Influence of bank concentration and institutions on capital structure: New international evidence

Abstract

This paper analyzes the effect of bank market concentration and institutions on capital structure in 39 countries. Results for 12,049 firms over 1995–2004 indicate that firm leverage increases with greater bank concentration and stronger protection of creditor rights, but drops with stronger protection of property rights. The results also indicate that greater bank concentration substitutes for creditor protection and asset tangibility to reduce the agency cost of debt between shareholders and debtholders. Weaker protection of property rights raises the agency cost of external funds, leading to the preferential use of internal funds as posited by the pecking order theory. The trade-off theory, however, is more valid in countries with stronger protection of property rights.

Introduction

Traditional analysis of determinants of capital structure has focused on firm characteristics mainly related to the extent of agency costs and asymmetric information (Jensen, 1986, Harris and Raviv, 1991). These studies have generally analyzed samples of US companies by comparing the trade-off theory (TOT) with the pecking order theory (POT). Recent papers argue that capital structure is affected not only by firm-level variables, but also by country-level variables, or a firm's legal and institutional environment.¹ We add to this growing literature. Our main contributions are the analysis of: 1) how the determinants of firm leverage vary across countries depending on bank concentration and the legal and institutional system, and 2) how bank concentration interacts with the legal and institutional system to address conflicts of interest between shareholders and debtholders. This analysis lets us to examine the different validity of the TOT and POT across countries.

Empirical literature suggests that an effective legal system favors a firm's use of external funds, while companies in poorer contracting environments depend more on internal funds (Demirgüc-Kunt and Maksimovic, 1999, Qian and Strahan, 2007). A well-developed legal system affects not only the extent of external funding but also increases the proportion of long-term debt (Demirgüc-Kunt and Maksimovic, 1999, Giannetti, 2003, Fan et al., 2006).

Most international studies in this area of research use country-level data. Rajan and Zingales (1995), Booth et al. (2001), Giannetti (2003), and Fan et al. (2006) use firm-level data for cross-country comparisons of capital structure. Bae and Goyal (2004) and Qian and Strahan (2007) also analyze how country variables affect the characteristics of bank loans.

These papers offer mixed evidence on the influence of institutions on firm capital structure. Rajan and Zingales (1995) in an analysis of large listed companies in G-7 countries find that factors identified as cross-sectionally correlated with firm leverage in the United States are similarly correlated in other countries as well. Booth et al. (2001) suggest that the same determinants of capital structure prevail in ten developing countries. This research together suggests that institutional differences are unimportant in both developed and developing countries, although the authors compare separate regressions for each country and do not use explicit variables for the institutional environment in their estimations.

Giannetti (2003) explicitly introduces creditor right protection in the analysis of a sample of listed and unlisted companies in eight European countries. Her results suggest that the relevance of institutional variables depends on firm size. Larger listed companies have easier access to international financial markets, so their corporate finance decisions are less subject to the institutional constraints in domestic markets. Institutional characteristics, however, have a greater impact in unlisted companies. For these companies, stronger creditor protection makes loans for investing in intangible assets more available and guarantees access to long-term debt for firms in sectors with highly volatile returns. Fan et al. (2006) find that institutional factors are critical determinants of firm capital structure in a cross-section of 39 developed and developing countries. A country's legal and tax system, level of corruption, and the availability of information intermediaries explain a significant portion of the cross-country variation in leverage. Firms in common law countries, for example, have less leverage and use more long-term debt.

Finally, Bae and Goyal (2004) and Qian and Strahan (2007) show the relevance of institutional conditions for terms of bank loans. Bae and Goyal (2004) find lenders charge lower spreads on loans in countries with stronger property rights protection, while the latter find that stronger creditor rights enhanced loan availability.

Our work makes several contributions to this literature. First, we analyze the influence of bank concentration and institutional characteristics not only on firm leverage but also on the firm-level determinants of leverage. Our analysis shows that either the pecking order or trade-off theories apply differently across countries, depending on institutions and bank concentration. Weaker protection of property rights increases the agency cost of external funds, leading to the preferential use of internal funds as posited by the POT. The TOT, however, is more valid in countries with stronger protection of property rights.

Second, we analyze the interaction of bank concentration with a country's legal and institutional system (protection of property and creditor rights) to reduce agency costs and mitigate information asymmetries between shareholders and debtholders. The results suggest that greater bank concentration can substitute for creditor protection and asset tangibility to reduce the agency cost of debt.

Third, we analyze more countries than most previous studies. We include a sample of 12,049 firms in 39 countries over the period 1995–2004 (the same number of countries as Fan et al., 2006), compared to seven countries in Rajan and Zingales (1995), eight countries in Giannetti (2003), and ten in Booth et al. (2001). We thus can provide information on a greater range of institutional differences to give us a deeper understanding of how capital structure depends on institutions and on bank concentration.

Finally, we account for dynamic processes in firm leverage using generalized-method-of-moments (GMM) estimators developed by Arellano and Bond (1991) for dynamic panel data. GMM models are designed to handle autoregressive properties in the dependent variable (firm leverage) when lagged values are included as explanatory variables and endogeneity in the explanatory variables (other firm-specific characteristics) must be controlled for. Although the GMM method has been used in studies on capital structure focusing on a single country, it has not yet been applied in studies using international data.²

The rest of the paper is organized as follows. Section 2 discusses the influence of bank concentration and institutions on the determinants of firm capital structure and the hypotheses tested in the paper. Section 3 describes the characteristics of the database and methodology, while Section 4 discusses the empirical results. Section 5 checks the robustness of our basic results. Finally, Section 6 concludes the paper.