*3.1. Database Construction and Variables Presentation*

The dependent variables included in the analysis are included in Table 1. I chose three dependent variables, which explain to a certain extent the financial structure, namely, the rate of total debt, the rate of long-term and short-term debt. These debt ratios show which proportion of assets are financed by debt. A high value of these ratios reveals the leverage of the company, and also the financial risk. Debt ratios vary across industries, businesses with intensive capital such as transportation sectors or telecommunications have higher debt than other industries such as technology sector.


**Table 1.** Dependent variables included in the empirical study.

Source: Author's own work.

Estimating separate relationships for long-term and short-term debt rates (long-term and short-term debt over total assets) allows for an influence on the maturity of the debt structure as well as the leverage. Total assets are included as a size variable to test scale effects in the ratio of debt to total assets.

Table 2 below shows the independent variables classified according to the level of influence, namely, microeconomic, macroeconomic and corporate governance indicators.

(Barton et al. 1989; Titman and Wessels 1988) agreed that companies with high rates of profit will maintain a low rate of debt, because they are able to generate funds from internal sources, so the profitability indicator was included as a variable. Companies with very high growth rates will seek external sources of funding to support their growth rate. Auerbach (1985) also argues that the leverage is inversely proportional to the growth rate, because the tax deduction of interest expense is not significant for fast-growing firms. Michaelas et al. (1999) found a positive expected growth related to leverage and long-term debt, while (Chittenden et al. 1996; Jordan et al. 1997) found mixed evidence. Graham (1996) concluded that, in general, taxes affect the financial decisions of enterprises, but the impact is not major. Myers (1977) argues that tangible assets, such as fixed assets, can support a higher level of debt compared to intangible assets. Assets can be used as collateral to reduce potential agent costs associated with borrowing (Smith and Warner 1979; Stulz and Johnson 1985). The size of the company plays an important role in determining the financial structure of a company. Researchers have found that large firms are less likely to go bankrupt because they tend to be more diverse than smaller companies (Ang et al. 1982; Marsh 1982; Smith and Warner 1979; Titman and Wessels 1988) report a negative relationship between the debt and the size of the firms. Marsh (1982) argues that small firms, due to their limited access to the capital market, tend to rely heavily on loans. Titman and Wessels (1988) argue that small firms are less reliant on equity because they may face a higher cost per issue unit. Ooi (1999) argues that firms with relatively higher operational risk will have incentives to have a lower leverage than firms with more stable incomes. Öztekin and Flannery (2012) have observed that firms that have more liquid assets can use them as an internal alternative of funds instead of debt. I included four macroeconomic indicators, GDP per capita, inflation rate, interest rate, and market size in the study (Bartholdy and Mateus 2008; Demirgüç-Kunt and Maksimovic 1996, 1999). I also included the GDP per capita because with its growth, the countries become richer and implicitly there are more financing resources. Thus, I expect this indicator to be positively correlated with debt. Inflation provides a perspective on the stability of the national currency. Countries with a high inflation rate are associated with a high degree of uncertainty. In general, loans are nominal value contracts, and the inflation rate influences the value of loans, making them riskier. I expect the inflation rate to be negatively correlated with debt. When the interest rate increases, companies are no longer willing to resort to bank loans, because the cost of the loan is higher. Therefore, I expect the interest rate to be inversely proportional to the debt. The size of the market was included because it indicates how easy it is to access the market. The corporate governance indicators were also included in the empirical study to see if they influence the financial structure. Vintilă and Gherghina (2012) obtained mixed results regarding the relationship between the size of the board and the performance of the company. The paper had a database of 155 US companies listed from different industries and investigated the relationship between corporate governance mechanism, CEO characteristics and

company performance. It turned out that the number of board members is in a negative relationship with Tobin's Q, but in a positive relationship with ROA. From the point of view of the status of the CEO, no results have been obtained that suggest a relationship with the performance of the company, whether or not he is chairman of the board. Therefore, I expect that the size of the board, the status of the CEO and the existence of the committees will not influence the financial structure.


**Table 2.** Independent variables included in the empirical study.

Source: Author's own work.

The general objective is to analyze the factors which have an influence on the financial structure. Firstly, I will start from a set of hypotheses, which will be tested afterwards, in accordance with the studies mentioned above.

**Hypothesis 1 (H1).** *There is a positive relationship between size and debt ratio (Chaklader and Chawla 2016; Cortez and Susanto 2012; Song 2005).*

**Hypothesis 2 (H2).** *There is a positive relationship between tangibility and debt ratio (Chaklader and Chawla 2016; Cortez and Susanto 2012; Song 2005).*

**Hypothesis 3 (H3).** *There is a negative relationship between growth opportunity and debt ratio (Alipour et al. 2015; Cortez and Susanto 2012; Psillaki and Daskalakis 2009).*

**Hypothesis 4 (H4).** *There is a negative relationship between liquidity and debt ratio (Alipour et al. 2015; Chaklader and Chawla 2016).*

**Hypothesis 5 (H5).** *There is a positive relationship between the tax rate and debt ratio (Alipour et al. 2015).*

**Hypothesis 6 (H6).** *There is a negative relationship between profitability and debt ratio (Alipour et al. 2015; Chaklader and Chawla 2016; Cortez and Susanto 2012; Nenu et al. 2018).*

**Hypothesis 7 (H7).** *There is a negative relationship between inflation rate and debt ratio (Bokpin 2009; Chadha and Sharma 2015; Demirgüç-Kunt and Maksimovic 1999).*

**Hypothesis 8 (H8).** *There is a negative relationship between interest rate and debt ratio (Bartholdy and Mateus 2008; Chadha and Sharma 2015; Demirgüç-Kunt and Maksimovic 1999).*

**Hypothesis 9 (H9).** *There is a positive relationship between GDP and debt ratio (Demirgüç-Kunt and Maksimovic 1996).*

**Hypothesis 10 (H10).** *There is a negative relation between board size and debt ratio (own consideration).*

**Hypothesis 11 (H11).** *There is a negative relation between presence of audit committee and debt ratio (own consideration).*

**Hypothesis 12 (H12).** *There is a negative relation between presence of nomination committee and debt ratio (own consideration).*

**Hypothesis 13 (H13).** *There is a negative relation between presence of remuneration committee and debt ratio (own consideration).*

**Hypothesis 14 (H14).** *There is a negative relation between CEO Status and debt ratio (own consideration).*

#### *3.2. Econometric Framework*

The influence factors were studied based on multiple regression model, using the method of least squares, data being structured as panel type:



*Financial\_structurei,t* = α<sup>0</sup> + α<sup>1</sup> × *Growthi,t* + α<sup>2</sup> × *PERi,t* + α<sup>3</sup> × *Stocki,t* + α<sup>4</sup> × *Sizei,t* + α<sup>5</sup> × *Etaxi,t* + α<sup>6</sup> × *C\_ri,t* + α<sup>7</sup> × *Boardi,t* + ε*i,t* (*Model 3*)

where *Financial\_structure* = TD, LTD, STD; α<sup>0</sup> = constant; α<sup>1</sup> ... α<sup>7</sup> = coefficients of the parameters; ε = error term; *t* = 2005 ... 2018; *i* = 1, 2, ... , 51.

The regression models are built based on the correlation matrix. The corporate governance variables are strongly correlated with each other and they were separated in three models. This situation is similar for the size indicator, which was calculated once as natural logarithm from total assets and once as natural logarithm from turnover. A macroeconomic indicator was also included in each model.
