the reference
period. In terms of region, firms are located mainly in the more affluent South and Southeast
(around 77 percent), The North and Northeast together make up 16 percent of the sample,
however the North alone accounts for only around 1.5 percent of the sample.6
In terms of industry, almost half the firms (46 percent) belong to the Garment and
Furniture sectors; over a fifth (21.7 percent) belong to the Machinery and Shoe and Leather
sectors, taken together. In terms of ownership, the vast majority of firms (94 percent) are
private domestic firms. Private foreign ownership and government ownership represent
5.3 percent and 0.4 percent of the sample respectively. Only seven firms are state-owned,
5. Our threshold is two years as the majority of Brazilian firms that leave the market do so within the
first two years (BNDES, 2003)
6. The Southeast, South, and Center-West are the richest regions, with per capita incomes of
R$ 9,316, R$ 9,387, and R$ 7,260, respectively. The Northeast and North are the poorest regions, with
incomes of R$ 3,255 and R$ 4,312 per capita, respectively. With regard to branch density, the Southeast
has the largest number of branches (9263), whereas the South and Center-West have 3446 and 1283
branches, respectively. The Northeast, the poorest region, has 2383 branches and North has only 623
branches. (Appendix Table A.1)
Table 1. The Dataset: Characteristics of Sample Firms
No. firms No. firms No. firms Manager's No. firms Sales No. firms
Region (%) Industry (%) Ownership (%) education (%) growth (%)
North 24 Food 127 Private 1549 Post 331 Increased 1042
(1.5) Processing (7.7) Domestic (94.4) Graduate (20.2) (64.6)
Northeast 238 Textiles 106 Private 86 Graduate 500 Decreased 390 Enterprise
(14.5) (6.5) Foreign (5.2) (30.5) (24.2)
Center-West 121 Garments 442 State 7 Incomplete 249 Unchanged 182
(7.4) (26.9) (0.4) University (15.2) (11.3) Size,
Southeast 713 Shoes & 173 Vocational 185
(43.4) Leather (10.5) Training (11.3) Financing
South 546 Chemicals 84 Secondary 158
(33.2) (5.1) School (9.6)
Machinery 183 Incomplete Sec. 62 Patterns,
(11.2) School (3.8)
Electronics 79 Primary School 95
(4.8) (5.8) and
Auto-parts 130 Incomplete 60 Credit
(7.9) Primary School (3.7)
Furniture 315 Constraints
(19.2)
Source: World Bank, Investment Climate Survey--Brazil, 2003. in
Brazil
7
8 World Bank Working Paper
of which six belong to the chemicals industry and one belongs to the electronics industry.
State owned firms are large; three have more than 500 employees, six out of seven have
annual sales of more than R$60 million per year. By contrast only 3.6 percent of private
domestic firms have more than 500 employees and only 8.5 percent have sales of over R$60
million per year. Foreign-owned firms account for 5 percent of the sample, and around
half are in the Machinery and Auto-parts industries. Foreign private firms are larger than
domestic private firms; a fifth have more than 500 employees, and over a third have sales
exceeding R$60 million.
Managers of about half the firms have completed university education. Yet, in 10 per-
cent of firms, the manager's education does not exceed primary school. In more techno-
logically intensive sectors such as Chemicals and Electronics, 80 percent of the managers
hold a post graduate degree.
Measures of Firm Size
Alternative criteria for classifying firm size were tested. The most widely used criterion in
Brazil is the number of employees, as defined by the Ministry of Industrial Development
and External Trade.7 This classification has also been adopted by the Brazilian Institute of
Geography and Statistics (IBGE) and the Institute for the Support of Micro and Small Firm
(SEBRAE).8
An alternative classification, based on sales volume, is used by Brazil's development
Bank (the BNDES).9 In addition, classification of firms by size deciles and quintiles was also
investigated. For the most part, the study uses only the first definition, since there appears
to be a high degree of co-movement of findings using alternative definitions. Using both the
sales criterion and the number of employees, micro and small firms represent the largest
share of the sample; around 70 percent taken together (Table 2). Micro firms form the
largest share of the sample according to the sales criterion (46 percent of firms, with annual
sales of around R$1.2 million); small firms represent the largest share on the employment
criterion (52 percent, employing between 20 and 99 workers). A breakdown of the sample
by firm size and by select firm characteristics is presented in Appendix Table A.2.
Construction of Other Variables
To test the hypotheses described above regarding firms' access to credit, the variables
described above were constructed as follows: Firms' performance is proxied by a series of
7. Minist�rio do Desenvolvimento Ind�stria e Com�rcio Exterior. Note that this classification leads
to an uneven distribution of firms in each sample category; a higher threshold for micro firms or a lower
threshold for large firms could have corrected this. However apart from its widespread use within Brazil,
this definition also coincidentally corresponds to that used by the Bank in all other ICA data analysis.
8. Instituto Brasileiro de Geografia e Estat�stica and Servi�o Brasileiro de Apoio �s Micro e Pequenas
Empresas.
9. Banco Nacional de Desenvolvimento Econ�mico e Social., or National Bank for Economic and
Social Development. SEBRAE uses a different definition for size according to sales. It follows the defini-
tion of Law 9841 of 10/5/99, in which a firm is classified as micro if its sales are lower than R$244,000;
small if its sales are equal or greater than R$244,000 and lower than R$1,200,000; and medium or large if
its sales are equal or greater than R$1,200,000.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 9
Table 2. The Dataset: Alternative Classifications of Firm Size
Number of Number Sales Number
employees (Nos.) of firms % (R$ 000 per year) of firms %
Micro 0 to 19 330 20 <1,200 736 46
Small 20 to 99 861 52 1,200 & <10,500 468 30
Medium 100 to 499 376 23 10,500 & <60,000 268 17
Large More than 500 75 5 60,000 170 7
500�999 53
1000�1999 12
2000�4999 7
>5000 3
Total 1642 100 1642 100
Source: World Bank, Investment Climate Survey--Brazil, 2003.
variables including sales growth, turnover (sales to asset ratio), and leverage. For regional
effects, five standard national regions are introduced as variables: North, Northeast, South,
Southeast, and Center-West. Dummy variables for these are weighted by regional income
per capita and by bank branch density. For industrial effects, nine industrial sectors are
introduced, using the standard industrial (CNAE) classification, weighted by capital inten-
sity, measured as the ratio of machinery and equipment costs to labor costs.10Managerial
education is captured at eight levels.11Firm ownership is classified in three categories; state-
owned, private domestic and private foreign. Bank ownership was classified similarly, for
each firm based upon the main bank the firm used.
Additional control variables include whether the firm age is below five years, and
whether or not the firm is an exporter (as measures of survival and competitiveness), firm
status (incorporated or not); membership of a trade group or association, and use of
external auditors, as measures of transparency. Finally, the proportions of the workforce
with higher education (proxied by the percentage of workforce that use computers), and
capacity utilization, were used as measures of innovation and capacity utilization.
The last group of variables, on bank relationships and creditworthiness, were mea-
sured by whether the firm has a unique bank relationship, whether the firm has collateral,
whether the firm has an overdraft or line of credit, and finally, by the ownership of the main
banking institution for each firm. A list of variables and their construction is given in
Appendix Table A.3.
10. Textiles, Auto-Parts, Chemicals, Food Processing, Electronics, Machinery, Furniture, Leather &
Shoes, and Garments.
11. Post graduate degree, university degree, incomplete university degree, vocational training after
secondary school, complete secondary school, incomplete secondary school, complete primary school,
and incomplete primary school.
10 World Bank Working Paper
Firm Size, Financing, Access to Credit, and Credit Constraints
Our analysis of access to financial services and firm size begins with a simple comparison
of financing patterns across firms of different sizes. This is followed by a more specific
question related to the role of size compared to performance and firm characteristics in
explaining access to credit. Two models have been specified, to test the robustness of
results obtained.
Firm Size and Financing Patterns
Based on data in the survey which provides a detailed breakdown of sources of funds
(internal capital, banks, trade credit, leasing, credit cards, government funds, and informal
sources), and separates these by uses (fixed and working capital, we use mean difference
tests to investigate whether the sources of funds vary significantly across firm sizes.12
Results are summarized in Table 3 later and detailed in Appendix Table A.4 and Appendix
Table A.5 . In terms of importance, for all firm sizes, and for both working capital and for
new investments, internal funds constitute the primary source of finance, especially for
fixed capital (55 percent, compared to 45 percent for working capital).13 Next in importance
as a source of finance, for both working capital and new investments, is credit from the
banking system, followed by trade credit, which for working capital contributes a sub-
stantial 14 to 16 percent of total financing. Informal sources can be important for working
capital finance. Leasing, credit card finance, and equity play a minor role as financing
sources.14
Looking at financing patterns across firms of different size, the findings which stand
out are first, that differentials by size may be more pronounced for fixed capital than for
working capital. In terms of the overall separation between external and internal funds,
large firms use significantly more external funds to finance new investments (59 percent
compared to 41�46 percent for other size categories). For working capital, differences are low
(44.2 compared to 41.2 percent, and there is no steady progression across size categories).
Trade credit too does not appear to vary systematically by firm size for working capital,
however its is surprisingly also important as a source of finance for new investments, and
here its importance does vary across firm size, representing around 12 percent for micro
firms and between 7 and 9 percent for other firm sizes.15For bank finance and for funding
12. F-tests and Chi-Squared-Tests. Note that these can only test for differences from the mean and
not for individual pairs of categories. Thus for example we cannot test whether the north is significantly
different from the south, or whether the southeast is significantly different from the north. We test for sig-
nificant differences in the use of internal funds across regions.
13. The results are corroborated by previous findings for Brazil. Brazilian firms primarily rely on
internal finance, secondly, on debt finance and thirdly, on equity (Junior and Melo, 1999), confirming the
Pecking Order theory. Equity finance represents a more important source of financing for larger firms
than for other firms reflecting the equity gap.
14. Credit card use for financing working capital varies significantly (at 5%) across firm size when
firms are classified according to sales only. Equity as source of financing for new investment varies sig-
nificantly across firm size, being more important for medium and large firms, when size is defined accord-
ing to sales and deciles and quintiles of sales.
15. Internal funds, local bank finance and trade credit represent around 80% of the total of the sources
of financing for all firm sizes.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 11
Table 3. Firm Size and Sources of Finance: Working Capital and New Investments
Working capital New investments
Micro Small Medium Large Micro Small Medium Large
No. of employees 0�19 20�99 100�499 >500 0�19 20�99 100�499 >500
Internal funds 44.2 43.3 44.8 41.2 58.7 57.8 54.0 41.0
Bank finance1
Foreign 0.8� 0.9� 1.7� 4.9� 0.0� 0.8� 2.6� 3.2�
Local private 10.8 12.7 12.6 8.5 5.7 6.9 5.4 1.4
Local public2 11.9* 15.2* 17.6* 25.2* 10.4� 14.1� 19.1� 34.5�
Of which 0.8� 1.9� 2.9� 6.0� 4.5� 6.5� 12.5� 25.3�
government funds
Trade credit 14.2 16.3 13.7 14.2 11.9* 8.6* 6.6* 9.2*
Leasing 0.5 0.9 0.8 0.3 2.2 3.1 3.5 5.0
Informal sources 10.5� 5.5� 1.8� 0.2� 4.4� 2.4� 0.4� 0.0�
Equity finance 2.7 2.7 4.7 1.8 3.5 3.8 6.0 4.0
Credit card finance 0.8 1.0 0.3 0.0 0.5 0.2 0.2 0.0
Others 3.6 1.5 2.0 3.7 2.7 2.3 2.2 1.7
Total (%) 100 100 100 100 100 100 100 100
Total no. of firms 328 860 373 72 247 716 324 64
1. This disaggregation does not derive directly from the questionnaire. Local commercial bank finance
is disaggregated into local private and local public finance according to the main bank the firm
does business with.
2. Government funds are included in the local public bank finance category.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: Based on World Bank, Investment Climate Survey data--Brazil, 2003.
from informal sources, there are significant differences across size categories for both fixed
and working capital. Informal sources are very important for working capital finance for
micro firms, representing 10.5 percent of working capital financing needs for micro firms,
compared to only 0.2 percent for large firms.16
Second, a larger percentage of firms among medium and large firms have overdrafts
or line of credit (81 and 83 percent respectively), compared to micro and small firms
(60 and 76 percent respectively). As firm size increases the amount available through an
overdraft or credit line as a percentage of sales increases sharply (from 33 percent for micro
firms to 546 percent for large firms). Moreover, micro and small firms are charged higher
interest rates on their overdrafts (around 5 percent) compared to medium and large firms
(3 and 4 percent respectively). Sample data suggests that as size increases, the number of
banks firms do business with also increases (Appendix Table A.6).
16. This also suggests that our later analysis of the impact of size on financing patterns could have been
enhanced if the use of specific credits requested or received was known. Unfortunately, information on
this has not been provided.
12 World Bank Working Paper
Third, separating banks by ownership, it emerges that public banks are more signifi-
cant providers of capital for larger firms.17 Micro firms use public banks for only 12 percent
of their working capital needs and 10 percent of new investment finance, in contrast to 25
and 34 percent for large firms. Private commercial banks by contrast appear to supply micro,
small and medium firms with a larger proportion of their needs than large firms, especially
working capital needs (11�13 percent, compared to 8.5 percent for large firms). Private
commercial banks account for a negligible proportion of large firms' working capital needs
(only 1.4 percent, compared to 5.4�6.9 percent for micro to medium firms). Foreign com-
mercial banks like public banks are far more important for large firms, and even provide
for a significant part of their working capital needs (5 percent), in addition to the finance
of fixed capital (3.2 percent).18
Sources of financing appear also to be affected by the other explanatory variables;
region, manager's education, industry and sales growth. Better off regions use a higher
proportion of external funds than poorer regions. Thus, the South uses less internal funds
and more commercial bank finance, for both working capital and fixed investments,
compared to other regions, while the North uses twice as much internal finance as other
regions. In terms of the number of bank relationships, as size increases, the number of
banks clearly increases ( Appendix Table A.7 ). In terms of region and education, firms in
the South work with a larger number of banks on average than firms from other regions.
