Does bank competition alleviate credit constraints in developing countries?

Abstract

Whether competition helps or hinders firms’ access to finance, particularly in the developing world, is in itself a much debated question in the economic literature and in policy circles. This paper considers the consequences of bank competition on credit constraints using firm level data covering 69 developing and emerging countries. In addition to the classical concentration measure, competition is assessed by computing three non-structural measures (Boone indicator, Lerner index and H-statistic). The results show that bank competition alleviates credit constraints and that bank concentration measure is not a robust predictor of a firm’s access to finance. The study highlights that bank competition not only leads to less severe loan approval decisions but also reduces borrower discouragement.

Introduction

Limited access to bank credit is viewed by many policy makers and academics as a major growth constraint for developing economies, in particular for small- and medium-sized enterprises (Beck and Demirguc-Kunt, 2006). As a result, numerous resources have been devoted to improving credit availability around the world. Competition in the banking sector is often cited as an important driver of access to credit. The traditional market power view argues that market power is detrimental in banking as well as in other industries as fiercer competition leads to lower costs and better access to finance (Besanko and Thakor, 1992, Guzman, 2000). In the presence of information asymmetries and agency costs, however, competition can reduce access by depriving banks of the incentive to build lending relationships (Petersen and Rajan, 1995).¹ Other contributions document that the quality of screening (Broecker, 1990, Marquez, 2002) and banks’ incentives to invest in information acquisition technologies (Hauswald and Marquez, 2006) are higher in less competitive markets. Therefore, the information hypothesis argues that access to credit for opaque borrowers can decrease when competition is harsher.

The answer to which of these views best describes the implications of competition on access to credit is ultimately an empirical issue. Despite the policy relevance of this question, recent works have failed to provide a clear conclusion on the effect of interbank competition on access to finance in developing countries. In line with the market power view, Beck et al. (2004) show that bank concentration increases the probability that firms will report finance as a major obstacle to growth. Chong et al. (2013) also confirm the market power view documenting that credit availability is restrained in concentrated markets in China. Adopting a similar approach on data from the Philippines, Tacneng (2014) provides support for the information hypothesis view as local concentrated banking markets are generally associated with an increase in credit accessibility. González and González (2008) also find results in line with the information hypothesis. A major concern with these studies involves the proxy of competition used; the degree of competition is assessed by market concentration. Several contributions have cast doubt on the consistency and robustness of the structural approach as an accurate indicator of competition in banking (Degryse et al., 2009). Carbó-Valverde et al. (2009b) document that this issue is not only a technical problem insofar as conclusions may be sensitive to the measure of competition employed.

A recent wave of works has attempted to provide more robust results by employing non-structural measures of competition. Claessens and Laeven (2005) examine the influence of banking competition on economic growth in 16 countries, using the Panzar–Rosse H-statistic as a non-structural approach to indicate market competition. They find that sectors heavily dependent on bank financing grow faster in countries where there is fierce bank competition. Liu and Mirzaei (2013) confirm the market power view. Fernández de Guevara and Maudos (2011) measure competition by employing both the Lerner index and the Panzar-Rosse H-statistic and find opposing evidence that the exercise of market power enhances economic growth. Using the Panzar-Rosse H-statistic, Hoxha (2013) reaches similar conclusions in favor of the information hypothesis. Recent papers directly investigate the relationship between the use of credit and bank competition. Using a panel data analysis on 53 developing countries, Love and Martinez Peria (2015) show that the use of bank loans is lower in more competitive markets, providing additional support for the market power view. However, in a sample of 33 countries, Mudd (2013) obtains evidence of a more complex relationship between bank competition and a firm’s probability of obtaining a line of credit. Competition (measured by the Panzar-Rosse H-statistic) has a positive but declining effect on a firm’s use of bank financing, which turns into a negative impact for competitive markets above the mean level. As a result, the evidence is mixed and results are sensitive to the sample and the measures of competition and credit availability employed. In this paper, we propose to provide more robust evidence by using different measures of competition and by considering a large number of developing countries.

