The determinants of leverage decisions: Evidence from Asian emerging markets

Quratulain Zafar; Wongsurawat, Winai; Camino, David Article The determinants of leverage decisions: Evidence from Asian emerging markets Cogent Economics & Finance Provided in Cooperation with: Taylor & Francis Group Suggested Citation: Quratulain Zafar; Wongsurawat, Winai; Camino, David (2019) : The determinants of leverage decisions: Evidence from Asian emerging markets, Cogent Economics & Finance, ISSN 2332-2039, Taylor & Francis, Abingdon, Vol. 7, Iss. 1, pp. 1-28, https://doi.org/10.1080/23322039.2019.1598836 This Version is available at: https://hdl.handle.net/10419/245223 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=oaef20 Cogent Economics & Finance ISSN: (Print) 2332-2039 (Online) Journal homepage: https://www.tandfonline.com/loi/oaef20 The determinants of leverage decisions: Evidence from Asian emerging markets Quratulain Zafar, Winai Wongsurawat & David Camino | To cite this article: Quratulain Zafar, Winai Wongsurawat & David Camino | (2019) The determinants of leverage decisions: Evidence from Asian emerging markets, Cogent Economics & Finance, 7:1, 1598836, DOI: 10.1080/23322039.2019.1598836 To link to this article: https://doi.org/10.1080/23322039.2019.1598836 © 2019 The Author(s). This open access article is distributed under a Creative Commons Attribution (CC-BY) 4.0 license. Published online: 17 Apr 2019. Submit your article to this journal Article views: 2579 View related articles View Crossmark data Citing articles: 1 View citing articles FINANCIAL ECONOMICS | RESEARCH ARTICLE The determinants of leverage decisions: Evidence from Asian emerging markets Quratulain Zafar 1 *, Winai Wongsurawat 2 and David Camino 3 Abstract: This study provides a stage-level analysis of firm-scale pooled data of 16 Asian countries to classified income economy-based data for various firmand country-specific predictors of leverage. Our analysis captures the selection impact of both microand macro-level determinants on capital structure with and without income economy-based models. The regression model evaluated the significance of predictor variables based on random effect model of panel data setting. The study further explores the issue of interest by looking at key individual regression models by income economy to avoid any potential loss of information. We argue that this approach provides a comprehensive and insightful set of determinants because of the newer dimension of income economy classification based on per-capita Gross National Product (GNP) defined by the World Bank. The estimating equations for financing determinants identify the additional variables of non-debt tax shield, liquidity, tax and GDP growth rate in case of Asian countries. Our study establishes that the core variables of tangibility, growth, size, and profitability retain their significance for leverage choice in both options during 2008–2014 in Asian economies. Furthermore, the findings show that the financing choices of firms in Asian Quratulain Zafar ABOUT THE AUTHORS Ms. Quratulain is Assistant professor at BUITEMS, Pakistan. She is currently pursuing her PhD in FinancialEconomics, from Asian Institute of Technology, Thailand with scholarship from HEC under Faculty Development program. Her research encompasses financial development system and relationship with financing choices in 16 Asian Countries during 2008-2014. Dr. Winai is an applied economist whose research interests span areas of government regulation, investment, entrepreneurship, and business strategy. From 2005-2007, worked as consultant at NERA Economic Consulting, New York, USA. He earned his PhD in managerial economics and strategy from Kellogg School of Management, Northwestern University. Dr. David is Professor of Finance and Director of University Institute of Law and Economics, at Universidad Carlos III de Madrid. He earned his Ph.D. degrees in Management and Business Economics from Universidad Autónoma de Madrid. His research interests are in Behavioral Finance, Investment Analysis, International Trade & Finance, Real Options, Risk Management. PUBLIC INTEREST STATEMENT We primarily contribute by classifying the micropanel data and macrovariables being effective determinants for three Income economies composed of 16 Asian countries as classified by the World Bank from 2008–2014, in contrast to previous studies in which countries are analyzed as pooled only. The research findings provide a significant contribution for: (1) Shareholders and market investors by developing a better understanding of the capital structure determinants for Asian firms. (2) Academicians will understand the role of financing decisions in maximizing the firm value. (3) The corporate managers can use the discussion on determinants as guidelines for their financing strategies to facilitate investment decisions. Our study is first of its kind to elaborate the impact of Income-economy wise basis where the expertise and policy application for financial development sector of upper middle income-economies can be replicated in lower middle-income economies as role models. Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 © 2019 The Author(s). This open access article is distributed under a Creative Commons Attribution (CC-BY) 4.0 license. Received: 05 November 2018 Accepted: 17 March 2019 First Published: 23 March 2019 *Corresponding author: Quratulain Zafar, School of Management, Asian Institute of Technology (AIT), Thailand E-mail: [email protected] Reviewing editor: Mohammed M Elgammal, Finance and Economics, University of Qatar, Qatar Additional information is available at the end of the article Page 1 of 28 regional markets are complemented by financial system development stages using the equity market, the bond market and the banking industry as proies. Subjects: Development Economics; Finance; Industry & Industrial Studies Keywords: leverage; firm-specific determinants; country-specific determinants; income economies; Asian countries JEL classification: G30; G20; F61; F62; C33 1. Introduction Research on capital structure and its relation to firm value was initiated by Modigliani and Miller (MM) (1958). They further revised the first theory in 1963 by incorporating tax shelters, bankruptcy costs, and asymmetric information in the capital structure. The trade-off theory for capital structure by Baxter (1967) and Kraus and Litzenberger (1973) indicates that the target capital structure choices of firms are determined by balancing the bankruptcy cost and tax-saving debt benefits against the cost of borrowing. Trade-off theory is strictly criticized by Myers and Majluf (1984) because tax-paying firms ruled out the theory of a conservative capital structure. Although tax benefits seem to be considerable according to Graham (2000), while Frank and Goyal (2003) explored the practical significance of trade-off theory. Consistent with Fama and French (2002), trade-off theory most convincingly explains a firm’sdeparture from an optimal capital structure due to its dynamic characteristics. The precise interpretation of pecking order theory (Lemmon, Roberts, & Zender, 2008) led to the concept of debt capacity, which suggests that firms should consider the equity issue only. This rigorous explanation has led researchers to focus on the concept of modified pecking order theory. Less than 20% of the firms follow pecking order theory prediction for debt and equity choices (Leary & Roberts, 2010). Frank and Goyal (2003)believethat pecking order theory is more relevant for large firms’choice of capital structure because of asymmetric information, whereas Byoun and Rhim (2005) prove that the pecking