An examination of managerial education suggests that firms where managers holds
post-graduate degrees use more finance from foreign banks and equity finance compared
to other firms. More educated managers also work with a larger number of banks
(Appendix Table A.8 ).
Access to Credit and Credit Constraints--Sample Frequencies
Moving from overall patterns of financing, to access to credit specifically, the next part of
the analysis examines the relation between constraints in access to credit and firm size,
performance, and other factors. Firms with access to credit are defined as those that
express a demand for credit, apply for a bank loan and receive it.19 Constrained firms are
those that express a demand for a bank loan but either (i) apply for a bank loan and are
rejected, or (ii) do not apply.20 The data shows that 59 percent of large firms have loans,
compared to 27 percent of micro firms. About 54 percent of large firms that did not apply
for credit reported that they did not need a loan, compared to 39 percent of micro firms.
About 61 percent of micro firms that did not apply for a bank loan reported other reasons
17. Local commercial banks were not separated into private and public banks in the data on financ-
ing sources. However the public bank share has been constructed by inference, using the name of the prin-
cipal bank provided by each respondent.
18. These results are similar to those in Kumar (2004) which reports that for individuals, private banks
were more active for small depositors and small loan segments than public banks.
19. This is access to credit in a narrow sense. In a wider definition, firms that do not have a loan but
also have no demand (either because there is no need or because they can finance their needs in other
ways) can also be defined as having access to credit.
20. Reasons cited in the questionnaire for not applying despite expressed demand include factors
related to the environment such as complicated application procedures, corruption in the allocation of
bank credit, or expectation of rejection, as well as cost related factors such as high interest rates or strict
collateral requirements.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 13
(such as application procedures, collateral requirements, interest rates, or expectations of
being rejected) compared to 46 percent among large firms. Only 2.7 percent of large firms
did not have a loan because their application was rejected, compared to 9.4 percent for
micro firms. About 38 percent of micro firms did not apply for bank loans (even though
they needed one) because of other reasons cited above. For large firms that percentage
corresponds to 18 percent.
Cost-related factors, in the form of high interest rates, are the principal reasons cited
for not applying for a loan, and for this, the proportion of affected firms is similar for all
firm sizes (Appendix Table A.9).
Application procedures and collateral requirements are next in importance, and these
represent a higher barrier for micro and small firms than medium and large firms. None
of the large firms failed to apply for a loan due to expectations of being rejected, unlike
micro and small firms. Corruption and expectations of being rejected are not reported as
important barriers.21
Around two thirds of all loans (67 percent) require collateral, which on average rep-
resents around 125 percent of loan value (Appendix Table A.10). Collateral is used for a
larger proportion of large firms' loans (81 percent) compared to micro firms (43 percent).
Buildings and machinery together form the largest share of collateral for firms of all sizes,
together representing around half of all collateral. The use of personal assets and intangible
assets as collateral does vary significantly across firm size. Large firms use less personal assets
(7 percent) compared to other firms (between 10 and 20 percent), but more intangible assets
(35, compared to 11 to 17 percent for other firms).
Looking at other factors which could affect access to credit and credit constraints, it is
found first a simple performance related variable, sales growth, does exhibit an association
with access to credit but the result is not significant statistically. Firms with decreasing
sales have a greater rejection rate (15.5 percent) compared to firms with increasing sales
(9.1 percent). And a large number of firms with declining sales do not apply for a loan
because they expect to be rejected (2.4 percent) compared to firms with increasing sales
(0.5 percent). Regional variations, by contrast, are significant. The percentage of firms with
loans is lower in the North (16.7 percent) than in the South (41.4 percent). And firms from
South are less credit constrained (28 percent) compared to firms from other regions
(between 31 and 46 percent).22 Managerial education does not vary significantly with the
percentage of firms that have loans though with regard to the reasons for not applying for
a loan (Appendix Table A.9), application procedures are a greater barrier for firms with less
educated managers compared to firms with more educated managers. About 18 percent of
the firms in which managers have incomplete primary education report application pro-
cedures to represent the main reason for not applying for a loan, compared to 5 percent of
firms in which the manager has a post graduate degree. About 40 percent of the firms with
the lowest educated managers report that loan application was the main reason for
21. An investigation of reasons for loan application rejection suggests lack of collateral and poor credit
history are the main factors. An analysis of size effects is limited since of the 193 observations, only 3 are
for large firms.
22. The requirement of collateral also varies significantly across regions. A smaller percentage of
firms in the North reported that financing required collateral (50%) compared to other regions (between
60% and 70%).
14 World Bank Working Paper
rejection, while only 12 percent of the firms with post graduate managers have reported
so. Finally, the percentage of firms across different industries that have a loan varies
between 30 and 40 percent but differences are not statistically significant.23
Relative Importance of Factors Affecting Credit: A Simple Model
Totestwhethersize,performance,industry,regionandmanager'seducationexplaintheprob-
ability of having a loan, we first estimate a maximum likelihood probit model incorporating
these variables, and estimate the marginal effects of these variables on access to credit as
defined above. Appendix Table A.11 reports the marginal effects. The results indicate first
that firm size dominates all other effects--region, industry, manager's education, firm
ownership, and performance. Small, medium and large firms respectively have probabilities
of having a loan which exceed micro firms by 9, 22, and 34 percentage points respectively.
The Relative Importance of Factors Affecting
Access to Credit An Alternative Model
In order to test the robustness of the results, an alternative estimation was undertaken,
using a two step maximum likelihood probit with sample selection, to deal with possible
selection bias between access to credit and demand for a loan.24 This model allows us to
estimate the probability of having a loan (or being unconstrained) given that the firm has
demand for a loan. In the first stage (first model) we estimate the probability of having
demand for a bank loan, and in a second stage (the second model) we estimate access to
credit defined by the probability of having a bank loan. The first model can be interpreted
as demand for credit and the second model as supply of credit. Firm characteristics and the
firm's willingness to invest25 explain the demand for credit. The supply of credit shall reflect
firms characteristics and the banks' evaluation of firms' risk.
Demand for credit = a + b firms'characteristics + d firm's willingness to invest + e
Supply of Credit = a + b firms'characteristics + d banks' evaluation of firms' risk + e
Firms' characteristics (which explain both models) are firm size, region, industrial
group, ownership, managers' education, capacity utilization, age, exporter status, corpo-
rate status, group membership, and innovative capacity (percentage of workers that use
a computer regularly). In addition to firm characteristics, demand for credit is also
explained by proxies for firm's willingness to invest--captured here by whether a bank
has an overdraft or line of credit,26 the percentage of inputs bought on credit and cited
23. Firmownershipisnotinvestigated,sincethesamplemaybeunrepresentative,withonly7state-owned
firms and 86 foreign firms out of 1642 firms,
24. The selectivity bias derives from the fact that only firms with demand for credit will be in the mar-
ket for a loan.
25. Theoretically the willingness to invest (apply for a loan) should consider the cost of alternative
sources of financing, including internal sources of financing.
26. In the first model (demand for credit) overdraft is capturing the availability of alternative resources
to the bank loan, whereas in the second model is capturing firms'worthiness.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 15
macroeconomic obstacles to growth (economic uncertainty, macroeconomic instability,
and cost of credit). The access to a bank loan model is explained by firms characteristics
(as described above) and by variables that aim to capture firms' risk--performance variables
(turnover, sales growth, leverage),27 information transparency (external auditor), the
nature of the banking relationship (unique or not), and whether the firm has an overdraft
and collateral or not.
Appendix Table A.12 reports the results, which indicate first that medium and large
firms have a greater probability of having loans than micro firms. Being a firm with more
than 500 employees increases the probability of having a loan by 25 percentage points
compared to firms with less than 20 employees (micro firms). Being a medium-sized firm
(100�499 employees) increase the probability of having a loan relative to micro firms by
15 percentage point. Apart from size, the other relevant variables included innovative
capacity, as measured by the percentage of workforce that uses computers. An increase of
one percentage point in this segment of the workforce increases the likelihood of having a
loan by 4 percentage points. Additionally, having an overdraft has a positive impact on the
probability of having a loan (by 16 percentage points). Note that having a unique bank
relationship decreases the probability of having a loan, by 11 percent.
Next, to further investigate differences in access which may arise from loan duration
(i.e., linked to the purpose of the loan), we split the sample into long term loans and short
term loans. Loans with a minimum duration of 24 months are classified as long term,
while loans below this threshold are deemed to be short term. This threshold represents
a popularly used distinction between loans for working capital and for loans for fixed
capital in Brazil.28Appendix Table A.13presents the main findings: access to long term loans
varies with firm size, and also with workforce education, creditworthiness (as measured
by overdrafts) and the numbers of banks firms do business with. By contrast, the only
significant variable in explaining loans for working capital (short term loans) is having an
overdraft facility. Firms that have an overdraft facility increase their probability of having
a short term loan by 6.5 percentage points. Firm size, unique bank relationships, and
percentage of workers that use computers play no role in explaining short-term loans.
Only the overdraft facility is relevant in explaining short-term loans, suggests that loans for
working capital are treated as extensions of overdrafts. This may imply that small firms
may have easier access to credit for keeping the business running, while facing greater
financing obstacles for new investments that allow growth and expansion.
The findings above that the firms that work with only one bank are more credit con-
strained are not in line with previous work (Rajan and Zingales 1994) which hypothesizes
that the establishment of a unique banking relationship can aid access to credit. Firms
appear to find it beneficial to build up a relationship with several institutions.29
27. To mitigate the endogeneity problem we use lagged variables.
28. At the BNDES bank, loans for working capital in Brazil are defined to have a maximum of 24 months
and loans for fixed capital have a minimum of 24 months and a maximum of 120 months.
29. The findings of Rajan and Zingales, 1994, focused on the effect of unique banking relationships
on lowering the cost of credit, however, rather than on raising quantitative access. In the present exercise
a specification with the numbers of banks as opposed to the unique versus multiple bank relationships
was also examined and results were similar.
16 World Bank Working Paper
Financial Institution Ownership and Access to Credit
The previous sections focus on the characteristics of the enterprises. This section aims to
characterize the finance provider, in particular the finance provider's ownership.
Domestic banks are the principal financial institutions which sample firms deal with,
and public banks (45 percent of enterprises) are somewhat more important, in terms of
numbers of firms, than private banks (42 percent or enterprises).30 Private foreign banks
are the principal institutions for only 12.7 percent of sample firms (Table 4).
Banco do Brasil, a public domestic bank, is the principal bank for 593 firms, or 36 percent
of the total sample. It is also the most important financial institution for small firms,
though micro firms appear to engage most importantly with the Caixa Economica Federal,
the second largest bank, also publicly owned. In contrast to Banco do Brasil, Caixa
Economica Federal's clients include few mid sized firms and no large firms. The second
most important bank for firm of all sizes is Bradesco, a privately owned domestic bank. Its
importance as the main bank does not vary across firm size.31
A larger percentage of firms which are clients of public banks have loans (53 percent)
compared to firms which are primarily private bank clients (42 and 45 percent, respectively).
Also a larger percentage of firms which are clients of public banks have overdrafts (80 percent)
compared to firms that work with private domestic and private foreign banks (70 and
76 percent, respectively). Furthermore, a lower percentage of firms that work primarily
with public banks have bank loan rejections (13 percent) compared to firms that work
with private domestic and private foreign banks (21 and 14 percent, respectively), and a
Table 4. Bank Ownership: No. and Percentage of Firms by Ownership Category
Type of institution No. of firms %
Domestic Private Banks 687 42.3
Foreign Private Banks 207 12.7
Public Banks 725 45.0
Total 1626 100
Source: World Bank, Investment Climate Survey--Brazil, 2003.
30. Data on bank ownership are not requested directly in the questionnaire, however firms are
asked to name the financial institution which they principally use. The ownership of the banks named
was classified based on data provided by the Central Bank of Brazil. Only one firm reports to be doing
business with BNDES, which is a large second tier (wholesale) lender to enterprises. However, funds
from BNDES are channeled through both public and private banks, as lines of credit.
31. There is no significant difference in the type of bank firms do business with across firm size.
However firm ownership seems to be correlated with bank ownership. State firms do more business
with public banks and less with foreign private banks. Foreign firms do less business with public banks
(25%) compared to private domestic firms (46%), and more with private foreign banks (22%) compared
to private domestic firms (12%).There are significant differences in the type of banks firms do business
with across regions. While the percentage of firms in the South that do business with public banks is 59%,
the same percentage is 22% in the North. However, differences across regions do not appear to follow
regional income differences, and industrial differences do not reflect relative factor intensity.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 17
Table 5. Access to Credit and Credit Constraints--Breakdown per Type of Bank
Private domestic bank Private foreign bank Public bank
Have a loan (%) 42.4� 44.9� 53.4�
Loan application rejected 20.8 14.3 12.6
Constrained 55.8� 53.1� 43.8�
Have overdraft 70.1� 75.8� 79.6�
Required collateral 67.3 65.2 67.4
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: World Bank, Investment Climate Survey--Brazil, 2003.
lower percentage of firms that work with public banks are constrained (44 percent) compared
to firms that work with private domestic and private foreign banks (56 and 53 percent,
respectively; see Table 5).