The principal contribution of this paper is to shed new light on the relationship between bank competition and access to credit by using several measures of competition and a large scope of developing countries. We also examine the potential channels through which bank competition may affect the use of credit by firms. Econometric analysis considers 28,642 firms from 69 developing countries across four continents. Firm-level variables are extracted from the World Bank Enterprises Surveys (WBESs) and country-level variables are taken from diverse sources. Following Popov and Udell (2012), those firms that were either discouraged from applying for a loan or were rejected when they applied are classified as credit constrained. Information on bank competition is extracted from the Global Financial Development Database. Competition in the banking industry is assessed by the degree of market concentration (share of assets held by the three largest banks) and by three non-structural measures (Boone indicator, Lerner index and Panzar-Rosse H-statistic).

The results show that financing constraints are alleviated in countries where banking markets are more competitive, irrespective of whether competition is measured by the Boone indicator or the Panzar-Rosse H-statistic, thus supporting the market power hypothesis. Employing the Lerner index to proxy market power suggests a similar conclusion, while the results are not statistically significant. The degree of concentration has no impact on credit availability. The empirical set-up allows us not only to test the information hypothesis against the market power hypothesis but also to investigate the channels by which competition affects credit availability. In particular, we investigate the determinants of a firm’s decision to apply for credit and a bank’s decision to approve or reject the request. The results show that competition not only leads to less severe loan approval decisions but that it also reduces borrower discouragement.

This paper makes several contributions to the literature on the relationship between bank competition and credit availability in developing countries by addressing a number of issues that have not yet been resolved in the literature. First, this paper considers different non-structural measures of competition to investigate the implications of competition on credit availability. There is no consensus regarding the best measure by which to gauge competition. As shown by Carbó-Valverde et al. (2009b), the choice of a particular indicator may influence conclusions regarding the implications of competition. By confronting the results from several non-structural measures, this paper gives a more complete picture of the role of competition on credit availability. Second, the identification of financially constrained firms remains a challenge. Existing studies measure credit constraints in terms of a firm’s perception of finance as an obstacle to growth (Beck et al., 2004) or the mix of internal and external funds (Carbó-Valverde et al., 2009b, Ryan et al., 2014). The survey measures are subject to perception bias (Ergungor, 2004) and the mix of funds requires data on firms’ balance sheets, rarely available for SMEs in developing countries. Recent works proxy credit constraints by the use of credit (Love and Martinez Peria, 2015, Mudd, 2013). The fact that few firms obtain credit is, however, not sufficient to prove constraints, since certain firms may not have a need for credit. This hypothesis is far from anecdotal in developing countries: indeed, our data documents that more than one third of firms had no need for credit. A comprehensive definition of credit constraints should measure the demand for credit that is unfulfilled by the existing supply of credit due to market imperfections. In this paper, we classify a firm as credit constrained if its demand is not fulfilled by the supply of funds due to market frictions. Specifically, among firms with a need for credit we classify as credit constrained firms whose loan application was turned down as well as those who were discouraged from applying in the first place. The third contribution involves country coverage. Our baseline model considers 69 developing countries including many low-income countries rarely considered in previous studies.

An additional contribution consists of not only testing the market power hypothesis vs. the information hypothesis but also investigating the channels by which competition affects credit availability (impact on lenders or borrowers). In this way, this work also adds to the empirical literature on borrower discouragement and approval/denial decisions in developing countries. Empirical studies are generally focused on the U.S. (Cole, 1998, Cavalluzzo et al., 2002, Han et al., 2009). Recent contributions have begun to investigate the determinants of a firm’s discouragement or a bank’s rejection decision in the case of developing countries (Bigsten et al., 2003, Brown et al., 2011, Chakravarty and Xiang, 2013) but often analyze the characteristics of firms or managers. Brown et al. (2011) underline that in Eastern Europe country characteristics play a significant role in borrower discouragement and in bank denial. Taking advantage of the broad scope of countries in our sample, we investigate a large set of country characteristics, including bank competition. The findings shed light on the important role played by country characteristics in explaining borrower discouragement and bank decisions.

The rest of the paper is organized as follows. Section 2 presents the data and describes the construction of variables and the empirical methodology. The baseline results are discussed in Section 3. Section 4 presents some tests in order to understand how competition affects credit availability and the final section concludes.