order is more relevant for small firms. By re-examining determinants, the authors continuously evaluate the validity of the competing capital structure theories to bridge the gap between theoretical explanations and financial practices concerning capital structure, particularly for Asia, because the empirical studies have provided inconclusive results (Brounen, de Jong, & Koedijk, 2006). We re-examine the validity of firm-specific and macroeconomic variables being portable determinants of capital structure in the Asian region in the pooled model form as well as in income economy groups. We will also test the explanatory power of the firm-specific factors representative of three major theories for their significant impact on capital structure. Considering the background of dynamic changes and the associated problems arising from inconsistent studies (Haron, 2014;Lemmon&Zender,2010), the authors of following study strongly call for identification of the most relevant and current developments in optimal capital structure studies specializing in Asia (Leary & Roberts, 2010; Sibindi & Makina, 2018). We primarily contribute to the literature by classifying the effective determinants for three income economies composed of 16 Asian countries as classified by the World Bank in contrast to previous studies in which countries are analyzed in pooled form only. Using the World Bank’s definition of income economies (IEs), we classify 16 Asian countries into three IEs based on the per-capita gross national product (GNP). In this study, the dataset is divided into lower middle-IEs (LMIEs), upper middle-IEs (UMIEs) and high-income economies IEs (HIEs), which are described in Table 3. We proceed to explore the differences in leverage choice when 16 sample countries are pooled as one group as compared to when the 16 countries are segregated on an income economy basis. Debt and equity are raised through institutions that include stock markets, banking institutions, bond issuing agencies, and investment banks. Thus, the selected firm variables are representative of all three capital structure theories and country-specific factors. The impacts of institutional differences are captured through country-specific determinants, including both stock market orientation, the development of the banking sector and bond markets. The institutional Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 2 of 28 environment remains distinctive around the continent, which may influence firms’operational and financing decisions (Desai, Foley, & Hines, 2004) In addition to contributing to existing knowledge, the authors seek to extend a major field study by De Jong, Kabir, and Nguyen (2008) with our sample of 16 Asian countries and to verify the stylized facts presented by Frank and Goyal (2009) that are applicable to Asian firms in recent times. We also find appropriate dimensions for inconclusive results by concentrating on the selected Asian economies (Haron, 2014). Re-examination of the variables in these countries will yield the latest empirical evidence on the determinants of capital structure. We will also discuss the applicability of the capital structure study of Rajan and Zingales (1995) for G7 economies, where tangibility, growth, firm size and profitability remain the major determinants. The objective of this paper is to examine the relevant determinants of capital structure in 16 Asian countries and three income economies, thus adding literature to the body of knowledge on the importance of firm-specific factors, such as profitability, net tangible assets, firm size and depreciation to total assets, and macroeconomic variables, such as stock market development, the banking industry and the bond market in addition to the GDP growth rate (GRT) and inflation (INF). Our study suggests that leverage choices of firms in Asian regional markets are complemented by financial system development stages. We count upon the assessment of Demirgüc¸ Kunt and Maksimovic (1994) regarding developing countries in which the differences in capital structure choices are attributed to the differences in the levels of economic and financial developments and institutional differences. We use the equity market, bond market and banking industry factors as proxies for the financial system in pooled and income economy systems and discover that lower middle-income economies have the highest mean leverage of 1.323, while high-income economies have the lowest mean leverage of 1.181 for the sample period 2008–2014. Our study extends the major fieldwork of Booth, Aivazian, Demirgüc¸ Kunt, and Maksimovic (2001) in terms of countries’ firm-specific and country-level determinants during the most recent period of 2008–2014 in 16 Asian economies. To identify the most significant determinants in the Asian context, we investigate micropanel data with country-level variables from 2008–2014 for the sample countries. China and India are the two fastest-growing major economies; our initial results show that India has the highest mean leverage of 1.567 for non-financial firms, followed by China’s leverage of 1.562 and Japan’s leverage of 1.549. In the bottom tier, Jordan has the third lowest leverage at 0.944, followed by Saudi Arabia as the second lowest at 0.887 and Malaysia as the lowest at 0.730. Furthermore, Asia has experienced some of the longest modern economic booms in the world, particularly in Japan (1950–1990), South Korea (1961–1996), and China (1978–2013), followed by current rapid economic growth in Philippines and India. Considering the synergistic outlook of Asia’s developing economies and their importance to the world economic outlook, the Asian region is selected to examine the macro and microfinancing determinants of Asian economies. Our empirical study aims to address the following questions: (1) What are the determinants of capital structure choice for three Asian income economy groups and 16 sample countries in Asia? (2) Do corporate financial leverage decisions differ significantly between the country-based model and income economy-based models for Asian firms? (3) Are the conventional capital structure models still applicable to both cases during the 2008– 2014 period? The paper is structured as follows: Background information is provided in Section 1, while Section 2 presents a literature review and discusses various capital structure theories and variable selection. The proposed capital structure model is presented in Section 3. The panel data regression analysis and the results are reported in Sections 4and 5. Section 6summarizes the main findings and concludes the article. Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 3 of 28 2. Literature review The puzzle of capital structure (Myers, 2001) remained unresolved for more than five decades. In a perfect capital market, when we interpret debt and equity as substitutable, firm value is assumed to be independent of capital structure choice as established by Modigliani and Miller (1958). Following studies since 1963, MM discarded the perfect capital proposition due to friction factors such as bankruptcy costs, transaction costs and agency costs and taxes, which has encouraged researchers to develop alternative theories of capital structure considering that capital structure is crucial for maximizing firm value across the globe. These theories include trade-off theory (Kraus & Litzenberger, 1973), pecking order theory (Myers, 1984), agency theory (Jensen & Meckling, 1976) and market timing theory (Baker & Wurgler, 2002; Demirgüç-Kunt & Maksimovic, 2002). These theories outline the impacts of determinants in either direction for various justifications (Bayrakdaroglu, Ege, & Yazici, 2013; Frank & Goyal, 2009;Haron,2014; Mahajan&Tartaroglu,2008;).HuangandRitter(2009) and Leary and Roberts (2010) precisely explain that capital structure theories are not mutually exclusive because none of them can explain all relevant factors independently. We follow the description by Kayhan and Titman (2007) to reconcile explanations between trade-off theory and pecking order theory by introducing modified pecking order theory. These authors reported that long-term capital structure is guided by Trade-off theory, while short-term decisions are influenced by pecking order theory. An extensive range of studies on capital structure provide a wide array of factors from three well-recognized theories: trade-off theory, pecking order theory and free cash flow theory (Myers, 2001; Sánchez-Vidal & Martín-Ugedo, 2005). Trade-off theory refers to capital structure as a trade-off between the costs of financial distress and the tax shield. The literature on trade-off theory discusses concerns about tax-shield benefits, financial distress costs including cash flow volatility, and possible bankruptcy costs. These issues have been extensively discussed in DeAngelo and Masulis (1980)andMyers(2001). According to pecking order theory, firms prefer retained earnings as an internal funding source over external funding. When funds from internal sources are insufficient to finance capital expenditures, firms consider borrowing from external sources easier rather than issuing equity (Myers & Majluf, 1984). According to agency cost-based theory, agency costs arise because of the use of debt in capital structures Hart and Moore (1995)andJensenand Meckling (1976). Jensen and Meckling (1976) introduced two major types of conflicts that emerge from agency costs. One type of conflict is between managers and shareholders, and the other is a conflict between shareholders and bondholders (Myers, 2001). Debt-mitigated agency conflicts between shareholders and managers can be found in many important studies, including Jensen and Meckling (1976) and Hart and Moore (1994). Research has shifted toward emerging countries to highlight the probable deviation in factors that may have emerged from the 1997 Asian financial crisis. From an early examination of seven leading industrial economies (Rajan and Zingales, 1995), we learn about the prevailing firm-specific determinants that considerably affect firms’capital structures across countries in parallel with the country-specific factors that influence leverage choices across those same countries. Early empirical studies on leverage decisions examined the determinants of capital structure for U.S. firms (Titman & Wessels, 1988), while studies by Rajan and And Zingales (1995)and Beck, Demirgüç-Kunt, and Maksimovic (2004) and La Porta, Lopez-de-Silanes, Shleifer, & Vishny (2000) focused on crosscountry comparative studies. By analyzing global patterns, it was established that capital structure varies with different institutional settings and macroeconomic variables. Booth et al. (2001)pioneered a study of emerging markets and indicated that the determinants of capital structure vary across developing countries. They also stated that capital structures in developing countries are affected by the same firm-specific factors that had previously been discussed in the context of developed countries. Some major studies by Huang and Ritter (2009) and Korajczyk and Levy (2003) highlighted the impact of macroeconomic variables on capital structure according to country-level differences. However, Deesomsak, Paudyal, and Pescetto (2004) argued that despite the importance of economic structure and cross-countries diversity, few studies have been conducted in ASEAN countries. Based on an analysis of 10 developing countries, Chen and Strange (2005)foundthat economic growth rates, inflation, and unemployment rates significantly and positively influenced capital structures and operational risks. Inflation increases the level of debt (Wu & Kim, 1988)while Chadegani, Nadem, Noroozi, and Madine (2011) showed that interest rates, inflation, and the GDP Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 4 of 28 negatively influence the debt-to-equity ratio, whereas exchange rates have a positive relation with leverage. However, differences exist in how some country-specific factors are related to GRTs and how capital markets impact leverage. Therefore, De Jong et al. (2008) examined both firmand country-specific factors in an equally divided sample of 42 developed and developing countries. Their analysis shows that in addition to the direct effects of country-specific factors on firms’leverage, indirect effects also exist when country-specific factors signify firm-specific factors. Core factors (Frank & Goyal, 2009) that show consistent signs and statistical significance across alternative data treatments are considered as control variables. The remaining factors, which are less consistent over a wide range of studies, are considered as test variables in this study. Bokpin (2009)showedthat control variables have significant impacts on capital structures in 34 emerging markets from 1990– 2006 and discussed the significant impact of stock market development, bank credit, inflation, and the GDP on leverage structure. Demirgüç-Kunt and Maksimovic (2002), in their study of 10 developing countries, conclude that the capital structure variables that are significant in the case of developing countries (including Jordan, Pakistan, Korea, Malaysia, Turkey, and Thailand) are like those that impact capital structure choices in the U.S., such as tangibility, liquidity, firm size and growth. Remarkable studies conducted on international capital structures with firmand countryspecific variables across countries include Booth et al. (2001), Deesomsak et al. (2004), Song and Philippatos (2004), Fan, Titman, and Twite (2012), and Hall, Hutchinson, and Michaelas (2004). The literature above identifies the determinants of capital structure for maximizing firm value and providing theoretical foundations for leverage decisions are subject to the empirical testing by field researchers. Scholars continue searching for and integrating theoretical explanations to determine debt and equity issues. Empirical studies demonstrate that basic firmspecific variables, including tangibility, size, ownership, tax, non-debt tax shield (NDTS), liquidity, dividends, profitability, risk, cash flow, expected growth rates, and inflation are important country-specific variables affecting leverage (Frank & Goyal, 2009). However, Kayo and Kimura (2011) prove that only firm-specific factors are important variables rather than macroeconomic variables across 40 sample countries. However, recent studies on capital structure in emerging markets by Delcoure (2007), Fauzi, Basyith, and Idris (2013), and M’ng, Rahman, and Sannacy (2017) provide contradictory results about both types of factors across Asian countries. For instance, the research on the economic effects of inflation (Barry, Mann, Mihov, & Rodriguez, 2008) and growth rates on capital structure are still lacking despite these factors’influences on firms’operating income and cash flow. From the literature review, we find that although empiricalstudiesareincreasinginindividual developing countries, few studies have been conducted in the Asian region (Deesomsak et al., 2004; Driffield & Pal, 2010). The present study adds valuable knowledge to the existing literature by empirically testing the determinants of capital structure in 16 Asian countries in pooled status as well as in segregated status of income economies as per the World Bank classification. The variables determining leverage across Asian firms in our 16 sample countries from Asia are shown in Table 1, which are identified in the present review of previous studies. 