To test whether access to credit varies according to bank ownership we split the sample
according to bank ownership--that is, into (i) firms that work mainly with public banks,
and (ii) firms that work mainly with private banks.
The results illustrate that, from the sample of firms that work primarily with a public
bank, large firms are the most likely to have a bank loan (Appendix Table A.14). However,
among firms that work mainly with private banks,32 larger firms are not more likely to have
bank loans than smaller firms. For private banks, firms with higher technological and
innovative capacity (as measured by the number of workers that use computers), with
greater rate of sales growth and that have an overdraft, are more likely to have a loan. Nev-
ertheless, firms that work with more than one bank and that are new (below five years old)
are less likely to have a loan. In sum, the results suggest that for public banks firm size is
the main indicator of credit worthiness, whereas private banks resort on other indicators
such as performance (sales growth), whether the firm is new and whether the firm has an
overdraft or credit line. Furthermore, the results suggest that among their clients, public
banks may tend to favor large firms over small firms.
To further investigate the effect of bank ownership on the likelihood of having a loan
we add interactive dummies (firm size times public bank dummy), to capture whether the
effect of working with a public bank and the probability of having a loan varies with firm size.
If public banks aim to address market failures we should expect that smaller firms that work
with public banks are more likely to have a bank loan compared to small firms that work
with private banks. The results reported show (Appendix Table A.15), however, that
smaller firms that work primarily with public banks are not more likely to have a loan than
small firms that work with private banks. Together, these results suggest that first, public
banks clearly do not give privileged access to credit to micro and small firms, and second,
that among their clients, public banks may tend to favor large firms over small firms.
32. Private domestic banks and private foreign banks are combined, to even sample size for these
two categroies.
18 World Bank Working Paper
A second approach adopted for the analysis of the role of public banks focused par-
ticularly on the lines of credit extended by Brazil's wholesale, second-tier development
bank, the BNDES, to other banks, public and private, for investment loans. These lines
of credit, which have a minimum duration of 24 months and a maximum duration of
120 months, are a huge source of investment funding in Brazil.33 Assuming that all loans
within this category are via BNDES credit lines, we estimate the probability of having a
loan from a public source (directly via a public bank or via these BNDES credit lines). We
expect small firms and export-oriented firms to be more likely to have bank loans than
non-exporters.
The results show, on the contrary, that larger firms are more likely to access to
loans. Being a large, medium, or small firm increases the probability of having a loan by
27 percentage points, 24 percentage points and 12 percentage points respectively, compared
to micro firms (Appendix Table A.16). We also find that though BNDES seeks to promote
exporting firms, they are not more likely to have access to credit than non-exporting firms.
BNDES' own statistics tend to confirm these findings. Although every year large firms
capture a lower share of BNDES resources, they still receive the greatest proportion at
present--70 percent in 2003.34
Financial Access as an Obstacle to Growth Compared to Other Variables
To conclude the analysis, we investigate the importance of financial access as a constraint to
growth, relative to other constraints (Appendix Table A.17). This analysis is based on a ques-
tion which asks respondents to rank potential obstacles to growth in order of importance.
Costs of financing are reported to be the main obstacle to growth for 57 percent of all firms.
Access to financing is ranked seventh (34.5 percent of respondents) after cost of financing,
tax rates, corruption, economic and regulatory policy uncertainty, and macroeconomic
instability; however the question is narrowly interpreted.35 Clearly firms face a number of
obstacles and cost of financing may be a greater overall barrier in Brazil than access.
The question examined here however is the differential impact of various obstacles, and
especially, financial obstacles, across firm size. Both access to financing and costs of financ-
ing are smaller obstacles to growth for larger firms relative to other sizes. Only 25 percent
of large firms rated access to finance as a "very high" obstacle to growth, in contrast to
34 percent for medium and small firms and 38 percent for micro firms. The cost of financing
is classified as a very high obstacle to growth by 45 percent of large firms and by 57 percent
by firms of other sizes.36 However significant results were obtained for the impact of firm
33. According to a source within BNDES, it is directly and indirectly responsible for around 25% of
credit provision in Brazil.
34. In 2002, micro and small, medium, and large firms received, respectively, 16%, 6% and 78%. In 2003,
micro and small, medium, and large firms received, respectively, 22%, 8% and 70% (BNDES sources).
35. The question asks whether financial access, and specifically collateral, may be a barrier. However
this may suggest a narrow interpretation of financial access and lead to some exclusion in responses.
36. The probability of classifying access to finance as the a very high obstacle to growth is 24% for large
firms and 30% for other firms. The probability of classifying cost of financing as a very high obstacle to
growth is 37% for large firms and between 42% and 47% for other firms. These probabilities are based on
an ordered logit model.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 19
Table 6. Firm Size and Finance Related Obstacles to Growth
Access to financing Cost of financing
Micro Small Medium Large Micro Small Medium Large
No. of employees 0�19 20�99 100�499 >500 0�19 20�99 100�499 >500
No obstacle 16.5 13.4 14.3 14.7 8.3 4.2 7.0 2.7
Low obstacle 7.1 8.3 9.2 13.3 3.4 3.1 4.0 2.7
Medium obstacle 17.1 16.2 17.0 21.3 7.4 7.7 7.8 13.3
High obstacle 21.1 28.1 25.3 25.3 23.1 28.0 24.1 36.0
Very high obstacle 38.2 34.1 34.2 25.3 57.8 57.0 57.2 45.3
Total 100 100 100 100 100 100 100 100
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: World Bank, Investment Climate Survey--Brazil, 2003.
size and other obstacles to growth. Larger firms are less likely to rate tax rates and corrup-
tion as very high obstacles to growth (Appendix Table A.18).37
Conclusion
This paper investigates the importance of firm size, firm performance, and other factors
which may affect firms' access to finance. The specific questions examined are, first, the
extent to which financing patterns vary across firm size. Second, we examine the extent to
which small firms may have less access to credit and face more credit constraints than larger
firms. Third, we investigate the relative importance of firm size, among other factors, in
assessing access to credit and credit constraints. Fourth, we examine the extent to which
characteristics of financial institutions, in terms of ownership, differentially affect firms'
access to credit. Our final question is an analysis of finance as a perceived obstacle to
growth, compared to other factors, and the importance of such perceived obstacles across
firms of different sizes. The analysis is undertaken in the context of Brazil, using a survey
dataset based upon an Investment Climate Assessment, which provides information on
variables not included in previous work, including information on multiple sources and
uses of credit, bank ownership, firm size and ownership, as well as location, industrial sec-
tor, and other data.
Results suggest, first, that sources of finance vary by firm size, and moreover, size may
affect access to investment financing more strongly than to working capital financing. The
absence of data on uses of credit, in our analysis of credit constraints may limit the quality
of its conclusions.
37. Similar results are obtained using an ordered probit (where the predicted outcome is rating the
obstacle as a `very high' obstacle). For instance large firms are less likely to classify tax rates and cor-
ruption as very high obstacles to growth than micro firms by respectively 11 percentage points and and
17 percentage points.
20 World Bank Working Paper
Since money is fungible, is the distinction between these categories relevant? We would
argue that although long term loans may be diverted towards short term uses, it may not
be possible to obtain sufficient volumes of short term resources to satisfy significant long
term investment needs. Moreover, formal financial institutions make a clear distinction
between such loans (for example, the BNDES bank lines of credit are not usually
extended for periods of below 24 months). Data which indicate a significantly higher
proportion of internal funds for investment financing for all size categories would tend
to support this.
Next, our results clearly indicate that size is an important determinant of credit access
and credit constraints. Large and medium firms are more likely to have a loan, and less
likely to have credit constraints. Moreover, size appeared to have a much more significant
effect on determining access to credit than performance-related variables. Also, there is an
effective quantitative limit in the allocation of credit to smaller sized borrowers. Whether
such an allocation of credit can be deemed to suggest the presence of credit market fail-
ures, however, is not clear. To the extent that smaller firms are genuinely more risky for
lenders and involve higher transaction costs, or to the extent that there is strong informa-
tional opacity (or unreliability) below a certain threshold, the findings above may not
imply market failures. However, the limited significance of performance variables suggests
at the least, that lenders do not significantly base their decisions to lend on performance.
In addition, the results did not corroborate the hypothesis of a robust industry, region, or
education effect.
The foregoing analysis was limited by a number of factors, however, which could affect
its results. First, as pointed out above, the ICA questionnaire does not permit distinctions
between loans requested or obtained for fixed capital, or working capital. Second, we did
not undertake an analysis of the extent to which other financing sources apart from bank
loans (for example, trade credit or informal sources) behaved with respect to size, per-
formance or other factors determining their credit availability. Third, the nature of the
performance variables used was limited; in particular, the questionnaire did not permit
direct investigation of profits or returns on equity or assets. It was particularly difficult to
devise robust measures of risk adjusted returns and the only variable we have used for this
was sales adjusted for and weighted by age, as a risk proxy. Nevertheless, the absence of
significance of performance variables is striking.
Results also indicate that firms that conduct business with one bank only decrease their
probability of having a loan. Admittedly, the number of banks used by a firm is also
strongly correlated with size. Firms with overdraft facilities and with greater innovation
capacity (as measured by the proportion of the workforce which is educated) also exhibit
easier access to credit and less credit constraints. The unimportance of the unique bank-
ing relationship differs from previous work in this area (for example, Peterson and Rajan
2002) and seems to suggest possible gains to firms of diversifying their sources of finance,
whether because of lending limits or other reasons.
Third, our results suggest that public banks are the main source of finance for all firm
sizes; however, public banks appear to favor large firms somewhat more than smaller ones,
among their clients, and there is no evidence to suggest that public banks are addressing
significantly addressing this group or that micro and small firms receive proportionally
more credit from public banks than other firms.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 21
Again, our results were rendered difficult by the limitations of the data, where the
question on sources of finance did not distinguish between banks on the basis of owner-
ship. Therefore the share of private versus public banks was constructed on the basis of
data providing the main bank relationship for each firm, rather than the bank at which a
specific loan application was made or rejected. Second, the questionnaire also fails to dis-
tinguish between direct and indirect sources of public bank funding. In the case of Brazil,
a substantial volume of firm financing, especially perhaps, investment financing, is pro-
vided by a wholesale bank, the BNDES, through lines of credit extended to both public and
private retail banks. Efforts were made to capture this effect both via assumptions on gov-
ernment funds, typically channeled via the BNDES to private banks, and by trying to iden-
tify second tier relending with the knowledge of the term for such loans.
Fourth and finally, cost of financing and access to financing are among the major rea-
sons reported as obstacles to growth for all firms; however other reasons such as taxation
and corruption are also important. Large firms are less likely to elect these as the major
obstacle to growth compared to smaller firms. However we fail to find a statistically sig-
nificant difference across firm size. Questionnaire difficulties again may explain this find-
ing as the question on financial access was narrowly phrased to focus on difficulties of
collateral provision.
Appendix
Table A.1. GDP, Population, and Branch Density per State
GDP per GDP No. Branch
capita (R$) (millions, R$) Population Branches per capita
North 4,312 57,027 13,225,186 623 21,228
Rond�nia 4,321 6,083 1,407,776 85 16,562
Acre 3,351 1,921 573,262 31 18,492
Amazonas* 7,169 20,736 2,892,454 132 21,913
Roraima 3,623 1,219 336,461 17 19,792
Par� 3,435 21,748 6,331,295 261 24,258
Amap� 4,523 2,253 498,121 19 26,217
Tocantins 2,590 3,067 1,184,170 78 15,182
Northeast 3,255 157,302 48,326,267 2383 20,280
Maranh�o* 1,796 10,293 5,731,069 247 23,203
Cear�* 2,858 21,581 7,551,085 348 21,699
Para�ba* 2,959 10,272 3,471,443 151 22,990
Bahia* 3,957 52,249 13,204,195 710 18,597
Piau� 1,941 5,575 2,872,231 108 26,595
Rio Grande do Norte 3,490 9,834 2,817,765 130 21,675
Pernambuco 3,962 31,725 8,007,320 425 18,841
Alagoas 2,649 7,569 2,857,305 117 24,421
Sergipe 4,514 8,204 1,817,457 147 12,364
Southeast 9,316 684,730 73,500,429 9263 7,935
Minas Gerais* 6,261 113,530 18,132,886 1828 9,920
Esp�rito Santo 7,148 22,538 3,153,050 315 10,010
S�o Paulo* 10,642 400,629 37,646,025 5484 6,865
Rio de Janeiro* 10,160 148,033 14,570,177 1638 8,895
South 8,387 213,389 25,442,828 3446 7,383
Santa Catarina* 8,541 46,535 5,448,425 811 6,718
Rio Grande do Sul* 9,129 94,084 10,306,058 1379 7,474
Paran�* 7,511 72,770 9,688,457 1256 7,714
Center-West 7,260 86,288 11,885,399 1283 9,264
Mato Grosso do Sul 6,505 13,736 2,111,606 220 9,598
Mato Grosso* 5,650 14,453 2,558,053 226 11,319
Goi�s* 4,898 25,048 5,113,924 545 9,383
Distrito Federal 15,725 33,051 2,101,812 292 7,198
Brazil 6,954 1,198,736 172,380,788 16998 10,141
Source: IBGE and Central Bank of Brazil.
23
24 World Bank Working Paper
Table A.2. The Dataset (Size, Region, Industry, Manager's Education, Sales Growth)
Size (No. of employees) Size (No. of employees)
Micro Small Medium Large Micro Small Medium Large
1�19 20�99 100�499 >500 Total 1�19 20�99 100�499 >500
Regions
North 8.3 66.7 20.8 4.2 100 14.8 16.0 11.2 12.0
Northeast 20.6 58.0 17.6 3.8 100 12.7 6.4 5.3 5.3
Southeast 21.1 53.4 21.2 4.3 100 45.5 44.3 40.2 41.3
South 16 49.6 28.9 5.5 100 26.4 31.5 42.0 40.0
Center-West 34.7 45.5 16.5 3.3 100 0.6 1.9 1.3 1.3
Total 100 100 100 100 100 100 100 100 100
Industry
Food Processing 12.6 35.4 39.4 12.6 100 7.0 4.8 8.2 14.7
Textiles 21.7 38.7 29.2 10.4 100 30.0 30.5 19.7 9.3
Garments 22.4 59.3 16.7 1.6 100 8.5 11.3 10.9 9.3
Shoes & Leather 16.2 56.1 23.7 4.0 100 3.9 5.9 3.7 8.0
Chemicals 15.5 60.7 16.7 7.1 100 13.6 9.4 13.0 10.7
Machinery 24.6 44.3 26.8 4.4 100 3.3 6.3 2.9 4.0
Electronics 13.9 68.4 13.9 3.8 100 4.8 6.7 11.4 17.3
Auto-parts 12.3 44.6 33.1 10.0 100 23.6 19.9 16.5 5.3
Furniture 24.7 54.3 19.7 1.3 100 0.0 0.0 0.3 0.0
Total 100 100 100 100 100 100 100 100 100
Manager's Education
Post-Graduate 10.9 42.0 32.9 14.2 100 10.9 16.2 29.0 62.7
Graduated Univ. 16.8 53.8 25.8 3.6 100 25.5 31.3 34.3 24.0
Incomplete Univ. 15.7 60.6 21.3 2.4 100 11.8 17.6 14.1 8.0
Vocational Training 28.1 55.7 15.1 1.1 100 15.8 12.0 7.4 2.7
Sec. School 23.4 55.1 20.9 0.6 100 11.2 10.1 8.8 1.3
Incomplete
Sec. School 30.6 58.1 11.3 0.0 100 5.8 4.2 1.9 0.0
Primary School 38.9 45.3 15.8 0.0 100 11.2 5.0 4.0 0.0
Incomplete
Primary School 43.3 51.7 3.3 1.7 100 7.9 3.6 0.5 1.3
Total 100 100 100 100 100 100 100 100 100
Sales Growth
Sales Increased 17.4 51.1 25.6 5.9 100 55.7 63.1 72.0 83.6
Sales Decreased 28.2 52.3 17.4 2.1 100 33.6 24.2 18.3 11.0
Sales Unchanged 19.2 58.8 19.8 2.2 100 10.7 12.7 9.7 5.5
Total 100 100 100 100 100 100 100 100 100
Source: World Bank, Investment Climate Survey, 2003.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 25
Table A.3. Definition and Construction of Variables
Basic variable Measures
Size Size dummies according to the number of employees: micro: 0�19;
small: 20�99; medium: 100�499; and large more than 499. Size is
also classified according to quintiles and deciles of the sales and
numbers of employees.
Performance/
Risk-adjusted performance
Rate of sales growth Percentage of sales growth (%)
Leverage Liabilities/capital (%)
Turnover Sales/assets (%)
Industry Nine sectors using CNAE classification: food processing, textiles,
garments, shoes and leather products, chemicals, machinery,
electronics, auto-parts, furniture. We also weigh the industrial
dummies by capital factor intensity.1
Region Five national regions: North, Northeast, South, Southeast, and
Center-West. We also weight those dummies by regional income
per capita and by branch density.
Ownership Three types of ownership: state, private domestic and private
foreign. 2
Education Eight levels of education: post graduate degree, university degree,
incomplete university degree, vocational training after secondary
school, complete secondary school, incomplete secondary school,
complete primary school, and incomplete primary school.
Relation with the banks/
credit worthiness proxies
Unique Bank Relationship (=1) if the firm does business with only one bank, (=0) if the firm
does business with more than one bank
Bank Ownership Three types of bank ownership: public, private domestic and
private foreign.
Overdraft or line of credit (=1) if the firm has an overdraft or line of credit, (=0) if the firm has
not an overdraft or line of credit
Collateral (=1) if the firm owns the buildings or land, (=0) otherwise
Competition, Credibility,
Capacity Use and
Innovation
New firm (=1) if the firm is below the age of two years old, (=0) ) if the firm
is above the age of five years old
Exports (=1) if the firm exports more than 10% of its production, (=0) if the
firm exports less than 10%
Credibility proxies
External auditor Annual financial statements are reviewed by an external auditor
Belongs to an (=1) if the firm belongs to an economic group, (=0) if the firm does
economic group not belong to an economic group
Status (=1) if the firm is a SA, (=0) if the firm is not a SA
26 World Bank Working Paper
Table A.3. Definition and Construction of Variables(Continued)
Basic variables Measures
Belongs to a producer (=1) if the firm belongs to a producer or trade association, (=0) if
or trade association the firm does not belong to a producer or trade association
Innovation and
Capacity Utilization
Computers use Workforce that regularly use computer in their jobs (%)
Capacity utilization 2002�2000 Average capacity utilization (%)
1. Factor intensity: capital (machinery and equipment) cost/labor costs.
2. The definitions of ownership follows the World Bank classification: (i) Private Domestic--firm with
a private domestic capital share that is (1) higher than the government capital share and higher
than the foreign capital share, and (2) the government share, and the foreign share if applicable,
is less than 10%; (ii) Private Foreign--firm with a foreign capital share that is (1) 10% or more and
(2) higher than the government capital share; and (iii) State--firm with a government capital share
that is (1) 10% or more and (2) higher than the foreign capital share (for the purpose of this classifi-
cation the private domestic capital share is irrelevant when the government capital share is 10% or
more).
Source: World Bank, Investment Climate Survey--Brazil, 2003.
28 World Bank Working Paper
Table A.4. Source of Finance--Working Capital
Size Region Education
university
graduate
Micro Small Medium Large North Northeast Center-west Southeast South Post. Graduate Incomplete
Internal funds 44.2 43.3 44.8 41.2 52.9� 45.9� 55.0� 46.3� 36.6� 40.7 44.8 42.0
Bank finance
Local1 21.9 26.0 27.3 27.7 18.5� 20.6� 15.1� 26.9� 28.5� 25.7 26.3 26.5
Local private 10.8 12.7 12.6 8.5
Local public2 11.9* 15.2* 17.6* 25.2*
Of which 0.8� 1.9� 2.9� 6.0�
government funds
Foreign Operations 0.8� 0.9� 1.7� 4.9� 1.0 0.9 1.0 1.4 1.1 2.4* 1.2* 0.8*
finance
Trade credit 14.2 16.3 13.7 14.2 9.4� 16.0� 13.2� 12.5� 19.1� 15.1 14.3 15.4
Leasing 0.5 0.9 0.8 0.3 0.8 0.2 0.9 0.7 1.1 0.5 0.5 1.1
Informal sources 10.5� 5.5� 1.8� 0.2� 2.3 7.4 5.8 4.9 5.3 3.7 5.1 6.1
Government funds 4.2 2.2 3.0 1.9 2.4 2.6 2.5 2.0
Equity finance 2.7 2.7 4.7 1.8 9.8 2.5 4.4 2.8 3.3 5.9� 3.0� 3.0�
Credit card finance 0.8 1.0 0.3 0.0 1.0 1.1 0.3 0.7 0.8 0.4 1.0 0.7
Others 3.6 1.5 2.0 3.7 0.0 3.3 1.3 1.9 1.7 3.0 1.3 2.3
Total 100 100 100 100 100 100 100 100 100 100 100 100
No. of firms 328 860 373 72 24 234 119 712 544 328 498 249
1. For firm size we disaggregate local finance into local private and local public. This disaggregration
does not derive directly from the questionnaire. Local commercial bank finance is disaggregated
into local private and local public finance according to the main bank the firm does business with.
2. Government funds are included in the local public bank finance category.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: World Bank, Investment Climate Survey--Brazil, 2003.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 29
Industry Sales growth
school
training secondary prim.
school shoes
school &
processing
Vocational Secondary Incomplete Primary Incomplete Food Textiles Garments Leather Chemicals Machinery Electronics Auto-parts Furniture Increased Decreased Unchanged
42.9 47.5 45.4 48.0 40.9 44.4 39.8 47.6 44.0 50.1 41.4 39.1 43.1 40.0 47.2� 35.7�42.1�
25.4 25.2 21.7 25.5 21.4 22.5 27.4 22.3 28.6 18.8 26.9 28.1 31.4 26.9 24.7 28.3 25.0
1.0* 0.3* 1.3* 0.8* 0.7* 2.4* 2.4* 0.7* 0.6* 3.0* 1.6* 0.7* 1.0 1.0* 1.6* 0.6* 0.7*
14.5 15.1 20.0 14.2 22.3 14.8 13.1 14.7 14.4 16.2 17.5 15.9 10.2 17.7 14.3 16.5 16.7
1.3 0.5 0.5 2.1 0.5 1.6 0.5 0.7 0.6 0.4 1.3 0.7 0.7 0.6 0.8 0.6 1.0
7.7 5.0 8.0 5.5 5.4 5.6 0.9� 6.9� 6.8� 2.2� 4.7 2.5� 2.0� 7.1� 3.9� 8.6� 5.9�
3.3 1.4 1.0 1.1 0.9 2.4 3.1 2.4 0.9 2.3 2.8 1.4 2.0 2.4 2.5* 2.2* 0.7*
1.9� 1.2� 1.8� 1.1� 3.5� 3.9� 7.0� 1.9� 2.5� 2.5� 1.2� 7.8� 7.3� 2.1� 3.0 2.9 4.0
1.1 0.7 0.2 0.4 0.6 1.0 0.5 1.4 0.3 1.2 0.3 0.7 0.2 0.4 0.6* 1.3* 0.3*
1.0 3.0 0.3 1.3 3.8 1.4 5.0 1.4 1.2 3.5 2.4 3.2 2.0 1.6 4.0 3.2 3.4
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
183 157 61 95 60 125 105 441 171 83 182 79 129 315 1038 390 181
30 World Bank Working Paper
Table A.5. Source of Finance: New Investments
Size Region Education
university
graduate
Micro Small Medium Large North Northeast Center-west Southeast South Post Graduate Incomplete
Internal funds 58.7 57.8 54.0 41.0 54.8� 61.0� 61.7� 59.6� 49.2� 53.4 56.7 57.9
Bank finance
Local1 15.7� 10.3� 9.7� 11.5� 17.0� 12.1 11.8 14.9
Local private 5.7 6.9 5.4 1.4
Local public2 10.4� 14.1� 19.1� 34.5�
Of which 4.5� 6.5� 12.5� 25.3�
government funds
Foreign Operations 0.0� 0.8� 2.6� 3.2� 0.0* 0.3* 0.3* 1.6* 1.3* 2.5* 1.5* 0.5*
finance
Trade credit 11.9* 8.6* 6.6* 9.2* 8.3 8.5 8.7 9.3 8.2 6.9 8.3 7.4
Leasing 2.2 3.1 3.5 5.0 1.7 1.2 0.4 4.1 3.4 2.9 3.1 3.3
Informal sources 4.4� 2.4� 0.4� 0.0� 2.5 3.5 2.9 2.2 1.8 1.3 2.1 3.1
Government funds 7.0� 7.0� 8.6� 5.9� 12.1� 10.8 9.5 7.6
Equity finance 3.5 3.8 6.0 4.0 9.1 4.0 6.1 3.7 4.5 8.4� 4.6� 1.7�
Credit card finance 0.5 0.2 0.2 0.0 0.9 0.4 0.1 0.2 0.3 0.1 0.5 0.3
Others 2.7 2.3 2.2 1.7 0.0 3.8 1.4 2.0 2.2 1.7 1.9 3.3
Total 100 100 100 100 100 100 100 100 100 100 100 100
No. of firms 247 716 324 64 23 178 110 569 471 276 429 200
1. For firm size we disaggregate local finance into local private and local public. This disaggregration
does not derive directly from the questionnaire. Local commercial bank finance is disaggregated
into local private and local public finance according to the main bank the firm does business with.