2.1. Variable selection The variables determining leverage across Asian firms are described in Table 1.Allthevariables are representative of each of the three core theories and will be regressed in addition to stylized variables as independent variables for panel regression to study the empirical relations identified in the literature review section. Our research discusses the influence of institutional differences captured through stock market orientation, size of the banking industry (SBI) and the size of the bond market (SBM) as country-specific factors. Our results support the empirical findings of Booth et al. (2001), Antoniou, Guney, and Paudyal (2008)andDeJongetal.(2008)indicating that the stock market has a significantly negative influence in the pooled model and income economy models. Our study supports the proposition that the institutional environment remains distinctive across Asian countries, with major influences on firms’financing decisions. Desai et al. (2004) and Chen and Strange (2005) found that economic growth rates, inflation, and Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 5 of 28 unemployment rates have significantly positive impacts on leverage. The variables are defined with corresponding measurements, and the data sources are summarized in detail in Table 2. 3. Data and empirical model 3.1. Data Our study sample covers 16 countries 1 across Asia, including China, Jordan, Malaysia, Thailand, Turkey, India, Indonesia, Pakistan, Philippines, Sri Lanka, Hong Kong, Japan, South Korea, Kuwait, Saudi Arabia, and Singapore, from the World Bank classification of 2008–2014. The unit of analysis in our study is a single firm, and 100 firms are selected from each country (except for Kuwait (66 companies)), Jordan (78 companies) and Saudi Arabia (98 companies). Data for firm-specific variables, country-specific variables and leverage are collected from Capital IQ by Standard & Poor’s. The firms in each country are selected using the criteria of the highest market capitalization and the availability of respective firm data over the sample period of 7 years. A few Asian countries are excluded because their data availability is fewer than 100 firms or less than 7 years. We primarily contribute by classifying the effective determinants for three Income economies composed of 16 Asian countries as classified by the World Bank in contrast to previous studies in which countries are analyzed in pooled form only. The World Bank divides the Asian income economies into four income groups: low-income economies, lower middle income economies, upper middle-middle income economies, and high-income economies. Income is measured using the gross national income (GNI) per capita in U.S. dollars converted from local currency using the World Bank Atlas method. We use the classification provided by the World Bank’s definition of income economies (IEs) for 16 Asian countries and grouped into 3 IEs. We proceed to explore the differences in leverage choice when the 16 sample countries are pooled as one group compared to when the countries are segregated on an income economy basis. The World Bank segregation of income economies 2 for 16 Asian countries is presented in Table 3. 3.2. Empirical model for the impacts of firmand country-specific determinants on leverage Both firmand country-specific variables are regressed to examine the determinants of leverage choices for firms. To ensure the robustness of both firmand country-specific variables, four regression models are considered. Table 1. Study variables Vector Variables Related theories 1Vector of firm-specific control variables Tangibility Trade-off theory Profitability Trade-off theory Growth Pecking order theory Size Pecking order theory Tax Trade-off theory 2-Vector of country-specific control variables GDP growth rate Macroeconomic factors Inflation Macroeconomic factors 3-Vector of firm-specific test variables Non-debt tax shield (NDTS) Trade-off theory Liquidity Pecking order theory Dividend payout ratio Pecking order theory Business risk Agency cost theory Cash flow Agency cost theory Ownership structure Agency cost theory 4-Vector of country-specific test variables Size of the equity market Macroeconomic factors Size of the bond market Macroeconomic factors Size of the banking industry Macroeconomic factors Adapted from: Deesomsak et al. (2004)and Tarek I.Eldomiaty (2008) Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 6 of 28 Table 2. The study variables with definitions, measurements, and data sources Variables Definition Data source Measurement 1Leverage (LEV) Leverage (LEV ratio is calculated by dividing a company’s total liability by its stockholders’ equity to assess how much debt is being used by a company to finance its assets relative to the shareholder value). Graham, Leary, and Roberts (2015), Gaud, Hoesli, and André Bender (2007), Drobetz and Fix (2005) and Demirgüc¸ Kunt and Maksimovic (1994) Calculated Debt-toEquity ratio = Total liability/ shareholders’equity Firm-specific variables 2Tangibility (TANG) Tangibility is defined as fixed assets over the book value of total assets. Booth et al. (2001) and Rajan and And Zingales (1995) Capital IQ by standard & poor’sTangibility = Gross amount of fixed assets/ total assets 3Size (SZE) Firm size is defined as the natural logarithm of total sales. De Jong et al. (2008), Gaud et al. (2007) and Booth et al. (2001) Capital IQ by standard & poor’sSize = (Natural log) Ln of sales 4Profitability (PRT) Profitability is measured by normalizing a firm’s earnings before interest and taxes (EBIT) with total assets. De Jong et al. (2008), Deesomsak et al. (2004) and Booth et al. (2001) Capital IQ by standard & poor’sProfitability = EBIT/total assets 5Tax (TAX) Effective average tax rate for the year is directly extracted from Compustat Global. De Jong et al. (2008) Capital IQ by standard & poor’sTax = Total income tax/EBIT 6Growth opportunity (GRW) Growth opportunity is defined as the market value of total equity over the book value of total equity. Gaud et al. (2007), Booth et al. (2001), and Bevan and Danbolt (2004) Capital IQ by standard & poor’sGrowth = Market-to-book ratio = Market value of equity/book value of equity 7Non-debt tax shield (NDTS) The NDTS is measured by depreciation over the book value of total assets. Deesomsak et al. (2004), Song and Philippatos (2004) and Huang and Song (2006) Capital IQ by Standard & Poor’sNDTS = Depreciation/total assets (Continued) Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 7 of 28 Table 5. Cross-country summary statistics of country-specific variables This table presents the crosscountry summary statistics for all 16 countries from the World Bank data sources during 2008–2014 Country GRT INF SEM SBM SBI China 8.88 0.033 83.815 46.785 114.285 (1.238) (0.017) (23.270) (2.204) (9.551) Hong Kong 2.69 0.034 456.221 42.084 165.582 (2.861) (0.013) (65.030) (12.958) (26.264) India 6.93 0.095 80.022 35.679 45.705 (2.062) (0.029) (15.422) (1.899) (2.676) Indonesia 5.66 0.067 38.869 14.673 26.481 (0.625) (0.030) (7.688) (2.047) (2.689) Japan 0.21 0.000 76.749 237.387 176.985 (3.006) (0.009) (12.736) (21.553) (2.453) Jorden 3.74 0.052 135.782 0.000 74.019 (1.871) (0.021) (40.201) (0.000) (5.862) Korea 3.18 0.029 91.207 104.589 63.850 (1.750) (0.011) (9.029) (5.332) (11.417) Kuwait 1.56 0.047 87.026 0.000 103.555 (5.521) (0.022) (25.882) (0.000) (5.322) Malaysia 4.60 0.025 135.327 108.759 23.863 (2.847) (0.011) (15.482) (4.813) (12.051) Pakistan 3.07 0.112 22.557 28.228 22.236 (1.204) (0.041) (8.379) (2.259) (3.896) Philippines 