2. Government funds are included in the local public bank finance category.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: World Bank, Investment Climate Survey--Brazil, 2003.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 31
Industry Sales growth
school
training secondary prim.
school shoes
school &
processing
Vocational Secondary Incomplete Primary Incomplete Food Textiles Garments Leather Chemicals Machinery Electronics Auto-parts Furniture Increased Decreased Unchanged
54.3 57.9 58.3 61.2 53.9 52.8*46.2* 61.3* 59.2* 52.3 53.3* 56.3*51.1*57.3* 57.9*52.5* 52.4*
14.5 14.1 14.3 14.7 14.5 13.7 11.7 13.8 13.7 12.2 14.9 13.6 14.5 10.4 13.1 12.7 15.3
0.6* 0.8* 0.0* 0.0* 0.0* 1.1� 6.1� 0.6� 0.3� 1.3� 1.3� 1.5� 0.8� 1.0� 1.4 0.5 1.8
10.9 8.4 9.6 9.6 19.9 7.8* 16.2* 9.2* 8.0* 3.5* 8.5* 8.8* 4.6* 9.8* 7.1 13.0 10.0
5.0 2.2 3.9 1.5 2.8 3.8 1.4 1.7 2.4 2.3 5.3 1.4 5.2 4.4 3.2 2.0 4.7
3.5 2.8 1.5 1.9 1.4 1.7 0 2.8 2.7 2 2.4 1.2 1.8 2.9 1.6 3.0� 4.0�
5.6 8.5 6.2 7.6 1.9 12.6 10.3 5.7 6.9 11.9 10.0 5.1 9.5 9.5 9.3* 7.7* 4.5*
3.9� 1.9� 2.2� 1.5� 2.8� 5.8� 5.5� 2.0� 3.8� 7.5� 3.4� 10.1� 9.6� 2.6� 3.9 5.8 4.1
0.1 0.2 0.0 0.1 0.0 0.1 0.0 0.4 0.0 1.4 0.4 0.4 0.0 0.1 0.2 0.5 0.1
1.7 3.2 4.1 1.9 2.8 0.6 2.4 2.6 3.1 5.6 0.7 1.5 2.8 2.1 2.2 2.1 3.1
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
145 126 46 74 54 110 83 365 143 70 152 68 109 249 906 286 138
32 World Bank Working Paper
Table A.6. Overdrafts, Credit Lines and Trade Credit
Size Region
university
graduate
Micro Small Medium Large North Northeast Center-West Southeast South Post Graduate Incomplete
No. Inst. Firms 2.1 3.0 4.8 8.2 3.0* 3.0* 2.9* 3.4* 3.8* 4.5* 3.7* 3.1*
business with
Overdraft or credit line
Firms with 60.2� 75.7� 82.9� 80.8� 83.3� 61.9� 66.1� 75.7� 79.7� 81.1� 76.5� 75.8�
overdraft facility
or line of credit (%)
Of which used (%) 44.4 46.6 46.9 51.9 36.15 49.5 44.3 45.5 47.8 46.1 45.4 49.4
Average interest rate 5.1� 4.9� 3.9� 3.1� 5.6 4.8 4.3 4.7 4.3 3.7� 4.2� 4.7�
(monthly rate, %)
Trade credit
Firms that use 76.9 82.1 82.6 84.9 70.8 79.7 77.7 80.3 84.6 83.2 83.1 79.5
supplier credit
Inputs purchased 79.5* 82.2* 85.2* 82* 85.9� 73.1� 73.2� 84.3� 83.5� 84.8* 84.3* 79.8*
on credit
Sales paid before 6.8 5.3 6.1 8.8 1.4 6.8 7.4 5.8 5.8 6.6 6.6 5.9
delivery
Total 100 100 100 100 100 100 100 100 100 100 100 100
No. of firms 330 861 376 75 23 178 110 569 471 276 429 200
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 33
Managers' education Industry Sales growth
school
training secondary prim.
school shoes
school &
processing
Vocational Secondary Incomplete Primary Incomplete Food Textiles Garments Leather Chemicals Machinery Electronics Auto-parts Furniture Increased Decreased Unchanged
3.0* 2.7* 2.5* 2.5* 2.3* 4.1 4.5 2.7 3.0 5.5 3.9 2.9 4.1 3.1 3.7 2.9 3.1
71.4� 69.6� 67.7� 58.9� 66.6� 74.6 75.2 70.1 72.3 73.8 78.0 79.7 82.9 74.5 78.2� 68.7� 67.6�
45.9 49.2 41.9 45.8 50.3 47.456.2 43.1 50.1 40.0 45.8 35.6 48.7 49.6 45.7*54.5* 56.9*
5.2� 5.9 5.5 5.7 4.8 3.6 4.2 5.1 4.5 3.5 4.9 4.5 4.2 4.8 4.4 5 4.6
82.0 81.0 77.4 80.0 68.3 76.2 74.3 80.7 82.1 72.6 83.5 83.5 85.3 84.8 81.1 83.8 76.9
77.3*80.9* 78.3* 83.1*85.3* 77.5� 79.9� 79.8� 86.1� 77.2� 84.1� 90.9� 90.2� 81.0� 82.7 82.3 80.3
5.3 4.6 8.1 3.0 5.4 3.3� 5.5� 3.5� 1.9� 9.2� 12.6� 3.1� 3.0� 10.2� 6.2 6.4 4.3
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
145 126 46 74 54 110 83 365 143 70 152 68 109 249 906 286 138
34
World
Bank
Table A.7. Firm Size and Number of Banks Firms Do Business with Working
Size (no. employees)--frequencies Size (no. of employees)--%
No. of Micro Small Medium Large Micro Small Medium Large Paper
banks No. firms 0�19 20�99 100�499 >500 0�19 20�99 100�499 >500 Total
0 10 5 5 0 0 50 50 0 0 100
1 273 112 133 25 3 41.0 48.7 9.2 1.1 100
2 464 114 291 55 0 24.8 63.3 12.0 0.0 100
3 351 63 217 64 7 17.9 61.8 18.2 2.0 100
4 197 21 97 71 8 10.7 49.2 36.0 4.1 100
5 119 7 58 47 7 5.9 48.7 39.5 5.9 100
6 69 4 27 35 3 5.8 39.1 50.7 4.3 100
7 33 1 6 22 4 3.0 18.2 66.7 12.1 100
8 0 0 9 21 6 0.0 25.0 58.3 16.7 100
9 8 0 2 4 2 0.0 25.0 50.0 25.0 100
10 35 0 5 14 16 0.0 14.3 40.0 45.7 100
>10 38 0 9 16 13 0.0 23.7 42.1 34.2 100
Total 1597 327 859 374 69 20.1 52.7 23.0 4.2 100
Average 2.1 3.0 4.8 8.2
Source: World Bank, Investment Climate Survey--Brazil, 2003.
36 World Bank Working Paper
Table A.8. Size, Region, Education, Industry, and Sales Growth Effects on Access to
Credit and Credit Constraints
Size Region Education
university
graduate
Micro Small Medium Large North Northeast Center-west Southeast South Post Graduate Incomplete
Total no. of firms 329 860 374 73 24 236 121 711 544 328 499 249
Have a bank loan 27.1� 31.9� 43.9� 58.9� 16.7� 33.1� 28.9� 32.1� 41.4� 38.7 36.7 30.9
(% of total
no. of firms)
Do not have a 72.9� 68.1� 56.1� 41.1� 83.3� 66.9� 71.1� 67.9� 58.6� 61.3 63.3 69.1
bank loan (% of
total no. of firms)
Total (% of total 100 100 100 100 100 100 100 100 100 100 100 100
no. of firms)
Do not have a loan
Rejected (% of do 12.9 10.2 8.6 6.7 10.0 12.7 10.5 9.1 11.3 8.5 8.5 13.4
not have a loan)
Did not apply (% of 87.1 89.8 91.4 93.3 90.0 87.3 89.5 90.9 88.7 91.5 91.5 86.6
do not have a loan)
Total (% of do not 100 100 100 100 100 100 100 100 100 100 100 100
have a loan)
Did not apply
No need (% of did 39.2* 44.5* 51* 53.6* 38.9 42.8 51.9 44.4 45.9 47.8 44.8 47.6
not apply)
Other reasons2 60.8* 55.5* 49* 46.4* 61.1 57.2 48.1 55.6 54.1 52.2 55.2 52.4
(% of did not apply)
Total (% of did 100 100 100 100 100 100 100 100 100 100 100 100
not apply)
Total of firms 47.7 40.7 29.7 20.5 54.2 41.9 38.0 40.4 34.6 34.5 37.3 40.2
constrained (% of
total no. firms)
Application was 9.4 7.0 4.8 2.7 8.3 8.5 7.4 6.2 6.6 5.2 5.4 9.2
rejected (% of total
no. firms)
Did not apply 38.3 33.7 24.9 17.8 45.8 33.5 30.6 34.2 27.9 29.3 31.9 30.9
(% of total no. firms)
1. For firm size we disaggregate local finance into local private and local public. This disaggregration
does not derive directly from the questionnaire. Local commercial bank finance is disaggregated
into local private and local public finance according to the main bank the firm does business with.
2. Government funds are included in the local public bank finance category.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: World Bank, Investment Climate Survey--Brazil, 2003.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 37
Industry Sales growth
school
training secondary prim.
school shoes
school &
processing
Vocational Secondary Incomplete Primary Incomplete Food Textiles Garments Leather Chemicals Machinery Electronics Auto-parts Furniture Increased Decreased Unchanged
184 158 62 95 60 126 105 441 173 84 182 79 129 315 1041 390 182
36.4 27.2 32.3 33.7 35 37.3 40 35.4 31.8 35.7 33.5 30.4 38.0 33.3 34.9 35.9 35.7
63.6 72.9 67.7 66.3 65 62.7 60 64.6 68.2 64.3 66.5 69.6 62.0 66.7 65.2 64.1 64.3
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
11.1 12.2 11.9 11.1 12.8 13.9* 4.8*11.9* 8.5* 7.4* 7.4* 3.6* 7.5* 15.2* 9.1� 15.6� 6.8�
88.9 87.8 88.1 88.9 87.8 86.1* 95.2*88.1* 91.5*92.6* 92.6* 96.4* 92.5* 84.8* 90.9� 84.4� 93.2�
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
39.8 46.5 40.5 41.1 44.1 35.3 43.3 45.8 60.8 44.0 48.2 50.0 34.3 40.0 46.1 44.8 43.2
60.2 53.5 59.5 58.9 55.9 64.7 56.7*54.2 39.256 51.850 65.7 59.9 53.9 55.2 56.8
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
40.8 43.0 43.5 42.1 40.0 43.7 35.2 38.5 30.1 38.1 36.8 35.4 41.9 43.8 37.8 39.7 37.9
7.1 8.9 8.1 7.4 8.3 8.7 2.9 7.7 5.8 4.8 4.9 2.5 4.7 10.2 6.0 10.0 4.4
33.7 34.2 35.5 34.7 31.7 34.9 32.4 30.8 24.3 33.3 31.9 32.0 37.2 33.7 31.8 29.7 33.5
38 World Bank Working Paper
Table A.9. Reasons for Not Applying for a Bank Loan and Reasons for Bank Loan Rejection
Size Region
university
Reasons for not
graduate
applying for a
bank loan Micro Small Medium Large North Northeast Center-west Southeast South Post Graduate Incomplete
Do not need loans 39.4* 44.7* 51.3* 53.6* 38.9 42.8 51.9 44.4 45.9 47.8 44.8 47.6
Applications 13.5* 9.4* 7.3* 0.0* 0.0 11.6 11.7 9.2 9.3 4.89 8.3 8.2
procedures
Collateral 7.7* 9.2* 4.2* 3.6* 5.6 8.7 9.1 7.3 7.5 6.52 7.3 10.9
requirements
Interest rates 36.5 33.2 33.0 32.1 50.0 36.2 23.4 35.0 32.7 35.9 35.1 30.6
are too high
Corruption in 1.0 0.6 0.0 3.6 0.0 0.0 1.3 0.5 1.1 0.5 0.35 0
the allocation
of bank credit
Did not expected 1.4 0.8 0.5 0.0 0.0 0.0 1.3 1.4 0.4 0.5 0.7 1.4
to be approved
Others 0.5* 2.3* 3.7* 7.1* 5.6 0.7 1.3 2.3 3.2 3.8 3.5 1.4
Total (%) 100 100 100 100 100 100 100 100 100 100 100 100
Total no. of firms 208 524 191 28 18 138 77 437 281 184 288 147
Reasons for
rejection
Lack of collateral 41.9 41.7 22.2 50.0 50.0 30.0 33.3 14.7 41.7 35.3 51.9 43.5
Incompleteness of 19.4 16.7 11.1 0.0 0.0 15.0 0.0 14.7 16.7 11.8* 3.7* 21.7*
the application
Lack of feasibility 3.2 10.0 22.2 0.0 0.0 10.0 0.0 17.6 8.3 5.9 7.4 13.0
of the project
Poor credit history 25.8 21.7 38.9 50.0 50.0 35.0 66.7 44.1 16.7 41.2 29.6 21.7
Others 9.7 10.0 5.6 0.0 0.0 10.0 0.0 8.8 16.7 5.9 7.4 0.0
Total (%) 100 100 100 100 100 100 100 100 100 100 100 100
Total no. of firms 2 20 3 200 36 17 27 23
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: World Bank, Investment Climate Survey--Brazil, 2003.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 39
Managers' education Industry Sales growth
school
training secondary prim.
school shoes
school &
Processing
Vocational Secondary Incomplete Primary Incomplete Food Textiles Garments Leather Chemicals Machinery Electronics Auto-parts Furniture Increased Decreased Unchanged
39.8 46.5 40.5 41.1 44.1 35.3 43.3 45.8 60.7 44.0 48.2 50.0 34.2 40.1 46.1 46.1 46.1
13.415.8 8.1 12.5 17.7 10.3 6.7 11.2 10.3 8.0 8.9 0.0 6.8 12.4 9.1 9.1 9.1
5.8 8.9 2.7 8.9 8.8 8.8 8.3 7.6 7.5 8.0 7.1 7.7 13.7 5.1 7.2 7.2 7.2
36.9 24.6 45.5 35.7 29.4 38.2 40.0 32.7 18.7 36.0 27.7 36.5 42.5 40.1 34.0 34.0 34.0
1.9 1.0 2.7 1.8 0 1.5 0.0 0.8 0.0 0.0 0.0 1.9 1.4 0.6 0.7 0.7 0.7
1.9 0 0 0 1.5 0.0 0.4 0.9 0.0 1.8 1.9 1.4 0.6 0.5 0.5 0.5
0 2.97 0 0 4.4 1.7 1.6 1.9 4.0 6.3 1.9 0.0 1.1 2.4 2.4 2.4
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
103 101 37 56 34 68 60 251 107 50 112 52 73 177 614 210 108
38.5 21.4 20.0 28.6 40.0 27.3 33.3 38.2 40.0 25.0 55.6 50.0 66.7 34.4 37.1 46.2 25.0
7.7*35.7* 0.0* 28.6* 40.0* 9.1 0.0 26.5 10.0 25.0 11.1 0.0 0.0 15.6 21.0 10.3 12.5
7.7 21.4 0.0 14.3 0.0 0.0 33.3 5.9 10.0 25.0 11.1 0.0 0.0 15.6 8.1 10.3 25.0
23.1 21.4 60.0 0.0 0.0 36.4 33.3 23.5 30.0 0.0 22.2 50.0 33.3 25.0 22.6 28.2 25.0
23.1 0.0 20.0 28.6 20.0 27.3 0.0 5.9 10.0 25.0 0.0 0.0 0.0 9.4 11.3 5.1 12.5
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
13 14 5 7 5 11 3 34 10 4 9 2 6 32 62 39 8
40 World Bank Working Paper
Table A.10. The Importance of Collateral and Shares of Collateral
Size Region
university
graduate
Micro Small Medium Large North Northeast Center-west Southeast South Post Graduate Incomplete
Loans that 42.7� 63.1� 82.9 81.4� 50.0* 70.5* 71.4* 60.5* 72.0* 78.7� 66.1� 70.1�
required collateral
Collateral as % 121.6 131.6 119.7 117.9 110 139.9 120.3 128.5 118.0 125.4 117.3 115.8
of the loan value
Share of collateral
Buildings, land 33.7 25.9 30.0 18.8 50.0 35.5 25.0 23.0 28.7 28.6 26.6 29.2
Machinery 24.5 23.1 21.0 20.0 0.0 14.3 15.8 21.5 26.8 19.4 19.2 28.1
Intangible assets 11.2 15.6 16.6 34.8 0.0 10.1 26.0 23.5 13.3 20.5 15.9 13.9
Personal assets 14.2 20.9 10.2 7.0 50.0 21.8 16.3 16.0 11.6 10.8 17.2 16.4
Other 16.4 14.6 22.2 19.4 0.0 18.4 17.0 16.1 19.7 20.7 21.1 12.4
Total (%) 100 100 100 100 100 100 100 100 100 100 100 100
Source: World Bank, Investment Climate Survey--Brazil, 2003.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 41
Managers' education Industry Sales growth
school
training secondary prim.
school shoes
school &
processing
Vocational Secondary Incomplete Primary Incomplete Food Textiles Garments Leather Chemicals Machinery Electronics Auto-parts Furniture Increased Decreased Unchanged
56.7� 53.5� 80.0� 62.5� 47.6� 78.7� 73.8� 54.5� 58.2� 73.3� 68.9� 58.3� 85.7� 73.3� 69.4 64.3 58.5
157.0138.8131.8 123.8 108.6 135.8 122.5130.7 107.7115.9 129.6 141.1 111.9125.7 123.2 121.5148.6
32.0 21.1 6.3 33.5 36.1 21.8 24.4 31.1 27.5 26.3 31.1 31.4 19.6 29.4 23.6 33.7 38.4
17.9 41.1 28.1 23.0 16.0 20.6 15.8 22.6 21.6 12.2 27.4 14.3 32.7 21.0 24.9 16.3 17.2
18.9 15.2 12.5 16.0 28.9 19.7 28.1 13.2 20.3 19.5 13.5 14.3 21.5 14.7 17.2 14.5 23.2
13.4 15.2 25.0 15.0 19.0 13.4 20.5 16.4 14.1 4.9 12.5 25.7 11.7 17.3 16.2 14.9 9.3
17.8 7.4 28.1 12.5 0.0 24.5 11.3 16.7 16.6 37.0 15.6 14.3 14.4 17.7 18.0 20.7 11.8
100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
42 World Bank Working Paper
Table A.11. Regression Results--Firm Characteristics, Performance and the
Probability of Having a Loan
Having a loan Having a loan1 Having a loan2
Size
Small firms 0.090 0.087 0.087
(2.20) (2.13) (2.13)
Medium firms 0.222� 0.219� 0.219�
(4.56) (4.50) (4.50)
Large firms 0.338� 0.336� 0.336�
(4.28) (4.25) (4.25)
Industry
Food processing 0.110* -0.002 -0.002
(1.75) (0.21) (0.21)
Textile 0.098 -0.029* -0.029*
(1.20) (1.78) (1.78)
Shoes and Leather products 0.096 -0.014 -0.014
(1.28) (0.26) (0.26)
Chemicals 0.034 -0.007 -0.007
(0.37) (1.02) (1.02)
Machinery 0.075 -0.008 -0.008
(1.03) (0.68) (0.68)
Electronics 0.099 -0.001 -0.001
(1.03) (0.16) (0.16)
Auto-parts 0.016 -0.009 -0.009
(0.21) (1.57) (1.57)
Furniture 0.053 -0.025 -0.025
(0.82) (1.32) (1.32)
Region
South 0.068* 0.026* 0.180*
(1.89) (1.87) (1.87)
Center-West -0.006 -0.003 -0.092
(0.09) (0.11) (0.11)
North -0.186 -0.150 -2.890
(1.26) (1.26) (1.26)
Northeast -0.005 -0.007 -0.026
(0.11) (0.14) (0.14)
Firm Ownership
State firms 0.212 0.213 0.213
(0.91) (0.92) (0.92)
Foreign firms -0.031 -0.031 -0.031
(0.38) (0.38) (0.38)
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 43
Table A.11. Regression Results--Firm Characteristics, Performance
and the Probability of Having a Loan (Continued)
Having a loan Having a loan1 Having a loan2
Education
University degree 0.005 0.006 0.006
(0.10) (0.13) (0.13)
Incomplete university -0.056 -0.057 -0.057
(1.03) (1.03) (1.03)
Vocational training after secondary school 0.003 0.002 0.002
(0.05) (0.04) (0.04)
Secondary school -0.103 -0.104 -0.104
(1.62) (1.64) (1.64)
Incomplete secondary school -0.053 -0.054 -0.054
(0.61) (0.62) (0.62)
Primary School -0.002 -0.003 -0.003
(0.03) (0.05) (0.05)
Incomplete primary school 0.010 0.009 0.009
(0.12) (0.10) (0.10)
Performance
Sales growth* -0.030 -0.030 -0.030
(0.96) (0.96) (0.96)
Observations 1116 1117 1117
* Sales growth of 2001. Note: We exclude from the analysis firms that do not need a loan.
Control dummies : Micro firms, Garments industry, Southeast, Post Graduate, and decreasing rate and
unchanged sales growth.
1. Regional dummies are weighted by regional income. Southeast (the richest region) is the con-
trol dummy. Industry dummies are weighted by capital intensity ratio (capital costs/labor
costs). Garments industry (the lowest capital intensity ratio) is the control dummy.
2. Regional dummies are weighted by branch density. Southeast (the region with the largest branch
density) is the control dummy. Industry dummies are weighted by capital intensity ratio (capital
costs/labor costs). Garments industry is the control dummy.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
44 World Bank Working Paper
Table A.12. The Impact of Firm Size on the Likelihood of Having a Loan: Model 2
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)2
Size
Small 0.034 0.048
(0.63) (0.89)
Medium 0.154 0.172
(2.33) (2.59)
Large 0.248� 0.252
(2.42) (2.32)
Performance
Turnover (sales/assets) -0.000 -0.000
(0.88) (0.95)
Leverage 0.000 0.000
(0.20) (0.23)
Sales growth 0.001 0.001
(1.27) (1.37)
Firm characteristics
Exporter 0.002 0.000
(0.05) (0.01)
SA 0.034 0.043
(0.33) (0.43)
Group 0.038 0.035
(0.49) (0.45)
Capacity utilization 0.001 0.001
(0.86) (0.86)
New firm -0.077 -0.077
(1.37) (1.37)
% workforce that use computers 0.003 0.003
(2.44) (2.44)
External auditor 0.050 0.050
(0.90) (0.76)
Collateral 0.014 0.026
(0.32) (0.61)
Relation with banks
Overdraft 0.158�
(3.40)
Bank unique relationship -0.111 -0.138
(2.02) (2.45)
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 45
Table A.12. The Impact of Firm Size on the Likelihood of Having a Loan:
Model 2 (Continued)
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)2
Other control variables
Industry Yes Yes
Region Yes Yes
Firm ownership Yes Yes
Education Yes Yes
Observations 1088 1088
Wald chi2 74.55 64.96
1. This refers to firms who have demand for a loan and have received a loan. The universe here is lim-
ited to firms which demand for a loan. This models concerns to the second stage model of the two
step maximum likelihood probit: supply of credit model.
2. The dummy which controls for whether firms have an overdraft or not is excluded from this
specification.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: Authors' calculations based on World Bank, Investment Climate Survey--Brazil, 2003.
46 World Bank Working Paper
Table A.13. The Likelihood of Having a Loan According to Its Duration
Long term loans Long term loans Short term loans Short term loans
(Including (Excluding (Including (Excluding
overdrafts)1 overdrafts)1,2 overdrafts)1 overdrafts)1,2
Size
Small 0.066 0.080 -0.024 -0.034
(1.18) (1.36) (1.21) (1.15)
Medium 0.208� 0.229� -0.034 -0.030
(2.95) (3.15) (0.96) (0.89)
Large 0.248 0.246* 0.113 0.048
(1.95) (1.89) (0.77) (0.78)
Performance
Turnover
(sales/assets) -0.004 -0.005 0.000 0.000
(1.48) (1.61) (0.68) (0.61)
Leverage 0.000 0.000 0.000 0.001
(1.16) (0.24) (0.67) (0.71)
Sales growth 0.002* 0.002* 0.001 -0.000
(1.73) (1.79) (0.42) (0.38)
Firm characteristics
Exporter 0.032 -0.035 0.041 0.035
(0.62) (064) (1.26) (1.21)
SA -0.004 0.001 -0.005 0.043
(0.04) (0.01) (0.68) (0.75)
Group 0.077 0.076 -0.029 -0.014
(0.96) (0.91) (0.30) (0.35)
Capacity utilization 0.001 0.002 -0.000 0.001
(1.00) (1.34) (1.01)
New firm -0.034 -0.042 -0.039 -0.052
(0.62) (0.68) (0.69) (1.78)
% workforce that 0.005� 0.005� -0.001 -0.001
use computers (3.39) (3.37) (1.17 (1.52)
External auditor 0.054 0.050 0.012 -0.006
(0.97) (0.85) (0.34)
Collateral 0.028 0.043 -0.014 -0.022
(0.64) (0.93) (1.04) (0.96)
Relation with banks
Overdraft 0.141� 0.065
(2.88) (2.33)
Bank unique -0.177� -0.216� 0.029 0.032
relationship (3.04) (3.54) (0.73) (0.97)
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 47
Table A.13. The Likelihood of Having a Loan According to Its Duration (Continued)
Long term loans Long term loans Short term loans Short term loans
(Including (Excluding (Including (Excluding
overdrafts)1 overdrafts)1,2 overdrafts)1 overdrafts)1,2
Other control
variables
Industry Yes Yes Yes Yes
Region Yes Yes Yes Yes
Firm ownership Yes Yes Yes Yes
Education Yes Yes Yes Yes
Observations 1088 1088 1088 1088
Wald chi2 76.38 70.07 63.00 41.99
1. This refers to firms who have demand for a loan and have received a loan. The universe here is lim-
ited to firms which demand for a loan. This models concerns to the second stage model of the two
step maximum likelihood probit: supply of credit model.
2. The dummy which controls for whether firms have an overdraft or not is excluded from this
specification.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: Authors' calculations based on World Bank, Investment Climate Survey--Brazil, 2003.
48 World Bank Working Paper
Table A.14. The Impact of Bank Ownership on the Firm's Likelihood of Having a
Loan--Model 2--Sample Split by Bank Ownership
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)1,2
Public Private Public Private
bank bank bank bank
Size
Small 0.036 0.000 0.035 0.012
(0.43) (0.01) (0.43) (0.19)
Medium 0.171* 0.090 0.171* 0.107
(1.80) (1.17) (1.81) (1.31)
Large 0.298 0.167 0.294 0.208
(2.39) (1.20) (2.40) (1.39)
Performance
Turnover (sales/assets) 0.003 -0.002 0.002 -0.002
(0.55) (1.62) (0.55) (1.62)
Leverage 0.002 0.002 0.002 0.003
(0.60) (0.76) (0.59) (0.76)
Sales growth 0.001 0.002 0.001 0.002
(0.30) (2.31) (0.31) (2.07)
Firm characteristics
Exporter -0.001 -0.019 -0.002 -0.025
(0.01) (0.31) (0.03) (0.40)
SA 0.028 0.118 0.026 0.118
(0.20) (0.09) (0.19) (0.01)
Group -0.122 0.115 -0.122 0.094
(0.89) (1.67) (0.89) (1.11)
New firm 0.411 -0.694� 0.122 -0.016
(0.89) (1.47) (0.89) (2.45)
Capacity utilization 0.001 0.003 0.002
(0.30) (0.93) (0.32) (1.31)
% workforce that use computers 0.011 0.013 0.003 0.002
(1.26) (1.33) (1.27) (1.33)
External auditor -0.266 0.071 -0.078 0.078
(0.92) (1.24) (0.92) (1.23)
Collateral -0.002 0.001 -0.002 0.019
(0.04) (0.03) (0.03) (0.39)
Relation with banks
Overdraft -0.008 0.161�
(0.08) (3.19)
Bank unique relationship -0.070 -0.122* -0.070 -0.170
(0.83) (1.79) (0.84) (2.49)
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 49
Table A.14. The Impact of Bank Ownership on the Firm's Likelihood of Having a
Loan--Model 2--Sample Split by Bank Ownership (Continued)
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)1,2
Public Private Public Private
bank bank bank bank
Other control variables
Industry Yes Yes Yes Yes
Region Yes Yes Yes Yes
Firm ownership Yes Yes Yes Yes
Education Yes Yes Yes Yes
Observations 500 582 500 582
Wald chi2 33.93 57.89 33.92 49.67
1. This refers to firms who have demand for a loan and have received a loan. The universe here is lim-
ited to firms which demand for a loan. This models concerns to the second stage model of the two
step maximum likelihood probit: supply of credit model.