5.21 0.044 61.241 31.458 55.066 (2.323) (0.015) (13.310) (1.469) (9.364) Saudi Arabia 4.43 0.052 74.951 0.000 99.645 (3.682) (0.019) (15.491) (0.000) (10.829) Singapore 5.12 0.035 153.741 57.064 100.391 (5.007) (0.021) (17.233) (2.556) (7.987) Sri Lanka 6.20 0.099 26.370 0.000 123.331 (2.358) (0.063) (8.189) (0.000) (15.732) Thailand 2.90 0.027 70.605 57.149 123.331 (3.224) (0.011) (14.081) (5.842) (15.732) Turkey 4.88 0.081 152.774 47.455 38.391 (5.372) (0.501 (42.589) (8.549) (8.219) F-ratio 3.159 143.94 105.51 377.72 131.45 (.000) (.000) (.000) (.000) (.000) Income Economies Lower middle 5.41 0.08 45.81 22.01 54.36 (2.183) (0.05) (24.44) (13.18) (37.44) Upper middle 5.00 0.08 92.94 52.02 74.81 (3.689) (0.25) (45.34) (35.40) (41.96) High-income 2.87 0.03 161.14 78.21 121.60 (3.986) (0.02) (143.11) (82.61) (41.03) F-ratio 6.287 15.90 14.30 8.29 24.78 (.003) (.000) (.000) (.000) (.000) Figures in parentheses are standard deviations. F-ratios are reported for comparisons of the means of the variables under study by ANOVA between countries and between income economies. The numbers in parenthesis associated with F-ratios are their p-values. Source: Author’s calculation Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 14 of 28 are associated with higher private sector debt ratios. In Table 5, we also find that bond markets with 78.21 and the banking industry with 121.60 are highly developed in case of high-income economies, followed by middle-income economies. A distinctive number of studies by Booth et al. (2001), Deesomsak et al. (2004), Song and Philippatos (2004), Bokpin (2009) and Fan, Titman, and Twite (2012) discuss international capital structures with firmand country-specific variables across countries, supporting our research findings. We use equity market, bond market and banking industry factors as proxies for the financial system in pooled and income economy systems and discover that lower middle-income economies have the highest mean leverage of 1.323 and high-income economies have the lowest mean leverage of 1.181 for the sample period 2008–2014. This result contrasts with the findings of Booth et al. (2001) who report that developing countries have considerably lower amounts of long-term debt. Research has proven that the impact of institutional differences is captured through country-specific determinants such as stock market orientation and developments in the banking industry and bond markets. Having documented the trends for the equity market role from the cross-country summary statistics in Table 5, we observe that the mean for stock market orientation is highest for highincome economies. From the cross-country summary statistics of macroeconomic variables in Table 5, we find a high degree of variation across all countries. Such as the highest mean GDP growth rate (GRT) is found for China at 0.097, while the lowest mean GRT is found for Japan at 0.004, we observe that the GRT is highest for the middle-income economy group with 0.72, and the lowest mean value is found for high-income economy with 0.01. The mean inflation value is highest for the lower middle-income economy is highest with 0.08, and the lowest value is found for the high-income economy group with mean 0.03. Our study supports the evidence reported by the International Financial Institution (IFI) showing that developments in the stock market represent a natural progression in a country’s development, indicating that economic institutions’development is accelerated when a country achieves a higher level of economic development. We also observe that the variation impact is also reduced when we consider the cross-income economy group summary; for example, the GDP growth rate (GRT) is highest for the middle-income economy group. Meanwhile, inflation is highest for the lower middle-income economy group and lowest for the high-income economy group. Stock market orientation, bank market orientation and bond market orientation are all lowest in the lower middle-income economies and highest in the high-income economies. 5. Empirical results and discussion In this study, the dataset of 16 Asian countries is considered in pooled data form. For further analysis 16 Asian countries are is divided into lower middle, upper middle and high-income economies to understand the cross-country differences and crossincome-economy group similarities (Table 1). To investigate our research questions, we advance to explore differences in leverage choices when 16 sample countries are pooled as one group compared to when the countries are segregated on an income economy basis. Our results add value to the existing literature when we will contribute by classifying effective determinants for three Income economies composed of 16 Asian countries as classified by the World Bank in contrast to previous studies in which countries’data are analyzed in pooled form only. Using the World Bank’s definitions of income economies (IEs), we classify 16 Asian countries into 3 income economies based on the GNP. We run firm-level random-effect regressions where leverage is the dependent variable, and firmspecific factors and country-specific factors are explanatory variables for each of the 16 countries in our dataset. The determining factors for leverage choice in Asian economies and countries from 2008–2014 are presented in Table 8. But initially, an examination of the collinearity diagnosis test (Table 6) and the correlation matrix of the sample data (Table 7) provides some insights about the robustness of our results. We check for multicollinearity by observing the Variance Inflation Factor (VIF) associated with explanatory variables and find that there is no serious issue of multicollinearity since the VIF values for all explanatory variables are not higher than 1.46 (Table 6). The Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 15 of 28 Table 6. Multicollinearity diagnostic test for the independent variables included in the models Model Unstandardized coefficients Standardized coefficients t Sig. Collinearity statistics B Std. error Beta Tolerance VIF (Constant) 1.190 .060 19.792 .000 TANG −.218 .031 −.064 −6.995 .000 .899 1.112 SZE .065 .003 .179 19.107 .000 .855 1.169 PRT −1.454 .106 −.128 −13.673 .000 .863 1.159 TAX .334 .059 .051 5.665 .000 .922 1.085 GRW .068 .004 .170 18.923 .000 .927 1.078 NDTS −.270 .195 −.015 −1.387 .165 .686 1.458 LIQ −.230 .007 −.298 −32.710 .000 .902 1.109 PAYR .023 .026 .008 .877 .380 .961 1.041 BRSK1 −.574 .166 −.031 −3.462 .001 .949 1.053 LFCFMCAP −.082 .024 −.030 −3.427 .001 .991 1.009 Ownership .135 .037 .033 3.683 .000 .955 1.047 GRT −.003 .004 −.007 −.837 .403 .943 1.061 INF .268 .102 .027 2.627 .009 .691 1.446 Dependent variable: Leverage Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 16 of 28 Table 7. Correlation matrix of leverage, which also proves the absence of multicollinearity LEV TANG SZE PRT TAX GRW NDTS LIQ PAYR BRSK LFCF Ownership Leverage 1 Tangibility −.109** 1 Size .226** .053** 1 Profitability −.101** −.031** .125** 1 Taxation .132** .026** .253** −.027** 1 Growth .144** −.044** .019* .215** .018 1 NDTS −.045** .220** −.057** .008 −.021* −.027** 1 Liquidity −.347** −.153** −.178** .114** −.088** .019* −.086** 1 Payout ratio .014 .050** −.025** .063** .094** .020* −.025** .036** 1 Business risk −.110** −.048** −.082** .137** .000 .081** .050** .007 −.006 1 LFCF −.010 −.010 .018 .082** −.008 .024* .001 .040** .015 −.006 1 Ownership .042** −.013 .068** .034 .030** −.005 −.034** −.002 −.091** −.015 .012 1 Superscripts ** and * indicate that the correlation is significant at 1% and 5% levels, respectively. Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 17 of 28 Table 8. Effects of Firmand Country-specific variables on Leverage for 16 countries using average annual data from 2008–2014 Variables Model 1 Model 2 Model 3 Model 4 LMIEs UMIEs HIEs LMIEs UMIEs HIEs Intercept 0.721*** 1.157*** 0.779*** 0.586*** 0.526*** 0.922* 1.432*** 0.048* (0.069) (0.178) (0.139) (0.141) (0.100) (0.507) (0.392) (0.493) TANG −0.227*** −0.249*** −0.169*** −0.444*** −0.234*** −0.165** −0.492*** −0.292*** (0.043) (0.044) (0.086) (0.076) (0.065) (0.086) (0.075) (0.066) SZE 0.071*** 0.069*** 0.062*** 0.120*** 0.083*** 0.064*** 0.145*** 0.115*** (0.006) (0.007) (0.011) (0.015) (0.008) (0.012) (0.019) (0.012) PRT −0.726*** −0.752*** −0.645*** −0.536*** −1.605*** −0.695*** −0.582*** −1.680*** (0.105) (0.106) (0.202) (0.148) (0.220) (0.202) (0.150) (0.219) TAX 0.365*** 0.353*** 0.349*** 0.512*** 0.304*** 0.290*** 0.506*** 0.319*** (0.051) (0.051) (0.096) (0.112) (0.067) (0.096) (0.114) (0.067) GRW 0.086*** 0.088*** 0.077*** 0.096*** 0.107*** 0.082*** 0.100*** 0.109*** (0.004) (0.004) (0.006) (0.008) (0.007) (0.006) (0.009) (0.007) NDTS −0.229** −0.149** −0.121** −0.308** −1.607*** −0.115** −0.643*** −1.560*** (0.109) (0.065) (0.059) (0.130) (0.289) (0.045) (0.235) (0.289) LIQ −0.141*** −0.144*** −0.187*** −0.104*** −0.125*** −0.177*** −0.117*** −0.125*** (0.008) (0.008) (0.016) (0.012) (0.012) (0.016) (0.013) (0.012) PAYR 0.027 0.028 0.012 −0.016 0.104*** 0.018 −0.029 0.125*** (0.024) (0.023) (0.049) (0.035) (0.037) (0.049) (0.035) (0.038) BRSK −0.362** −0.332** −0.227** −0.547*** −0.210** −0.268** −0.525*** −0.205** (0.138) (0.138) (0.108) (0.210) (0.107) (0.122) (0.201) (0.093) LFCF −0.007 −0.007 −0.041 0.013 −0.039 −0.027 0.011 −0.036 (0.019) (0.019) (0.043) (0.026) (0.030) (0.043) (0.026) (0.030) Ownership 0.120 0.131* 0.353*** −0.219* −0.228 0.252* −0.131 −0.225 (0.075) (0.076) (0.129) (0.120) (0.204) (0.132) (0.119) (0.204) (Continued) Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 18 of 28 Table 8. (Continued) Variables Model 1 Model 2 Model 3 Model 4 LMIEs UMIEs HIEs LMIEs UMIEs HIEs GRT - −0.605*** - - - −0.464*** −0.112*** −0.267** (0.208) (0.116) (0.055) (0.122) INF - 0.153 - - - 1.819*** −0.039 1.787** (0.119) (0.430) (0.124) (0.707) SEM - −0.334*** - - - −0.541*** −0.158 −0.249** (0.063) (0.166) (0.126) (0.113) SBM - −0.032 - - - 0.250* −0.412*** −0.231*** (0.038) (0.131) (0.091) (0.061) SBI - 0.145* - - - 0.201 −0.026 0.482** (0.079) (0.259) (0.147) (0.192) Observations 10,794 10,794 3500 3346 3948 3500 3346 3948 R-Square 0.17 0.17 0.17 0.18 0.18 0.17 0.21 0.19 Wald Chi-square 1128.35 1163.96 388.58 395.88 577.28 403.19 428.63 611.10 (.000) (.000) (.000) (.000) (.000) (.000) (.000) (.000) Breusch and Pagan LM test 10,244.38 10,221.45 2073.29 4126.54 3909.24 2074.80 3866.66 3817.23 (.000) (.000) (.000) (.000) (.000) (.000) (.000) (.000) The table presents the regression results of the effects of firmand country-specific variables on leverage using annual average data from 2008–2014 estimated from Eq 3.1–3.2. The data represent 16 Asian countries and 3 Asian income economies using average annual data from 2008–2014. The independent variable is leverage, and the regression coefficients are reported for the independent variables previously defined in Table 2. The Robust Standard errors are reported in parentheses. Superscripts ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively. Observations are the number of panel observations in the regressions. The Wald chi-square and Breusch and Pagan (1980) test results for random effects are reported for each model. The numbers in parenthesis associated with Wald chi-square and Breusch and Pagan test are their p-values. Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 19 of 28 results are presented in the following correlation matrix (Table 7), which shows that the pair-wise correlations of the variables generally do not appear to indicate any concern over the multicollinearity problem in estimating the regression. In this study, we study 11 firm-specific variables and estimates the relationships among all of them. Pearson correlation analysis provides an early sign that SZE, TAX GRW, and Ownership are positively significantly related to leverage and are significantly different from zero at the 1% level. On the other hand, the variables of TANG, PRT, NDTS, LIQ, and BRSK are negatively significantly related to leverage at a 1% level. Among the explanatory variables, highest (and significant) correlation is between firm size and taxation such that their correlation coefficient is 0.253 (p-value 0.000). And to a lesser extent, NDTS is correlated with TANG (0.220; p-value = 0.000). We believe that this level of correlation between independent variables may not pose any possible threat of multicollinearity for our further regression analysis. 5.1. Country-wise and income-economy based impacts of firm-specific determinants on leverage The results reveal that tangibility, profitability, non-debt tax shield, liquidity, business risk, and GDP growth rate have significant negative effects on firms’leverage across all four models. However, firm size, taxation, and business growth have significant positive impacts on leverage. Nevertheless, certain factors showed partial effects over leverage, indicating that these are not significant in pooled models (Models 1 and 2), whereas their effects are significant in certain individual models based on income economies. For example, the effects of payout ratio and ownership are shadowed in the pooled models, whereas in the Income economy-based segregated models (3 and 4), these factors showed their differentiated impacts. From pooled model 2, the stock market and bond market have significant negative effects on leverage. However, the effects of stock market development are not apparent in the case of upper middle-income economy in segregated models. The size of the banking industry (SBI) showed a significant positive relationship with leverage in pooled model 2. Interestingly, the role of bank market is quite apparent in the high-income economy only. Ownership is positively related to leverage in model 2 (combined micro and macro factors). A positive relationship with leverage implies that firms in countries with concentrated private ownership show more leverage usage compared to those in countries with more publicly owned firms. The positive correlation between leverage with tangibility is supported by Rajan and And Zingales (1995) and De Jong et al. (2008) in their studies across countries. Nevertheless, our research finds that tangibility is negatively associated with leverage across four models because firms’asset tangibility is mainly associated with agency costs of debt according to agency theory (Jensen & Meckling, 1976). Risky firms prefer to use fewer fixed assets, thus restraining managers from using more debt than the optimal level. Our results advocate the findings of Booth et al. (2001) regarding the negative relation of tangibility in their study of 10 countries. Research over time indicates that the value and risk of a firm’s assets are not the only determinants of the level of borrowing and that certain types of assets also have a significant impact. Therefore, the companies holding more tangible assets with extensive secondhand markets are expected to borrow less than those holding more valuable or intangible assets. Furthermore, the use of collateral plays an important role in countries with relatively weaker creditor protection (Vinh.Vo, 2017); thus, emerging countries can be rationally accepted as part of the group of countries with weak credit protection. Since the need for collateral is more pronounced in traditional bank lending, the role of asset tangibility is expected to