2. The dummy which controls for whether firms have an overdraft or not is excluded from this
specification.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: Authors' calculations based on World Bank, Investment Climate Survey--Brazil, 2003.
50 World Bank Working Paper
Table A.15. The Impact of Bank Ownership on the Firm's Likelihood of Having a
Loan--Model 2--Consolidated Sample
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)1,2
Size
Small 0.040 0.036
(0.56) (0.50)
Medium 0.211 0.211
(2.47) (2.44)
Large 0.274 0.263*
(1.97) (1.89)
Public Bank 0.137 0.137
(1.46) (1.42)
Small firm--Public Bank -0.009 -0.004
(0.09) (0.04)
Medium firm--Public Bank -0.098 -0.097
(0.81) (0.79)
Large firm--Public Bank -0.077 -0.072
(0.40) (0.37)
Performance
Turnover (sales/assets) -0.003 -0.003
(0.77) (0.76)
Leverage 0.000 0.000
(0.17) (0.22)
Sales growth 0.002 0.002
(1.52) (1.52)
Firm characteristics
Exporter -0.002 -0.007
(0.05) (0.13)
SA 0.026 0.024
(0.26) (0.23)
Group 0.038 0.038
(0.49) (0.45)
New firm -0.077 -0.080
(2.52) (1.35)
Capacity utilization 0.001 0.003
(1.07) (1.07)
% workforce that use computers 0.003 0.003
(2.21) (2.32)
External auditor 0.046 0.049
(0.83) (0.83)
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 51
Table A.15. The Impact of Bank Ownership on the Firm's Likelihood of Having a
Loan--Model 2--Consolidated Sample (Continued)
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)1,2
Collateral 0.012 0.011
(0.27) (0.27)
Relation with banks
Overdraft 0.374�
(2.87)
Bank unique relationship -0.144 -0.149
(2.52) (2.53)
Other Control variables:
Industry Yes Yes
Region Yes Yes
Firm ownership Yes Yes
Education Yes Yes
Observations 1084 1084
Wald chi2 33.93 33.93
1. This refers to firms who have demand for a loan and have received a loan. The universe here is lim-
ited to firms which demand for a loan. This models concerns to the second stage model of the two
step maximum likelihood probit: supply of credit model.
2. The dummy which controls for whether firms have an overdraft or not is excluded from this
specification.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: Authors' calculations based on World Bank, Investment Climate Survey--Brazil, 2003.
52 World Bank Working Paper
Table A.16. Probability of Having a Loan from a Public Bank or a BNDES Credit Line
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)1,2
Size
Small 0.040 0.036
(0.56) (0.50)
Size
Small firm 0.121 0.135
(2.15) (2.58)
Medium firm 0.243� 0.253�
(3.55) (4.04)
Large firm 0.270 0.261
(2.47) (2.61)
Performance
Turnover (sales/assets) -0.000 -0.000
(0.54) (0.39)
Leverage 0.001 0.001
(0.40) (0.52)
Sales growth 0.002* 0.008
(1.85) (0.20)
Firm characteristics
Exporter -0.002 0.014
(0.04) (0.29)
SA -0.023 -0.079
(0.23) (0.89)
Group -0.010 0.013
(0.13) (0.18)
Capacity utilization 0.000 0.001
(0.54) (1.07)
New firm 0.004 -0.018
(0.02)
% workforce that use computers 0.005� 0.005�
(3.17) (3.86)
External auditor 0.066 0.040
(1.17) (0.80)
Collateral 0.055 0.072*
(1.25) (1.83)
Relation with banks
Overdraft 0.166�
(3.18)
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 53
Table A.16. Probability of Having a Loan from a Public Bank or a BNDES
Credit Line (Continued)
Having a loan Having a loan
(Including overdrafts)1 (Excluding overdrafts)1,2
Bank unique relationship -0.059 -0.121
(0.99) (2.27)
Other Control Variables:
Industry Yes Yes
Region Yes Yes
Firm Ownership Yes Yes
Education Yes Yes
Observations 1088 1088
Wald chi2 68.58 75.21
1. This refers to firms who have demand for a loan and have received a loan. The universe here is lim-
ited to firms which demand for a loan. This models concerns to the second stage model of the two
step maximum likelihood probit: supply of credit model.
2. The dummy which controls for whether firms have an overdraft or not is excluded from this
specification.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Source: Authors' calculations based on World Bank, Investment Climate Survey--Brazil, 2003.
54
World
Bank
Working
Table A.17. Obstacles to Growth--Firm Size and Other Factors
Access to Cost of Tax Economic and regulatory Macroeconomic Paper
financing financing administration Tax rates policy uncertainty instability Corruption
Small 0.015 0.076 0.004 0.073 0.053 0.022 -0.094
(0.20) (0.95) (0.06) (0.98) (0.71) (0.30) (1.22)
Medium -0.027 -0.005 -0.082 0.011 -0.018 0.037 -0.238�
(0.31) (0.05) (0.98) (0.12) (0.22) (0.44) (2.73)
Large -0.203 -0.137 -0.259* -0.282 -0.037 0.175 -0.454�
(1.53) (1.01) (1.95) (2.12) (0.28) (1.23) (3.66)
Control Variables
Industry Yes Yes Yes Yes Yes Yes Yes
Region Yes Yes Yes Yes Yes Yes Yes
Firm ownership
Education
Observations 1616 1623 1636 1641 1639 1637 1634
Note: Micro firms is the control dummy.
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
�
Regression estimates based on World Bank, Investment Climate Survey--Brazil, 2003.
Table A.18. The Relative Importance of Obstacles to Growth and Firm Size
No Degree of obstacle Weighted Differences across
obstacle Low Medium High Very high Total average38 size test
Tax rates 3.2 2.3 10.1 33.4 51.1 100 32.71 (0.058)*
Cost of Financing (e.g. interest rates) 5.6 3.3 7.9 26.5 56.7 100 32.54 (0.056)*
Economic and regulatory policy uncertainty 2.9 4.5 16.8 32.8 43.1 100 30.89 (0.185)
Macroeconomic instability (inflation, exchange rate) 2.8 4.0 18.3 33.5 41.4 100 30.67 (0.592)
Enterprise
Corruption 10.4 7.9 14.5 20.1 47.1 100 28.56 (0.000)�
Tax administration 7.7 6.6 19.6 33.4 32.7 100 27.68 (0.421)
Access to Financing (e.g., collateral) 14.3 8.5 16.8 25.9 34.5 100 25.78 (0.352)
Size,
Labor regulations 10.4 9.6 23.2 29.9 27.0 100 25.37 (0.032)
Anti-competitive or informal practices 10.2 10.9 22.5 28.6 27.8 100 25.29 (0.346) Financing
Crime, theft and disorder 16.6 14.1 17.1 20.8 31.4 100 23.63 (0.000)�
Skills and education of available workers 12.9 14.9 32.5 28.9 10.7 100 20.94 (0.015)
Patterns,
Legal system/conflict resolution 21.3 17.5 28.4 19.9 13.0 100 18.60 (0.007)�
Customs Regulations 30.4 9.4 22.4 20.4 17.4 100 18.50 (0.000)�
Trade Regulations 30.8 12.2 22.2 19.4 15.4 100 17.64 (0.000)� and
Business Licensing and Operating permits 26.6 19.2 24.4 18.2 11.6 100 16.90 (0.192) Credit
Transportation 39.2 17.4 24.1 14.1 5.2 100 12.87 (0.465)
Electricity 45.8 15.8 18.2 13.7 6.6 100 11.97 (0.207) Constraints
Standards and Quality (INMETRO) 40.4 22.6 21.1 10.3 5.6 100 11.81 (0.003)�
Access to Land 52.0 13.7 14.5 13.2 6.6 100 10.87 (0.000)�
Patents and Registered Trademarks (INPI) 47.2 19.9 16.8 10.1 6.0 100 10.78 (0.015) in
Telecommunications 66.5 14.2 13.1 4.6 1.6 100 6.06 (0.344) Brazil
Statistical significance: * significant at 10%, significant at 5%, and significant at 1%.
� 55
Source: World Bank, Investment Climate Survey--Brazil, 2003.
References
Aaronson, Daniel, Raphael Bostic, Paul Huck, and Robert Townsend. 2000. "Supplier
relationships and small business use of trade credit." Federal Reserve Bank of Chicago
Working Paper 2000�28.
Angelini, P., R. Di Salvo, and G. Ferri. 1998. "Availability and cost of credit for small
businesses: customer relationships and credit cooperatives." Journal of Banking &
Finance 22:925�954.
Avery, Robert, Raphael Bostic, and Katerine Samolyk. 1998. "The role of personal wealth
in small business finance." Journal of Banking & Finance 22:1019�1061.
Baldwin, John, Guy Gellatly, and Valerie Gaudreault. 2002. "Financing innovation in new
small firms: new evidence from Canada." Analytical Studies Branch, Research Paper
Series No. 11F0019 No. 190.
Bank of England. "Finance for small firms." Various reports.
Barro, Robert, and Xavier Sala-I-Martin. 1995. "Economic Growth." New York,
McGrawhill.
Bechetti, Leonardo, and Giovanni Trovato. 2002. "The determinants of growth for small
and medium sized firms. The role of the availability of external finance." Small Busi-
ness Economics 19:291�306.
Beck, Thorsten, Ross Levine, and Norman Loayza. 2000. "Finance and the sources of
growth." Journal of Financial Economics 58:261�300.
Beck, Thorsten, Asli Demirgu�-Kunt, and Vojislav Maksimovic. 2002. "Financing patterns
around the world: The role of institutions." Policy Research Working Paper 2905,
World Bank.
------. 2003. "Financial and legal institutions and firms size." Policy Research Working
Paper 2997, World Bank.
Berger, Allen, Seth Bonime, Lawrence Goldberg, and Lawrence White. 2000. "The dynam-
ics of market entry: the effects of mergers and acquisitions on de novo entry and small
business lending in he banking industry." Journal of Business. Forthcoming.
Berger, Allen, Asli Demirgu�-Kunt, Ross Levine, and Joseph Haubrich. 2004. "Bank con-
centration and competition: an evolution in the making." Journal of Money, Credit,
and Banking. Forthcoming.
Berger, Allen, Laurence Goldberg, and Lawrence White. 2001. "The effects of dynamic
changes in bank competition on the supply of small business credit." European Finance
Review 5:115�139.
Berger, Allen, Iftekhar Hasan, and Leora Kappler. 2004. "Further evidence on the link
between finance and growth: an international analysis of community banking and eco-
nomic performance." Journal of Economics Services Research. Forthcoming.
Berger, Allen, Leora Kappler, and Gregory Udell. 2001. "The ability of banks to informa-
tionally opaque small business." Journal of Banking & Finance 25:2127�2167.
57
58 World Bank Working Paper
Berger, Allen, Anthony Saunders, Joseph Scalise, and Gregory Udell. 1998. "The effects
of bank mergers and acquisitions on small business lending." Journal of Financial
Economics 50:187�229.
Berger, Allen, and Gregory Udell. 1994. "Lines of credit and relationship lending in
small firm finance." The Wharton Financial Institution Center Working Paper
94�11.
------. 1998. "The economics of small business finance: the roles of equity and debt
markets in the financial growth cycle." Journal of Banking & Finance 22:613�673.
Berkowitz, Jeremy, and Michelle White. 2002. "Bankruptcy and small firms' access to
credit." NBER Working Papers 9010.
Berlin, Mitchel, and Loretta Mester. 1998. "On the profitability and cost of relationship
lending." Journal of Banking & Finance 22:873�897.
Bester, Helmut. 1987. "The role of collateral in credit markets with imperfect information."
European Economic Review 31(1987):887�899.
Bigsten, Arne, Paul Collier, Stefan Dercon, Marcel Fafchamps, Bernard Gauthier, Jan
Gunning, Mans Soderbom, Abena Oduro, Remco Oostendorp, Cathy Patillo, Francis
Teal, and Albert Zeufack. 2000. "Credit constraints in manufacturing enterprises in
Africa." Journal of African Economies 12(1):104�125.
Binks, Nartin, and Christine Ennew. 1996. "Growing firms and credit constraint." Small
Business Economics 8:17�25.
------. 1997a. "Smaller business and relationship banking: the impact of participative
behaviour." Entrepreneur Theory and Practice 21(4):83�92.
------. 1997b. "The relationship between U.K. banks and their small business customers."
Small Business Economics 9:167�178.
Blasio, Guido. 2003. "Does trade credit substitute for bank credit? Evidence from firms-
level data." IMF Working Paper 03/166.
BNDES. 2003. "Firmas e emprego demografia das firmas brasileiras." Informe-se BNDES
No 50.
Boot, Arnoud, and Anjan Thakor. 1994. "Moral hazard and secured lending in an infinitely
repeated credit market game." International Economic Review 35(4):899�920.