be more prominent in bank-oriented economies. Larger firms tend to have higher leverage, coherent to our study which is also proving the positive impact of firm size on leverage, thus supporting the notion of Rajan and And Zingales (1995) and M’ng et al. (2017) in their research on Thailand, Malaysia and Singapore implying that larger firms in terms of assets tend to have higher leverage, and that larger and more diversified firms face a lower default risk. Our finding regarding firm size is consistent with those of Deesomsak et al. (2004), Huang and Song (2006) and Shahjahanpour, Ghalambor, and Aflatooni (2010) according to trade-off theory. Pursuant to pecking order theory, profitable firms choose to Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 20 of 28 use internally generated funds and lower leverage (Myers & Majluf, 1984). We find negative and significant results in all of the models, which are supported by key studies such as Al-Sakran (2001), Chen (2004), Gaud et al. (2007) and Alves and Ferreira (2011). We also find a positive and significant impact of taxation on leverage across all four models for Asian countries and three economies, confirming the hypothesis proposed by Fan et al. (2012) in which they suggested that negative significant values arise because of dividend relief tax systems, whereas positive values arise because of classical tax systems; for such countries, the value of the tax gain from leverage is positive (De Jong et al., 2008). In our case, only Thailand and Turkey are found to be following dividend tax relief systems, and their negative relation has been overshadowed by those countries with positive gains. Contrary to the results of Booth et al. (2001), we find a significant relation between debt ratios and tax policy for developing Asian countries. Our study finds that firm growth positively and significantly impacts leverage because highgrowth firms use more external borrowing, which is well supported by the pecking order theory. Investment in high-risk projects is backed by the fact that large firms have higher growth opportunities, which eventually increase their likelihood of bankruptcy. Additionally, creditors may be unwilling to lend the funds at lower rates because the expected growth may decrease to 0%. Eventually, creditors will be unwilling to lend funds at low rates. Myers and Majluf (1984), Deesomsak et al. (2004) and Gaud et al. (2007) have empirically shown the positive relationship between leverage and firm growth. For developed countries and high-income economies, we find a stable and linear relationship between firm leverage and the NDTS because with more securable assets when leverage is increased, the NDTS is a contributive variable for securing a firm’s assets. Our study, in conjunction with Huang and Song (2006), indicates that the NDTS has a negative influence because firms with a high NDTS enjoy a tax benefit, thus negatively influencing leverage. Deesomsak et al. (2004) estimated the negative coefficients for the non-debt tax shield (NDTS) statistically significant for Thailand, Malaysia, Singapore, and Australia, thus supporting tax-based capital structure theories. In the context of a highly perceived bankruptcy risk and the rising cost of borrowing, the NDTS shows greater relevance to the leverage decision. Companies in Switzerland also support the trade-off theory in demonstrating the negative relationship between leverage and NDTS (Drobetz & Fix, 2005). Likewise, the coefficient associated with liquidity is negative and significant. According to the Pecking order theory, when Asian firms use their internal funds with decreasing levels of external financing, liquidity will shrink (Vo, 2017). Our regression verifies the results of Deesomsak et al. (2004), Mazur (2007), Viviani (2008) and Shahjahanpour et al. (2010). Thus, firms with less liquid equity employ more debt in their capital structures. The modified turnover exhibits negative and significant coefficients. Lower turnover implies less liquidity because firms with low liquidity carry more debt. Finally, when we measure liquidity using the modified liquidity ratio, we find that the coefficient of the modified liquidity ratio shows a significant negative impact on a company’s leverage (De Jong et al., 2008; Udomsirikul, Jumreornvong, & Jiraporn, 2011). The probability of financial distress is defined as business risk, which is assumed to be negatively related to leverage according to trade-off theory. Business risk is negatively related to debt because earnings are volatile when the environment is uncertain (Maria, Petr, & Anna, 2010). Our study shows that business risk is negatively and significantly correlated with leverage, supporting the study of 45 countries by Cheng and Shiu (2007) indicating that a firm with a lower business risk will use higher debt. Regarding the negative effect of business risk on leverage, our results are consistent with the findings of Huang and Song (2006) and De Jong et al. (2008). Trade-off theory indicates that firms with a relatively higher business risk prefer to reduce debt in their capital structures. The likelihood of defaulting on debt increases with a high business risk, causing an increase in financial distress costs. The coefficient for business risk is negative for six countries in Booth et al. (2001), a study which includes four countries contained in our dataset: South Korea, Pakistan, Thailand, and Turkey. Although dividends as a significant variable are not included in our final model, the payout ratio has a significant positive impact on the debt–equity ratio only for HIEs. We suggest that the companies in such economies have strong financial health and therefore can afford to pay Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 21 of 28 dividends. Empirical research by Chang and Rhee (1990) supports that firms with high payout ratios are likely to borrow more than firms with low payout ratios. Moreover, Beattie, Goodacre, and Smith (2004) and Frank and Goyal (2009) argue that companies will pursue debtraising options to support their dividend payouts. In the expansion stage of economy development, companies prefer to finance ventures with debt while assuming the risk. Ownership is among the factors that have partial and prominent effects on leverage in the income economy-based models (3 and 4), as shown in Table 8. Ownership shows a significant positive impact on leverage in the case of lower middle-income economies (LMIEs) and upper middle-income economies (UMIEs), implying that privately owned firms in LMIEs prefer to use debt. Authors observe that there are more privately owned firms in LMIEs that prefer to use debt rather than equity. This finding is complemented by the fact that bonds (as a private source of funding) also have a positive and significant effect on leverage. We find a negative and significant impact of ownership on leverage in UMIEs. As discussed, the negative sign signifies the concentration of publicly owned firms in those countries that belong to the UMIEs group. China has the highest proportion (77%) of the publicly owned firms concentrated in the telecommunication, transportation, chemical, energy, utilities, and construction sectors. The high proportion of publicly owned Chinese companies in this UMIEs group overshadows other countries in the same group that have zero percentage of public ownership, notably Jordan and Turkey. State-owned enterprises (SOEs) are favored by Government policies despite reductions in their formal privileges. Such policies provide SOEs with unlimited access to loans from state-administered banks (Attaoui & Poncet, 2011; Guariglia & Yang, 2016). In key strategic industries, some SOEs also hold important monopolistic positions (Allen, Qianb, & Qian, 2005; Ding, Guariglia, & Knight, 2013). 5.2. Country-wise and income-economy-based impacts of country-specific determinants on leverage Thus far, we have discussed the effect of leverage on firm-specific determinants in both pooled models and income economy-based models. Next, we investigate the impact of country-specific factors with respect to income economy groups and pooled models. The variables of GDP growth rate (GRT), inflation (INF), size of the equity market (SEM), size of the bond market (SBM), and size of the banking industry (SBI) are regressed over corporate leverage for the 2008–2014 period. Cheng and Shiu (2007), in their study of 45 countries, show that Ln GDP (a proxy for the GRT) has a significant negative coefficient, indicating that firms in wealthier countries have less leverage than those in poorer countries. Our findings in Table 5also indicate that high-income economies have lower leverage with a mean of 1.181, whereas lower middle-income economies have the highest mean of 1.323. At all the stages of economic development, financial development improves capital allocation, boosts aggregate growth, and helps the poor through this channel. However, the distributional effect of financial development and thus the net impact on the poor depends on the level of economic development. Our regression results confirm the direct impacts of several country-specific factors on corporate leverage across pooled models and income-economies models unlike the indirect impacts of country-specific determinants on leverage by De Jong et al. (2008). Bokpin (2009) proposed that the effects of country-specific determinants on capital structures are subject to the choice of measuring the capital structure in the case of most countries. The GDP growth rate (GRT) among other factors such as creditor rights protection and bond market development, the GRT consistently shows a statistically significant impact on capital structure (De Jong et al., 2008). Following De Jong et al. (2008), we use the GRT to analyze the effects of countries’economic conditions. Our results for the GDP growth rate (GRT) show a negative relation with debt because debt financing becomes more attractive as the inflation rate increases, since corporations’real tax shelters attributable to interest deductions increase as inflation increases. The GRT is an aggregate of the magnanimity of a given country, thus propounding healthy investment opportunities for investors in that country and supporting investors with good growth opportunities. Our findings are consistent with those of Bokpin (2009) and Kayo and Kimura (2011) indicating that the GRT has a Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 22 of 28 negative and statistically significant relationship with corporate leverage. We find that inflation (INF) is positively and significantly related to leverage in the cases of lower middle and highincome economies. The size of the equity market (SEM) indicates the size of the stock market relative to the size of the economy. Cheng and Shiu (2007), De Jong et al. (2008) and Kayo and Kimura (2011) find that stock market orientation is an important variable used to evaluate the impacts of macroeconomic variables on capital structure. The SEM influences the tendency to issue equity rather than debt. Firms in countries with strong capital markets have comparatively easy access to equity funds; therefore, a negative relation is expected between the SEM and leverage. We find a significant negative impact on leverage in cases of lower middle and high-income economies, implying that leverage decreases when capital markets are improved. Investors begin to use observable return capital production technology under weak conditions, which is correlated with equity issues. Therefore, a voluminous equity market increases the economic growth, that allows a lower debt–equity ratio over time, thus showing a typical pattern of development. Our study supports the fact that high levels of equity market activities are supported by a lower aggregate ratio of debt to equity that is associated with an increase in per-capita output. The stock market is a vehicle for diversifying risk when firms are also using banks or other financial intermediaries as sources to meet their financial requirements (Antoniou et al., 2008; De Jong et al., 2008). For investors, two types of complementary financial services exist: risk diversification and information processing. Therefore, we can say that the leverage of firms may increase with a better-developed stock market because developed stock markets help individuals to price and diversify risks more easily. It is well known that with developed stock market owners have more diversified investment opportunities, and they can also take projects that otherwise would not have been practicable. The study of 42 countries by De Jong et al. (2008) mentions that trading bonds are easier when the bond market in a country is well developed, causing firms to have high leverage. In our study, we use the size of the bond market (SBM) to capture the financial depth of the financial sector relative to the economy. In addition to the bond market, creditor rights protection and GDP growth rates (Antoniou et al., 2008; Booth et al., 2001; Kayo & Kimura, 2011) show significantly positive relation with corporate leverage. Based on cross-country macro statistics, we find an overall increasing trend in the Bond market. In pooled model 2, we find that the SBM is not significantly showing the impact on leverage because those countries are in the majority in the pooled model where bond markets are not well developed. However, when we consider the segregated models based on Income economies, we find that the bond market is positively and significantly related to leverage in lower middle and upper middle-income economies. This indicates that firms in this group of countries prefer to use bond market credit rather than bank market credit for long-term financing to remain unaffected by yearly economic fluctuations (Tomschik, 2015). The development of bond markets arguably improves legal systems, thus mitigating agency problems and protecting debt holders. Highly developed bond market in a country allows firms to achieve a high leverage ratio; however, contrary to this proposition, we find that bond markets in the upper middle and high-income economies of Asia have an inverse effect on leverage. Similarly, we know that Beck et al. (2004) suggest that bond market development is positively related to the development and bank finance for large firms because lower expected return projects are conditioned to less frequent verification and are usually aimed at firms in mature economies. The size of the banking industry (SBI) represents the financial depth of sector relative to the economy. Private credit issued by deposit money banks and other financial institutions to the GDP is a standard indicator in the finance and growth literature. Cheng and Shiu (2007) confirm that countries with powerful banking sectors can issue more loans, indicating a positive correlation between bank loans and the debt ratio. Countries with higher levels of private credit relative to the GDP have shown faster growth, with positive effects on poverty reduction (Beck, Demirgüç-Kunt, & Levine, 2000). The banking industry is positively and significantly affecting leverage in pooled model 2. We can say that better availability of private credit through banks is associated with Zafar et al., Cogent Economics & Finance (2019), 7: 1598836 https://doi.org/10.1080/23322039.2019.1598836 Page 23 of 28