Brewer, Elijah, Hesna Genay, William Jackson, and Paula Worthington. 1996. "How are
small firms financed? Evidence from small business investment companies." Economic
Perspectives 20(6).
Carpenter, Robert, and Bruce Peterson. 2002. "Is the growth of small firms constrained by
internal finance?" The Review of Economics and Statistics 84(2):298�309.
Carrol, Robert, Douglas Holts-Eakin, Mark Rider, and Harvey Rosen. 2000. "Personal
income taxes and the growth of small firms." NBER 7980.
Cavaluzzo, Ken, and John Wolken. 2002. "Small business loan turndowns, personal wealth
and discrimination." Board of Governors, Federal Reserve System. Processed.
Chakravarty, Sugato, and James Scott. 1999. "Relationships and rationing in consumer
loans." Journal of Business 2(4)523�544.
Chittenden, Francis, and Martyn Robertson. 1993. Small Firms. Sage Publications, Inc.
Clarke, George, Robert Cull, and Maria Soledad Martinez Peria. 2001. "Does Foreign Bank
Penetration Reduce Access to Credit in Developing Countries? Evidence from Asking
Borrowers." World Bank Working Papers 2716.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 59
Clarke, George, Robert Cull, Maria Soledad Martinez Peria, and Susana Sanchez. 2002.
"Bank Lending to Small Businesses in Latin America: Does Bank Origin Matter?"
Journal of Money, Credit and Banking. Forthcoming.
Cole, Rebel. 1998. "The importance of relationships to the availability of credit." Journal
of Banking & Finance 22:959�977.
Demirg��-Kunt, Asli, and Vojislav Maksimovic. 1996a. "Financial constraints, use of
funds, and firm growth." Policy Research Working Paper 1671, World Bank.
------. 1996b. "Stock market development and firm financing choices." The World Bank
Economic Review 10(2).
------. 1998. "Law, finance, and firm growth." The Journal of Finance 53(6):2107�2137.
------. 1999. "Institutions, financial markets and Firm debt maturity." Journal of Finan-
cial Economics 54:295�336.
------. 2001. "Firms as Financial Intermediaries: Evidence from Trade Credit Data."
World Bank. Processed. http://econ.worldbank.org/files/2517_wps2696.pdf
Demsetz, Harold, 1973. "Industry Structure, Market Rivalry, and Public Policy." Journal
of Law & Economics 16(1):1�9.
Diamond, Douglas. 1989. "Reputation acquisition in debt markets." The Journal of Politi-
cal Economy 97(4):828�862.
Fazarri, Steven, and Michael Athey. 1987. "Asymetric information, financing constraints,
and investment." The Review of Economics and Statistics 69(3):481�487.
Ferri, Giovanni, and Marcello Messori. 2000. "Bank-firm relationship and allocative effi-
ciency in the Northeast and Central Italy and in the South." Journal of Banking &
Finance 24:1067�1095.
Fluck, Zsuzsanna. 1999. "Capital Structure Decisions in Small and Large Firms: A life-cycle
theory of financing." New York University, Leonard N. Stern School Finance Depart-
ment Working Paper Series No. 99�069.
Frazer, Garth. 2004. "Which Firms Die? A Look at Exit from Manufacturing in Ghana."
University of Toronto. Processed.
Freel, Mark. 1999. "The financing of small firm product innovation within the UK." Tech-
novation 19:707�719.
Gallup, John, Jeffrey Sachs, and Andrew Mellinger. 1998. "Geography and economic devel-
opment." NBER Working Paper No. 6849.
Gertler, Mark, and Simon Gilchrist. 1994. "Monetary policy, business cycles, and the behav-
ior of small manufacturing firms." The Quarterly Journal of Economics 109(2):309�340.
Giudici, Giancarlo, and Stefano Paleari. 2000. "The provision of finance to innovation: a
survey conducted among Italian technology based small firms." Small Business Eco-
nomics 14:37�53.
Goldberg, Lawrence, and Lawrence White. 1998. "De novo banks and lending to small
business: an empirical analysis." Journal of Banking & Finance 22:851�867.
Griliches, Z. 1998. "R&D and productivity, the econometric evidence." Chicago, Univer-
sity Chicago Press.
Guiso, Luigi, Paola Sapienza, and Luigi Zingales. 2002. "Does local financial development
matter." NBER Working Paper No. 8923.
Hamilton, Robert, and Mark Fox. 1998. "The financing preferences of small firm owners."
International Journal of Entrepreneurial Behavior & Research 4(3).
60 World Bank Working Paper
Hancock, Diana, and James Wilcox. 1998. "The credit crunch and the availability of credit
to small business." Journal of Banking & Finance 22:983�1014.
Hanson, James. 2004. "The role of state owned financial institutions: policy and practice."
SixthAnnual Financial Markets and Development conference, Washington, D.C.
Harris, Milton, and Artur Raviv. 1990. "Capital structure and the informational role of
debt." The Journal of Finance 45(2):321�349.
Hart, Peter, and Nicholas Oulton. 1996. "Growth and size of firms." The Economic Journal
106(438):1242�1252.
Hubbard, R. Glenn. 1998. "Capital-Market imperfections and investment." Journal of Eco-
nomic Literature 36(1):193�225.
Hughes, A., and D.J. Storey. 1994. Finance and the Small Firm. London: Routledge.
Johnson, Simon, John McMillan, and Christophe Woodruff. 2002. "Property rights and
finance." NBER Working Paper No. 8852.
Jordan, Judith, Julian Lowe, and Peter Taylor. 1998. "Strategy and financial policy in UK
small firms." Journal of Business and Accounting 25(1 & 2):1�27.
King, Robert, and Ross Levine. 1993. "Finance and growth: Schumpeter might be right."
The Quarterly Journal of Economics 108(3):717�737.
Klapper,Leora,VirginiaSarria-Allende,andVictorSulla.2002."Smallandmedium-sizeenter-
prise financing in Eastern Europe." World Bank Policy Research Working Paper 2933.
Kocenda, Evzen, and Jan Svejnar. 2003. "Ownership and firm performance after large-scale
privatization" William Davidson Working Paper No. 471a.
Kumar, Anjali, and Thorsten Beck, Cristine Campos, Soumya Chattopadhyay. 2004.
"Access to Financial Services in Brazil." World Bank. Processed.
Kumar, Jayesh. 2003. "Ownership Structure and corporate firm performance." Indira
Gandhi Institute of Development Research. Processed.
La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny. 1997.
"Legal determinants of external finance" NBER Working Paper No. 5879.
------. 2002. "Government ownership of banks." The Journal of Finance 57(1):265�301.
Laeven, Luc. 2001. "Financial Liberalization and Financing Constraints: Evidence from
Panel Data on Emerging Economies." World Bank. Processed.
Laursen, Keld, Volker Mahnke, and Per Vejrup-Hansen. 1999. "Firm growth a knowledge
structure perspective." Processed.
Levenson, Alec, and Kristen Willard. 2000. "Do Firms Get the Financing They Want? Mea-
suring Credit Rationing Experienced, by Small Businesses in the U.S." Small Business
Economics 14:83�94.
Levy, Brian. 1993. "Obstacles to Developing Indigenous Small and Medium Enterprises:
An Empirical Assessment." The World Bank Economic Review 7(1)65�83.
Long, Michael, John Wald, and Jingfeng Zang. 2003. "A cross sectional analysis of firm
growth options." Processed.
Lopez-Gracia, Jose, and Cristina Aybar-Arias. 2000. "An empirical research to the finan-
cial behavior of small and medium sized companies." Small Business Economics
14:55�63.
Love, Inessa. 2002. "Financial development and financing constraints: international evidence
from the structural investment model." Review of Financial Studies. Forthcoming.
Lucas, Robert (1998) "On the mechanisms of economic development" Journal of mone-
tary economics 22, 3�42.
Enterprise Size, Financing Patterns, and Credit Constraints in Brazil 61
Majumdar, Sumit, and Pradeep Chhiber. 1998. "State investors and state owner: conse-
quences of firm performance in India." Economic Development and Cultural Change.
Meyer, Laurence. 1998. "The present and future roles of banks in small business finance."
Journal of Banking & Finance 22:1109�1116.
Myers, Stewart. 1984. "The capital structure puzzle." The Journal of Finance 39(3):575�592.
Myers, Stewart, and Nicholas Majluf. 1984. "Corporative financing and investment
decisions when firms have information that investors do not have." Journal of
Financial Economics 13:187�221.
Modigliani, Franco, and Merton Miller. 1958. "The cost of capital, corporation finance and
the theory of investment." American Economic Review 48:261�297.
Natke, Paul. 1999. "Financial repression and firm self-financing of investment: empirical
evidence from Brazil." Applied Economics 31:1009�1019.
Nickell, Stephen, and Daphne Nicolitsas. 1995. "How does financial pressure affect firms."
Centre for Economic Performance Discussion Paper No. 266.
Nkusu, Mwanza. 2003. "Interest rates, credit rationing, and investment in developing
countries." IMF Working Paper 03/63.
Papadimitriou, Dimitri, Ronnie Phillips, and Randal Wray. 1994. "An alternative in small
business finance." Levy Institute Public policy brief No. 12/1994.
Peek, Joe, and Eric Rosengren. 1998. "Bank consolidation and small business lending: It's
not just bank size that matters." Journal of Banking & Finance 22:799�819.
Petersen, Mitchell, and Raghuram Rajan. 1994. "The benefits of lending relationships:
evidence from small business data." The Journal of Finance 49(1):3�37.
------. 2002. "Does distance still matter? The information revolution in small business
lending." The Journal of Finance 42(6):2533�2570.
Pollard, Jane. 2003. "Small firm finance and economic geography." Journal of Economic
Geography 3(4):429�452.
Rajan, Raghuram, and Luigi Zingales. 1998. "Financial dependence and growth." The
American Economic Review 88(3):559�586.
Romer, Paul. 1990. "Endogenous technological change." Journal of Political Economy
98:71�102.
Rousseau, Peter, and Paul Wachtel. 2000. "Equity markets and growth: cross-country evi-
dence on timing and outcomes, 1980�1995." Journal of Banking and Finance
24:1933�1957.
Russo, Paolo, and Paola Rossi. 2001. "Credit constraints in Italian industrial districts."
Applied Economics 33:1469�1477.
Saito, Katrine, and Delano Villanueva. 1981. "Transactions costs of credit and the small
scale sector in the Philippines." Economic Development and Cultural Change
29(3):631�639.
Schiantarelli, Fabio, and Alessandro Sembenelli. 1996. "Form of ownership and financial
constraints." World Bank Policy Research Paper 1626.
Sharpe, Steven. 1990. "Asymmetric information, bank lending and implicit contracts: a
stylized model for customer relationships." The Journal of Finance 45(4):1069�1087.
Short, Helen. 1994. "Ownership, control, financial structure and the performance of
firms." Journal of Economic Surveys 8(3):203�249.
Solow, R. 1957. "Technical change and the aggregate production function." Review of eco-
nomics and Statistics 39:312�320.
62 World Bank Working Paper
Stiglitz, Joseph, and Andrew Weiss. 1981. "Credit rationing in markets with imperfect
information." The American Economic Review 71(3):393�410.
Strahan, Philip, and James Weston. 1998. "Small business lending and the changing struc-
ture of the banking industry." Journal of Banking & Finance 22:821�845.
Terra, Maria. 2003. "Credit constraints in Brazilian firms: evidence from panel data."
Revista Brasileira de Economia 57(2):443�464.
Topalova, Petia. 2004. "Overview of the Indian Corporate Sector: 1989�2002." IMF Work-
ing Paper 04/64.
Vinning, Aidan, and Anthony Boardman. 1992. "Ownership versus competition: efficiency
and public enterprise." Public choice 73:205�239.
Wachtel, Paul. 2003. "How much do we really know about growth and finance?" Economic
Review--Federal Reserve of Atlanta (First quarter 2003):33�47.
Watson, Robert, and Nick Wilson. 2002. "Small and Medium size enterprise financing: a
note of the empirical implications of a pecking order." Journal of Business Finance and
Accounting 29(3 & 4):557�578.
Willmore, Larry. 1986. "The comparative performance of foreign and domestic firms in
Brazil." World Development 14(4):289�502.
Wilson, Nicholas, and Barbara Summers. 2002. "Trade credit terms offered by small firms:
survey evidence and empirical analysis." Journal of Business Finance and Accounting
29(3 & 4):317�351.
Enterprise Size, Financing Patterns, and Credit Constraints in
Brazil is part of the World Bank Working Paper series. These
papers are published to communicate the results of the Bank's
ongoing research and to stimulate public discussion.
This study investigates the importance of firm size with
respect to access to credit, relative to firm performance and
other factors which may affect creditworthiness--such as
management education, location, and the industrial sector to
which the firm belongs.
The principal findings are that size strongly affects access to
credit, compared to performance as well as other variables,
suggesting quantitative limitations to credit access. Looking
at short versus long-term loans, the impact of size on access
to credit is greater for longer terms. Regarding ownership of
the lending institution, the study finds public financial institu-
tions are more likely to lend to large firms. Finally, examining
the role of financial constraints relative to other constraints
faced by the firm, financial access constraints may have a less
signficant differential impact across firms of different sizes
than other constraints, though cost of finance as a constraint
is very important.
World Bank Working Papers are available individually or by
subscription, both in print and online.
ISBN 0-8213-6129-5
TMxHSKIMBy361290zv":':,:-:*
THE WORLD BANK
1818 H Street, NW
Washington, DC 20433 USA
Telephone: 202 473-1000
Internet: www.worldbank.org
E-mail: feedback@worldbank.org